Professional Documents
Culture Documents
Eric Quémerais
Martin Snow
Roger-Maurice Bonnet
Editors
Cross-Calibration
of Far UV Spectra of
Solar System Objects
and the Heliosphere
ISSI Scientific Report Series
Volume 13
Cross-Calibration of Far UV
Spectra of Solar System
Objects and the Heliosphere
123
Editors
Eric Quémerais Martin Snow
LATMOS-IPSL Laboratory Atmospheric & Space Physics
Université Versailles-Saint-Quentin University of Colorado
Guyancourt, France Boulder, CO, USA
Roger-Maurice Bonnet
International Space Science Institute
Bern, Switzerland
Detecting and measuring the Sun’s ultraviolet emission absorbed by the Earth’s
atmosphere has been one of the earliest scientific objectives of space research and
continues, even more than 50 years after the launch of the first sounding rockets
and artificial satellites, to be of interest to several space research programs and mis-
sions. This is essentially because the electrical and chemical balance of planetary
atmospheres is directly determined by the strongly variable ultraviolet radiation
from the Sun below 300 nm. Of particular interest is the presence in the solar
spectrum of the intense emission of the resonance Lyman-α line of atomic hydro-
gen at 121.6 nm and its equivalent from neutral helium at 58.4 nm. These strong
resonance emissions from the two most abundant atoms in the universe are at the
origin of the interplanetary ultraviolet background, which results from light scat-
tering by these atoms present in the interplanetary medium and the heliosphere.
Measuring with the best precision this background emission allows a study of the
penetration of interstellar gas from the Milky Way into the solar system, thereby
offering a unique tool to study the dynamics of our solar system as it moves through
its mother galaxy.
Measuring ultraviolet radiation in absolute terms is one of the most delicate
and difficult tasks for experimenters involved in solar and planetary projects. In
most cases, ultraviolet instruments are well calibrated on the ground but, unfortu-
nately, optics and detectors in the FUV are very sensitive to contaminants and it
is very challenging to prevent contamination before and during the test and launch
sequences of a space mission. Therefore, ground calibrations need to be confirmed
after launch and it is necessary to keep track of the temporal evolution of the
sensitivity of the instrument during the mission. This is even more important for
solar instruments directly illuminated by the Sun’s radiation, although this book
is primarily concerned with instruments designed for planetary and interplanetary
missions. Even more difficult is to reconcile several measurements made at dif-
ferent times with different instruments calibrated with different techniques, in the
laboratory before launch or in orbit, using celestial objects selected as radiation
standards.
Although the International Space Science Institute (ISSI) is mostly involved in
the analysis of data obtained from space missions, it has naturally devoted many
of its working tools to the calibration of various datasets in an attempt to estab-
lish common standards against which they might ultimately and quantitatively be
compared. For example, the former ISSI Working Group on the calibration of So-
lar Heliospheric Observatory (SOHO) instruments (The Radiometric Calibration
of SOHO, A. Pauluhn, M.C.E. Huber & R. von Steiger Eds., ISBN 1608-280X,
2002), was a successful endeavour to compare on a standard scale the intensities of
various solar phenomena observed by SOHO with different ultraviolet and X-ray
detectors.
Over the past seventeen years of ISSI’s existence, heliospheric research has been
one main area of investigation by this institute, as is illustrated by the list of ISSI
books published so far in the ISSI Space Science Series. These publications are
V
VI Foreword
accessible through the ISSI web page. Indeed, several of the authors of this study
have been involved one way or another in these previous ISSI activities.
Back in 2008, the main author of this project, E. Quémerais, responded to ISSI’s
annual call for proposals to convene an international team to study the possible
ways to calibrate far ultraviolet instruments after their launch. His proposal’s goal
was to gather as much information as possible on the best approaches for calibrating
future experiments. The ISSI Science Committee in the course of its review process
did not consider that proposal appropriate for the “International Team” category.
An ISSI “Team” offers only a few opportunities to work over one to two years
with the outcome of the team’s effort published in the peer-reviewed literature in
one or several usually unrelated articles dedicated to the subject. With the strong
support of ISSI, however, the Science Committee approved the formation of an ISSI
Working Group dedicated to the establishment of a defined set of standards for the
calibration of ultraviolet instruments on heliospheric and planetary missions. This
book published in the ISSI Scientific Report Series represents the outcome of the
working group.
Hopefully, more heliospheric and planetary instruments will be launched in
the future, acquiring more data with better detectors and benefiting from better
calibration techniques. A natural response of ISSI to such progress would be to
consider this work as just the beginning of a long-term effort aimed at a more
precise understanding of the phenomena, which characterise the interaction of our
solar system with its home galaxy. Therefore, this book could be considered to be
the first of a series dedicated to such an ambitious goal.
Foreword V
1 Introduction 1
Index 269
VII
—1—
Introduction
In this book we discuss the problems encountered for calibrating space-borne
instruments in the spectral range 50–300 nm. We also discuss the level of accuracy
that can be achieved, and review the different sources of emissions that can be
used in that range, i.e. the interplanetary background, stars, and solar system
bodies such as planets, moons, and comets. As most of the work presented here
is based on comparisons with older datasets obtained by experiments that have
been calibrated independently, we also analyse the consistency between existing
datasets. Whenever possible, we suggest a correction factor if an older dataset
is in disagreement with a more recent one. For example, a comparison of the
interplanetary background data acquired by the two ultraviolet (UV) spectrometers
on Voyager 1 and Voyager 2 with models and other more recent datasets shows that
it is indeed necessary to derive new calibration factors for these two instruments.
The interplanetary ultraviolet background emission was first observed at the end
of the 1960’s and is due to the backscattering of solar photons by interstellar hydro-
gen and helium atoms present in the interplanetary medium. The corresponding
resonance emissions (Helium at 58.4 nm and Hydrogen at 121.6 nm) can be observed
in every direction of the sky as long as the observer is sufficiently away from plan-
etary exospheres. For planetary missions, it is often possible to observe the inter-
planetary emission during their long cruise. During the last three decades, several
ultraviolet instruments have indeed observed these emissions. Hence, the possibil-
ity of using the interplanetary ultraviolet background to calibrate FUV instruments
in particular at the Hydrogen Lyman-α wavelength is of great importance. This
is a difficult task because the interplanetary Lyman-α emission is highly variable,
both spatially and temporally. The measured brightness strongly depends on the
position of observation in the heliosphere. It also depends on the direction of the
line of sight in the sky, mainly as a function of the angle between the line of sight
and the direction of the Sun. The interplanetary background emission also varies
with solar activity. Therefore, highly detailed models are necessary to represent all
these variations accurately.
The first section of the book (Chaps. 2 and 3) is logically dedicated to the
establishment of an accurate model of the spatial distribution of neutral interstellar
hydrogen in the heliosphere. Chapter 2 gives an overview of the main physical
phenomena that need to be taken into account for developing such a model based
on the interaction between the solar wind and the Local Interstellar Medium. It
presents a state-of-the-art 3D kinetic model of the interstellar hydrogen gas inside
the heliosphere based on the various inputs described in Chap. 3, in particular the
solar factors that are responsible for shaping the distribution of neutral interstellar
hydrogen in the heliosphere. This concerns the solar Lyman-α flux and the resonant
radiation pressure force acting on neutral hydrogen atoms in the heliosphere, as well
as the process of photoionization of heliospheric hydrogen, and how these evolve
in time and with heliolatitude. Chapter 3 also deals with changes in the neutral
hydrogen ionization process by charge exchange with protons of the solar wind,
and also with changes in electron impact ionization, induced by solar activity. The
chapter considers their variations with time, heliolatitude and solar distance.
Based on these studies, the heliosphere can be divided into two regions. In
the inner heliosphere, the hydrogen distribution is dominated by the influence of
the Sun. There, the interplanetary background is modulated by the changes in
radiation pressure, solar wind flux and solar wind ram pressure. In the outer he-
liosphere, the hydrogen distribution is dominated by the influence of the interface
region between the expanding solar wind and the ionized part of the local inter-
stellar medium.
Chapter 4 compares the results obtained with models, such as those presented
in Chaps. 2 and 3, to the different datasets so far obtained in the outer helio-
sphere with the Voyager ultraviolet spectrometer (UVS), ALICE New Horizons,
and in the inner heliosphere Solar Wind ANisotropies (SWAN) instrument on the
Solar Heliospheric Observatory (SOHO), and the Space Telescope Imaging Spec-
trograph (STIS) on the Hubble Space Telescope (HST) instruments. Because these
datasets do not necessarily overlap, models are used to bridge the gaps in distance
or time. Combining these datasets with the model results allows one to derive
calibration factors that give a coherent picture of the photometric behaviour of
the various instruments. For example, the STIS calibration of interplanetary back-
ground observations is based on International Ultraviolet Explorer (IUE) stellar
observations as well as on those obtained with SPectroscopy for the Investigation
of the Characteristics of the Atmosphere of Mars (SPICAM) on Mars Express, as
described in Chap. 7. SWAN is calibrated at Lyman-α through direct comparisons
with interplanetary background measurements made by STIS. Therefore the SWAN
calibration ultimately derives from the IUE stellar measurements. The model of
Chap. 2 is used to fit the observations of SWAN and its extrapolation to the outer
heliosphere, allowing a comparison with the observations of ALICE New Horizons
and of the UVS-Voyager background measurements. This example illustrates how
all these successive operations lead to a coherent calibration ensemble originating
from the IUE stellar observations.
Chapter 5 presents a comparison of three datasets obtained in the inner he-
liosphere where the influence of the solar parameters is most important. Some
systematic discrepancies between two of the datasets (SWAN SOHO and MASCS
MESSENGER) can be alleviated by changing the calibration of one of the instru-
ments. This demonstrates that comparison of interplanetary background data can
be used to inter-calibrate UV instruments.
Chapter 6 presents a new set of observations shown here for the first time
and performed by the ALICE New Horizons instrument between 9 and 17 AU
from the Sun. The Voyager 1 and Voyager 2 spacecraft visited this region of
the heliosphere nearly 30 years before the New Horizons spacecraft. The inter-
comparison of the Voyager and New Horizons datasets is therefore important for
our current study because the interplanetary background is rather easy to model
Introduction 3
Olga A. Katushkina
Space Research Institute (IKI) Russian Academy of Sciences,
Moscow, Russia
Lomonosov Moscow State University, School of Mechanics and Mathematics,
Moscow, Russia
Eric Quémerais
LATMOS-IPSL,
Université Versailles-Saint Quentin, Guyancourt, France
Maciej Bzowski
Space Research Center,
Polish Academy of Science, Warsaw, Poland
Abstract
We review the modern concepts of penetration of interstellar atoms of hydrogen
into the heliosphere up to 1 AU. Before entering into the heliosphere the atoms
penetrate through the region of the solar wind (SW) interaction with the local in-
terstellar medium (LISM). In the interaction region the atoms can exchange charge
with both solar wind and interstellar protons disturbed in the SW/LISM interac-
tion region. Charge exchange results in a disturbance of the pristine interstellar
atom flow in the interaction region, and, therefore, the parameters of interstellar
gas inside the heliosphere are different from their interstellar values. This makes
it more difficult to determine local interstellar parameters from measurements of
the interstellar atoms inside the heliosphere, but, on the other side, opens possi-
bilities to study the SW/LISM interaction region remotely. This paper overviews
the main physical phenomena and modern models of the SW/LISM interaction
and presents a state-of-art 3D kinetic model of the interstellar hydrogen gas in-
side the heliosphere. The distributions of the gas parameters are compared with
the distributions obtained in the context of the classical hot model. Quantitative
and qualitative differences are discussed. The state-of-art model is employed to
calculate spectra of the backscattered Lyman-α radiation as they would be mea-
sured at 1 AU and the zero, first and second moments of the spectra. It is shown
that the SW/LISM interaction imprints in the spatial and velocity distribution of
the interstellar atoms are revealed in the intensities, line-shifts, and line-widths
of the distribution functions. A qualitative comparison of the model results with
SOHO/SWAN data are presented.
Figure 2.1: A sketch of the structure of the interaction region between the solar
wind and the interstellar magnetic field proposed by Axford et al. (1963). The solar
wind is supersonic in Region I, and the solar magnetic field lines form Archimedean
spirals co-rotating with the Sun. The termination shock is located at a heliospheric
distance S. Beyond the termination shock, in region II (the boundary shell), solar-
wind protons and interstellar neutral hydrogen atoms exchange charge, and dissi-
pative effects permit oppositely-directed solar wind magnetic-field lines to merge
and form closed loops. S∗ is the distance to the boundary between the solar and in-
terstellar magnetic fields. Blobs of the solar plasma and magnetic field detach from
region II and move out into region III (interstellar space), where they gradually
diffuse away
velocity. Hence, the solar wind density and dynamic pressure are directed radially
away from the Sun and decrease as ∼ 1/r2 . At the distance where the dynamic
pressure (magnetic pressure and/or thermal pressure) equals the pressure of the
local interstellar medium (LISM), the solar wind passes through a standing shock—
the termination shock (TS), beyond which it slows down and becomes subsonic.
Nevertheless, it was assumed that in the region beyond the shock front (that we
call now the inner heliosheath), the individual velocities of solar wind protons re-
main of the order of ∼ 300 km/s due to high thermal velocities. The energetic
protons undergo charge exchange with neutral interstellar hydrogen. This process
leaves low-energy protons in the transition region beyond the shock, and provides
an isotropic source of high-energy neutral hydrogen atoms. Some of these atoms
(now called energetic neutral atoms—ENAs) move through the termination shock
(TS) back toward the Sun.
Patterson et al. (1963) calculated the distribution of the neutral hydrogen atoms
in interplanetary space assuming that the TS is an isotropic source of the neutrals,
taking into account their velocity, the ionizing solar radiation, and the charge
exchange with the solar wind. It was also assumed that (1) the post-shocked solar
10 2. Interstellar Hydrogen and Backscattered Lyman-α
wind protons suffer charge exchange with interstellar neutrals just in a thin shell
beyond the TS, (2) half of the solar wind flux is returned into the heliosphere
in the form of fast neutrals. Later, Hundhausen (1968) showed that most of the
neutral hydrogen observable in the vicinity of the Earth does not undergo charge
exchange between solar wind protons and interstellar neutral hydrogen near the
shock boundary of the heliosphere, as assumed by Patterson et al. (1963), but in a
region far beyond the shock. Hundhausen’s model predicted much smaller atomic
hydrogen density near Earth. Therefore, the neutral hydrogen density near Earth
could then only be maintained in the model if the shock is near 5 AU.
An alternative approach, that actually became a commonly accepted paradigm
for at least two decades, was suggested by Fahr (1968a). We should point out that
modern concepts of the ENAs originating in the inner heliosheath (Gruntman et al.
2001) are very similar to those of Patterson et al. (1963) and Hundhausen (1968).
Fahr (1968a) pointed out that the assumption of the random motion of inter-
stellar hydrogen made by Patterson et al. (1963) is very unlikely because the solar
system itself has a velocity of 20 km/s with respect to the local standard of rest
of nearby stars, and also interstellar hydrogen clouds are moving. It was shown
(Blum and Fahr 1970; Axford 1972) that the concept of the Strömgren sphere of
ionized gas around a hot star fails for the Sun, in case that there is a relative bulk
motion of the interstellar atoms with respect to the Sun. The Strömgren sphere
is determined by equating the flux of ionizing photons emitted by the Sun to the
total recombination rate (Strömgren 1939). In case of the Sun, the Strömgren
sphere has a radius of about 1,500 AU. It was estimated (Blum and Fahr 1970)
that about 90 % of the cold interstellar hydrogen with a velocity of 20 km/s will
enter the heliosphere without being photoionized or charge exchanged.
Fahr (1968a) has calculated the heliospheric distribution of the number density
of the cold interstellar hydrogen taking into consideration the macroscopic motion
of interstellar gas and the solar gravitational field. Blum and Fahr (1970) took into
account the losses of interstellar atoms due to charge exchange with the solar wind
protons and due to photoionization by solar radiation. Therefore, in the model, the
cold interstellar neutrals enter the solar system along Kepler trajectories and suffer
losses caused by the EUV-ionization and charge exchange with the solar wind
protons. The charge exchange process produces secondary fast neutrals having
the velocity of the solar wind. These particles move radially outwards without
significant losses within the solar wind (Fahr 1968b), and cannot be observed in
Lyman-α because they are Doppler shifted too far from the center of the solar line.
According to the prediction of the cold model of Blum and Fahr (1970), the
number density of interstellar H atoms are larger in the direction toward the
interstellar flow, and, therefore, the observed Lyman-α emission should have a
maximum, and the maximum should be toward the direction of the interstellar
H atom flow. Extraterrestrial UV radiation (most likely Lyman-α) was measured
from interplanetary probes Zond 1 (Kurt 1966, 1967), of Venera 2, 3, and 4 (Kurt
and Germogenova 1967; Kurt and Syunyaev 1968), and of Mariner 5 and 6 (Barth
1970). These measurements proved that the 15 % highly shifted night time Lyman-
α found earlier by Morton and Purcell (1962) in their rocket flight was indeed
of extra-terrestrial origin. It was found in these interplanetary probes measure-
ments that the backscattered Lyman-α emission is distributed inhomogeneously
Introduction: A Brief Historical Review 11
and features toward the direction of the galactic center. Because of this specific
direction of the emission maximum, a new “galactic” source of the diffuse radiation
was suggested as an alternative to Patterson et al. (1963) scenario.
The dilemma of the galactic or interstellar origin of the diffuse radiation was
resolved by OGO-5 spacecraft measurements (Bertaux and Blamont 1971; Thomas
and Krassa 1971; Lallement 2001). The apogee of OGO-5 was out of the geocorona,
i.e. the geocorona emission could be ruled out at apogee. The spacecraft was spin-
ning so that the extraterrestrial diffuse Lyman-α radiation could be mapped over
the sky. A broad maximum in one direction and a minimum in the opposite direc-
tion were observed. To find the distance to the region of the maximum emissivity
of the backscattered Lyman-α emission two sky-mappings of the diffuse Lyman-α
were performed on board the OGO-5 spacecraft in September 1970 and April 1970,
when the Earth (and the spacecraft too) was at opposite positions in its orbit. The
difference in the position of the maximum of diffuse Lyman-α emission as seen in
the two maps was by about 30◦ . This proved that the emission region was located
at few AU from the Sun and the displacement of the maximum was due to the
parallax effect that would be negligibly small in the case of a galactic source of the
emission. Analysis of three sky-maps of the OGO-5 allowed to find the direction of
the interstellar H atoms approaching the Sun (Bertaux et al. 1972). Newer values
of the direction of interstellar hydrogen inside the heliosphere are (252.5◦ , 8.9◦ ) in
the ecliptic J2000 coordinates (Lallement et al. 2010).
Therefore, the theoretical prediction made by Fahr (1968a) and Blum and Fahr
(1970), i.e. that interstellar H atoms penetrate into the heliosphere up to distances
of a few AU from the Sun, was confirmed by Lyman-α measurements made by
OGO-5 spacecraft instruments in 1969–1970 (Bertaux and Blamont 1971; Thomas
and Krassa 1971; Thomas 1972), and the “galactic” interpretation was ruled out.
An upper limit of galactic emission of 15 Rayleigh (compared to the observed 300–
500 R of the extra-terrestrial emission) was derived from hydrogen absorption cell
measurements on board Prognoz-5 and 6 (Lallement et al. 1984).
Since the interstellar temperature is ∼ 104 K, the real situation is drastically
different from the cold model considered above. The cold model fails for the down-
wind direction at the axis of symmetry, where it gives infinite densities in case when
the solar gravitation is larger than the radiation pressure (μ < 1). The quantity μ
is defined as the ratio of the solar gravitational attraction to the repulsion due to
radiation pressure and will be used throughout this document:
Actually, Danby and Camm (1957) were the first to solve the problem of the
motion of a cloud of gas particles in the gravitational field of a point mass. They
found an analytical formula for the velocity distribution function of the particles.
Later, the Danby–Camm formula was applied to the interstellar atoms moving in
the heliosphere and modified by including a loss function to take into account the
effects of charge exchange and photoionization. Details of the hot model formula-
tion will be given later in this paper (Lallement et al. 1985a; Izmodenov 2006).
12 2. Interstellar Hydrogen and Backscattered Lyman-α
The problem with finite temperature has been considered by Fahr (1971),
Thomas (1972), Feldman et al. (1972), Bertaux et al. (1972), Fahr (1974), Blum
et al. (1975), Meier (1977), Fahr (1978), and Wu and Judge (1979). Such a model
is called the classical hot model of hydrogen distribution. Very recently this clas-
sical hot model has been applied by Lee et al. (2012) to analyses of atom fluxes
measured at 1 AU by the Interstellar Boundary Explorer (IBEX). Measurements
of the backscattered solar Lyman-α radiation with a hydrogen absorption cell on
board the Soviet interplanetary probe Mars-7 (Bertaux et al. 1976) brought in-
formation on the spectral profile of the extraterrestrial Lyman-α radiation. More
accurate measurements with a hydrogen absorption cell flown on Prognoz-5, inter-
preted with the assumption of a Gaussian velocity distribution of H atoms, yielded
a temperature of (8.8 ± 1) · 103 K (Bertaux et al. 1977). However, a more accurate
interpretation of the spectral profile is only possible if the velocity distribution
function of the interstellar hydrogen is taken into account, even for the simple case
of μ = 1, in which atoms trajectories are straight lines.
In the 1980–1990s the classical hot model was widely used to interpret backscat-
tered solar Lyman-α measurements. However, very soon it became clear that effects
connected with the heliolatitudinal and solar cycle variations of the solar wind must
be taken into account.
It has been shown through modeling by Joselyn and Holzer (1975) that a non-
isotropic solar wind would strongly affect the distribution of atomic hydrogen in
the heliosphere, and that this could be observed in maps of backscattered solar
Lyman-α radiation. The signatures of the latitudinal variations were observed
in the backscattered interplanetary Lyman-α glow observations from Mariner 10
(Kumar and Broadfoot 1978, 1979; Witt et al. 1979, 1981), Prognoz 6 (Lallement
et al. 1985b; Summanen et al. 1993), Pioneer–Venus (Ajello et al. 1987; Lallement
and Stewart 1990), SOHO/SWAN (Bertaux et al. 1997, 1999), and by the Ulysses
GAS Lyman-α measurements (Pryor et al. 2003).
Lallement et al. (1985b) performed an analysis of Prognoz 5 and 6 measurements
by comparing the model results with the varying “a” parameter that is responsible
for the anisotropy of the solar wind. It was shown that the best agreement between
the data and the model is achieved for a 30–50 % decrease in the ionization rate
over the solar pole in comparison with the equatorial plane. The anisotropy of the
total ionization was studied in much more details by the SOHO/SWAN instrument
(e.g. Bertaux et al. 1997, 1999). Studies of the total ionization rate variations with
latitude and solar cycle were also performed by Pryor et al. (2003).
It follows from the classical hot model that due to net effects of solar gravitation,
radiation pressure, charge exchange and photoionization, the velocity distribution
of the interstellar atoms are disturbed at about 15–20 AU in the upwind direction.
In the downwind direction the “solar imprints” remain up to ∼ 100 AU. A typical
interstellar atom with a velocity of 20 km/s travels about 4 AU per year. The
time which is needed for the atom to pass through the region where the ionization
and gravitational effects are significant is therefore comparable with the 11 year
solar cycle. Since the solar radiation and solar wind flux both change during the
solar cycle, calculating the time-dependent variation of interstellar atoms within
the heliosphere requires that both be taken into account (Blum et al. 1993; Kyrölä
et al. 1994; Bzowski and Ruciński 1995; Ruciński and Bzowski 1995; Summanen
1996; Bzowski et al. 1997, 2002; Pryor et al. 2003; Quémerais et al. 2006).
The Solar Wind/Local Interstellar Medium Interaction 13
The variability of the solar factors exerts a significant influence on the hydrogen
density distribution within 10–20 AU from the Sun and is the most pronounced in
the downwind region. Ruciński and Bzowski (1995) have modeled the number
density of interstellar hydrogen in the frame of a time-dependent hot model. It
was shown that the departures of the density profiles from the stationary model
at different phases of the solar cycle are clearly visible up to 5 AU on the upwind
side (approximately the same occurs in the sidewind direction) and to 15 AU in
the downwind region. Further away from the Sun, the differences decrease and
practically vanish beyond 15–20 AU in the upwind and 50–60 AU in the downwind
directions, respectively.
Bzowski and Ruciński (1995) calculated the distributions of the H atom number
density by using a number of stationary models with “instantaneous” values of the
radiation pressure and ionization rate for the considered phase of the solar cycle.
Comparison of these distributions with the results of non-stationary model has
shown significant differences. Therefore, a time-dependent approach is essential to
accurately model the interstellar H atoms within the heliosphere.
It is clear from the historical review above that the development of the models
of the interstellar atoms inside the heliosphere is closely connected with the mea-
surements of the backscattered solar Lyman-α radiation, and the requirements of
the data analysis stimulated the theoretical model development.
It is important to underline that all of the models described above assume
that interstellar atoms have a pristine interstellar Maxwellian distribution func-
tion before they start to interact with the heliosphere. However, by the 1970s
(Wallis 1975) it was realized that before penetrating inside the heliosphere, inter-
stellar atoms of hydrogen pass through the so-called heliospheric interface that is
the region of the solar wind interaction with the charged component of the local
interstellar medium. At the present time, there is no doubt that the neutral and
plasma components interact in the heliospheric interface by charge exchange, and
these components should be treated self-consistently in the models by taking into
account their mutual interactions.
e
au s
li op
ck
He
ho
ws
H Bo Termina
t io
n
sh
o ck
SW
LISM
Figure 2.2: Qualitative picture of the SW interaction with the LIC. The heliopause
(HP) is a contact (or tangential) discontinuity which separates the solar wind
plasma and the interstellar plasma component. The termination shock (TS) is
formed due to the deceleration of the supersonic solar wind. The bow shock (BS)
may also exist if the interstellar plasma flow is supersonic. Four regions are distin-
guished: the supersonic solar wind (region 1); the solar wind flow between the TS
and the HP (region 2 or the inner heliosheath); the disturbed interstellar plasma
component flow (region 3 or the outer heliosheath); the undisturbed interstellar gas
flow (region 4)
Fluff) are in turn embedded in the Local Bubble that is believed to be made of hot
(∼ 106 K) and tenuous (∼ 10−3 cm−3 ) gas with a characteristic size on the order
of 100 pc (Welsh 2009).
The LIC is partially ionized (Lallement 2001). Interaction of the charged com-
ponent of the LIC with the solar wind plasma gives rise to the formation of the
interaction region, which is often called the heliospheric interface or heliosheath
(Fig. 2.2). The heliospheric interface has a complex structure, where the solar
wind and interstellar plasma, interplanetary and interstellar magnetic fields, inter-
stellar atoms of hydrogen, galactic and anomalous cosmic rays (GCRs and ACRs)
and pickup ions play important roles.
To get some insight into the structure and the physical processes inside the
interface using remote observations, a theoretical model must be employed.
The Solar Wind/Local Interstellar Medium Interaction 15
Theoretical studies of the heliospheric interface have been performed for more than
four decades, following the pioneering work by Parker (1961) and Baranov et al.
(1971). However, a complete theoretical model of the heliospheric interface has not
yet been constructed. The basic difficulty stems from the multi-component nature
of both the LIC and the solar wind. The LIC consists of at least four main com-
ponents: plasma (electrons and protons), hydrogen atoms, interstellar magnetic
field and galactic cosmic rays. The heliospheric plasma consists of particles of solar
origin (protons, electrons, alpha particles, etc.), pickup ions and energetic parti-
cle components that include, for example, the termination shock particles (TSP)
(Stone et al. 2005; Decker et al. 2005; Burlaga et al. 2005) and the anomalous cos-
mic ray (ACR) component. Pickup protons (or ions) are created by processes of
charge exchange, photoionization and electron impact ionization, and after being
ionized immediately picked up by the magnetic field. The pickup protons modify
the heliospheric plasma flow starting from ∼ 20–30 AU from the Sun. TSPs and
ACRs may also modify the plasma flow upstream of the termination shock and in
the heliosheath.
The first self-consistent stationary model of the interaction of the two-component
(plasma and H atoms) LIC with the solar wind was developed by Baranov and
Malama (1993). The interstellar wind was assumed to be an homogeneous paral-
lel supersonic flow, and the solar wind was assumed to be spherically symmetric
at Earth’s orbit. Under these assumptions, the heliospheric interface has an axi-
symmetric structure. The main physical process of this interaction is resonance
charge exchange (H atoms with protons), although the processes of photoionization
and ionization of H-atoms by electron impact can be important in some regions
of the heliosphere (for example, in the inner heliosheath or in the supersonic solar
wind). The significant effect of the resonance charge exchange is connected with
the large cross section of such collisions that is a function of the relative velocity
of colliding particles. However, it was discussed by some authors (Williams et al.
1997) that elastic H–H and H–proton collisions can be important in the problem of
the solar wind interaction with the local interstellar medium. This specific question
was discussed in detail by Izmodenov et al. (2000) and it was shown that the elastic
collisions are negligible.
The first self-consistent model of the SW/circumhelipspheric interstellar medium
(CHISM) interaction was developed by Baranov and Malama (1993). This is an
axisymmetric and stationary two-component model. The plasma component is
quasi-neutral and consists of electrons and protons. It is assumed that pickup
protons are assimilated into the plasma component immediately after ionization.
The plasma component is described as a fluid, and the Euler equations are solved
to get the spatial distribution of the plasma number density, np (r), bulk velocity,
Vp (r), and pressure Pp (r). The neutral component consists of hydrogen atoms and
is described kinetically. The two components interact by charge exchange. Pho-
toionization and electron impact ionization are taken into account in the model
as well.
The main results of the model can be described as follows. The interstellar
atoms strongly influence the heliospheric interface structure. The heliospheric
16 2. Interstellar Hydrogen and Backscattered Lyman-α
a b
Figure 2.3: Effect of the interstellar H atoms on the geometrical pattern of the in-
terface. (a) The heliospheric interface pattern in the case of a fully ionized circum-
heliospheric interstellar medium (CHISM), (b) the case of partly ionized CHISM.
Here BS, HP, and TS are the bow shock, the heliopause and the termination shock,
respectively. MD and TD are the Mach disk and the tangential discontinuity; RS
is the reflected shock that is formed in the case of a fully ionized plasma. These
results were obtained initially by Baranov and Malama (1993). Region 1 is the
supersonic solar wind, 2 is the inner heliosheath between the TS and HP, 3 is the
outer heliosheath between the HP and BS [From Izmodenov and Alexashov (2003)]
interface is much closer to the Sun in the case when the H atoms are taken into
account in the model as compared to the pure gas dynamical case as shown in
Fig. 2.3.
The termination shock becomes more spherical and the flow in the region be-
tween HP and TS becomes subsonic (the sonic lines disappear). The Mach disk
and the complicated tail shock structure, consisting of the reflected shock (RS) and
the tangential discontinuity (TD), disappear as well.
The supersonic flows upstream of the bow and termination shocks are modified
due to charge exchange with the neutral component. The supersonic solar wind
flow (region 1 in Fig. 2.3) is modified by charge exchange with the interstellar
neutrals. The new protons created by charge exchange are picked up by the solar
wind magnetic field. The Baranov–Malama model assumes immediate assimilation
of the pickup ions into the solar wind plasma. The solar wind protons and pickup
protons are treated as one-fluid, called the solar wind. The number density, velocity,
temperature, and Mach number of the solar wind are shown in Fig. 2.4a.
The effect of charge exchange on the solar wind is significant. By the time the
solar wind flow reaches the termination shock, it is decelerated by 15–30 %, heated
by a factor of 5–8, and loaded with the pickup proton component (approximately
20–50 %).
The Solar Wind/Local Interstellar Medium Interaction 17
a b c
1.6
TS
1.5 10−1 HP
ρ HP
ρR2
ρER2E 1.4 ρ
BS LIC
1.4
1.3
10−2
TS
1.2 BS
1.2
1.1 TS
np, cm−3
1.0 1.0 10−3
0 20 40 60 80 100 250 300 350 400 450 500 100 120 140 160 180 200
R (AU) R (AU) R (AU)
1.2 1.0 TS
6 1.6
BS
105 HP
4 1.4 HP
2 1.2 Tp, K
0 1.0 10 4
0 20 40 60 80 100 250 300 350 400 450 500 100 120 140 160 180 200
R (AU) R (AU) R (AU)
2.0 100
25 TS
Mach number
Mach number
20 1.5
10−1
15
1.0 BS
BS
10
10−2
0.5
5 HP
Mach number HP
TS −3
0 0 10
0 20 40 60 80 100 250 300 350 400 450 500 100 120 140 160 180 200
R (AU) R (AU) R (AU)
Figure 2.4: Plasma density, velocity, temperature, and Mach number upstream
of the termination shock (a), upstream of the bow shock (b), and in the helio-
sheath (c). The distributions are shown for the upwind direction. The solid curves
correspond to nH,CHISM = 0.2 cm−3 , np,CHISM = 0.04 cm−3 . The dashed curves
correspond to nH,CHISM = 0.14 cm−3 , np,CHISM = 0.10 cm−3 . VCHISM = 25.6 km/s
and TCHISM = 7,000 K in both cases [From Izmodenov (2000)]
18 2. Interstellar Hydrogen and Backscattered Lyman-α
The interstellar plasma flow is modified upstream of the bow shock by charge
exchange between the interstellar protons and secondary H atoms. These secondary
atoms originate in the solar wind. This leads to heating (40–70 %) and deceleration
(15–30 %) of the interstellar plasma before it reaches the bow shock (BS). The
Mach number decreases upstream of the BS and for a certain range of interstellar
parameters (nH,CHISM np,CHISM ) the bow shock may disappear. The solid curves
in Fig. 2.4b correspond to a small degree of ionization in the CHISM (np /(np +
nH ) = 1/6); the bow shock almost disappears in this case.
The interstellar neutrals also modify the plasma structure in the inner helio-
sheath. In a pure gasdynamic case (without neutrals) the density and temperature
of the postshock plasma are nearly constant. However, the charge exchange pro-
cess leads to a large increase in the plasma number density and decrease in its
temperature (Fig. 2.4c). The electron impact ionization process may influence the
heliosheath plasma flow by increasing the gradient of the plasma density from the
termination shock to the heliopause (HP, Baranov and Malama 1996). The influ-
ence of interstellar atoms on the heliosheath plasma flow is important, in particular,
for the interpretation of kHz-radio emissions detected by Voyager and for analysis
of the heliospheric ENA fluxes.
Charge exchange significantly alters the interstellar atom flow. The atoms newly
created by charge exchange have the velocity of their ion counterparts in the charge
exchange collisions. Therefore, the velocity distribution of these new atoms depends
on the local plasma properties in the location of their origin. It is convenient to
distinguish four different populations of atoms, depending on the region in the
heliospheric interface where the atoms were formed. Population 1 are the atoms
created in the supersonic solar wind up to the TS (region 1 in Fig. 2.3), population 2
are the atoms created in the inner heliosheath (region 2 in Fig. 2.3), and population
3 are the atoms created in the outer heliosheath (region 3 in Fig. 2.3). The atoms of
population 3 are often called the secondary interstellar atom component. We will
refer to the original (or primary) interstellar atoms as population 4. The number
densities and mean velocities of these populations are shown in Fig. 2.5I as functions
of heliocentric distance. The distribution function of H atoms, fH (r, w H ), can be
represented as a sum of the distribution functions of these populations: fH =
fH,1 + fH,2 + fH,3 + fH,4 . The Monte Carlo method allows us to calculate these four
distribution functions. These distributions were presented by Izmodenov (2001)
and Izmodenov et al. (2001) at 12 selected points in the heliospheric interface.
As an example, the distribution functions at the termination shock in the upwind
direction are shown in Fig. 2.5II for the four introduced populations of H atoms.
It is seen from this figure that the distribution functions of all H-atom populations
are not Maxwellian inside the heliosphere, i.e. the fluid approach is not correct
for describing the motion of neutral atoms. Comparisons of kinetic and different
multi-fluid approaches show significant differences in the results (Izmodenov and
Alexashov 2005; Mueller et al. 2008).
Original (or primary) interstellar atoms (population 4) are significantly fil-
tered (i.e. their number density is reduced) before reaching the termination shock
(Fig. 2.5I-a). The outer heliosheath is the main “filter” for these atoms. Since the
slow atoms have a small mean free path (due to both larger charge exchange cross
sections and smaller velocities) in comparison with the fast atoms, they suffer larger
The Solar Wind/Local Interstellar Medium Interaction 19
I II
a c
a c
Population 2
b d
b d
Figure 2.5: I. Number densities and velocities of the four atomic populations as
functions of heliocentric distance in the upwind direction. 1 designates atoms cre-
ated in the supersonic solar wind, 2 atoms created in the heliosheath, 3 atoms
created in the disturbed interstellar plasma, and 4 original (or primary) interstel-
lar atoms. Number densities are normalized to nH,CHISM , velocities are normalized
to VCHISM . It is assumed that nH,CHISM = 0.2 cm−3 , np,CHISM = 0.04 cm−3 . II.
Velocity distributions of the four atom populations—primary interstellar atoms
(population 4), secondary interstellar atoms (population 3), atoms created in the
inner heliosheath (population 2), atoms created in the supersonic solar wind (pop-
ulation 1)—at the termination shock in the upwind direction; wz is the projection
of the velocity vector on the axis parallel to the LIC velocity vector. Negative
values of wz indicate approach to the Sun. wx is the magnitude of the projection
of the velocity vector on the plane perpendicular to the interstellar velocity vector;
wR ,wθ are radial and tangential velocity components. All velocities are in km/s
[From Izmodenov et al. (2001)]
losses. This kinetic effect, called selection, results in a deviation of the interstellar
distribution function from a Maxwellian (Fig. 2.5II-a). The selection also results
in a ∼ 10 % increase in the primary atom mean velocity towards the termination
shock (Fig. 2.5I-c).
Secondary interstellar atoms (population 3) are created in the disturbed inter-
stellar medium by charge exchange between the primary interstellar neutrals and
protons decelerated in the vicinity of the heliopause. The secondary interstellar
atoms collectively make up the hydrogen wall, a density enhancement at the he-
liopause. The hydrogen wall was predicted by Baranov et al. (1991) and detected
in the direction of α Cen (Linsky and Wood 1996) by the Hubble Space Telescope.
At the termination shock, the number density of secondary neutrals is comparable
to the number density of the primary interstellar atoms (Fig. 2.5I-a, dashed curve).
20 2. Interstellar Hydrogen and Backscattered Lyman-α
The relative abundances of secondary and primary atoms entering the heliosphere
vary with the degree of interstellar ionization. The bulk velocity of population 3 is
about −18 ÷ −19 km/s. The sign “−” means that the population approaches the
Sun. One can see that the distribution function of this population is not Maxwellian
(Fig. 2.5II-b). The reason for the abrupt behavior of the distribution function for
wz > 0 is that the particles with significant positive wz velocities can reach the
termination shock only from the downwind direction. The distribution functions
of different H atom populations were calculated by Izmodenov et al. (2001) for
various upwind directions. The fine structures of the distribution functions of the
primary and secondary interstellar populations vary with direction. The direc-
tional variation of the velocity distribution reflects the geometrical pattern of the
heliospheric interface. The distribution functions of the interstellar atoms can be
a good diagnostic of the global structure of the heliospheric interface.
Another population (population 2) of the heliospheric hydrogen atoms are the
atoms created in the inner heliosheath by charge exchange with hot and compressed
solar wind and pickup protons. The number density of this population is by an or-
der of magnitude smaller than the number densities of the primary and secondary
interstellar atoms. This population has a minor importance for the interpretation
of Lyman-α and pickup ion measurements inside the heliosphere. Some atoms of
this population may probably be detectable by a Lyman-α hydrogen cell experi-
ment due to their large Doppler shifts (Quémerais and Izmodenov 2002). Recently
it was pointed out by Chalov and Fahr (2003) that charge exchange of these atoms
with solar wind protons may produce tails in the distribution function of pickup
ions that are measured at one or several AU from the Sun during quiet time periods.
Gruntman and Izmodenov (2004) showed that this population of H-atoms is a ma-
jor contributor to the density of interplanetary hydrogen at heliocentric distances
< 1 AU and could dominate in the downwind (interstellar wind) region under typ-
ical solar and interstellar conditions. Mass transport by heliospheric ENAs may
become especially important for determining the origin of the pickup ions attributed
to the inner source of neutral particles in the Sun’s vicinity.
Due to their high energies and large mean free path, a portion of the atoms
from this population penetrate upstream of the BS and modify the pristine inter-
stellar medium at large heliocentric distances. These atoms propagate freely in the
supersonic solar wind. Thus, these atoms are a rich source of information on the
plasma properties at the place of their birth, i.e. at the inner heliosheath. This
population of atoms is measured by the Interstellar Boundary Explorer (IBEX).
The last population of the heliospheric atoms are the atoms created in the
supersonic solar wind (population 1). The number density of this atom popu-
lation takes its maximum at ∼ 5 AU. At this distance, the number density of
the population is about two orders of magnitude smaller than the number density
of interstellar atoms. Outside the termination shock the density decreases faster
than r −2 , where r is the heliocentric distance (curve 1, Fig. 2.5I-b). The mean
velocity of population 1 corresponds to the bulk velocity of the supersonic solar
wind and is about 450 km/s. The distribution function of this population is also
not Maxwellian (Fig. 2.5II-d). The extended “tail” in the distribution function is
caused by the solar wind plasma deceleration upstream of the termination shock.
This “supersonic” atom population penetrates the interface and charge exchanges
The Solar Wind/Local Interstellar Medium Interaction 21
with interstellar protons beyond the BS. The process of charge exchange leads to
heating and deceleration of the interstellar gas upstream of the bow shock and,
therefore, to a decrease of the Mach number ahead of the bow shock.
Since 2003 the standard Baranov–Malama model has been modified by adding
solar wind alpha particles and interstellar helium ions to plasma components
(Izmodenov et al. 2003). Up to now, helium ions and alpha particles are not
included in the alternative models of other groups.
To evaluate possible effects of both interstellar ions of helium and solar wind
alpha particles, Izmodenov et al. (2003) performed parametric model calculations.
It was shown that the heliopause and the termination and bow shocks, are closer to
the Sun when the influence of interstellar helium ions is taken into account. This
effect is partially compensated by additional solar wind alpha particle pressure
that was also taken into account in the model. The net result is as follows: the
heliopause, termination and bow shocks are closer to the Sun by ∼ 12 AU, ∼ 2 AU,
∼ 30 AU, respectively in the model taking into account both interstellar helium
ions and solar wind alpha particles as compared to the model ignoring these ionized
helium components. Despite the fact that the net effect of interstellar helium ions
and solar alpha particles is rather small (∼ 7 % displacement for the heliopause,
∼ 10 % the BS, and ∼ 2 % for the TS), it still can be important for the interpretation
and prediction of experimental data related to the inner heliosheath region.
Aleksashov et al. (2000) explored the problem of solar wind interaction with the
CHISM for the case where the interstellar magnetic field is parallel to the relative
Sun/CHISM velocity vector. In this case, the model remains axisymmetric. It
was shown that the effect of the interstellar magnetic field on the positions of
the TS, BS and HP is significantly smaller when H atoms are included (Baranov
and Zaitsev 1995). The calculations were performed with various Alfvén Mach
numbers in the undisturbed CHISM. It was found that the √ bow shock straightens
out with decreasing Alfvén Mach number MA = VCHISM 4πρ/BCHISM (i.e. with
an increasing magnetic field strength in the CHISM). It gets closer to the Sun near
the symmetry axis, but recedes from it on the flanks. By contrast, the nose of the
heliopause recedes from the Sun due to the tension of magnetic field lines, while the
wings of the heliopause get closer to the Sun due to magnetic pressure. As a result,
the region of the compressed interstellar medium at the heliopause (or “pileup
region”) is reduced by almost 30 %, as the magnetic field increases from zero to
3.5 × 10−6 Gauss. It was also shown that the H atom filtration and heliospheric
distributions of primary and secondary interstellar atoms are virtually unchanged
over the entire considered range of the interstellar magnetic field (0–3.5 × 10−6
Gauss). The magnetic field has the strongest effect on the density distribution of
population 2 of H atoms, which increases by a factor of almost 1.5 as the interstellar
magnetic field increases from zero to 3.5 × 10−6 Gauss.
Izmodenov et al. (2005), as well as Izmodenov and Alexashov (2005, 2006),
studied the problem assuming that the interstellar magnetic field (IsMF) is inclined
with respect to the direction of the interstellar flow. In this case, the SW/CHISM
interaction region becomes asymmetric and the flow pattern becomes essentially
three-dimensional. Here we present new results obtained within the framework of
this model for the interstellar magnetic field BCHISM = 4.4 μG and α = 15◦ , where
α is the angle between the CHISM flow velocity and the IsMF direction.
22 2. Interstellar Hydrogen and Backscattered Lyman-α
-3.2 -2.9 -2.6 -2.3 -2.0 -1.7 -1.4 -1.2 -1.1 0.1 0.4 9.3E+03 2.0E+05 5.5E+05 9.0E+05 1.3E+06 1.6E+06
600 600 2.0E-07 1.2E-06 2.2E-06 3.2E-06 4.2E-06 5.1E-06
600
400 400
400
200 200
200
X (AU)
X (AU)
X (AU)
0 0
0
400
400 400
200
200 200
X (AU)
X (AU)
0
0 X (AU) 0
Figure 2.6: Isocontours of the proton number density (plot a), plasma tempera-
ture (plot b), magnetic field (plot c), total thermal plasma and magnetic pressure
(plot d), normalized hydrogen (plot e) and oxygen (plot f ) number densities in
the plane determined by the interstellar velocity and magnetic field vectors. The
streamlines of the plasma component (plot a), the magnetic field lines (plot c)
and “streamlines” of the hydrogen (plot e) and oxygen (plot f ) atoms are shown.
The termination shock and heliopause are shown as white curves [From Izmodenov
et al. (2009)]
Figure 2.6 shows the results of the model calculations in the xz plane. The xz
plane is determined by the Sun-CHISM relative velocity vector and the interstellar
magnetic field vector. The direction of the z-axis is chosen to be opposite to the
interstellar gas velocity vector. The plasma streamlines and isolines of the plasma
density are shown in panel a, the plasma temperature in panel b, and the total
(plasma thermal plus magnetic field) pressure in panel d. Panel c shows the mag-
netic field lines and isocontours of the interstellar magnetic field. The heliopause
and the termination shock become asymmetric with respect to the direction of the
interstellar flow due to the asymmetric pressure of the IsMF (panel c).
The interstellar magnetic field pressure pushes the heliopause and the termi-
nation shock towards the Sun as compared to the model without magnetic field.
The Solar Wind/Local Interstellar Medium Interaction 23
Quantitatively the effect of the IsMF on the location of the TS in the directions of
Voyager 1 and Voyager 2 is shown in Table 2.1. The TS distances in this table are
obtained in the framework of the stationary 3D kinetic-MHD model that is under
discussion, but corrected according to a kinetic-gasdynamic model of the interface
by Izmodenov et al. (2008) that takes into account solar cycle effects (Izmodenov
2009). It is seen from the table that the models with BCHISM = 4.375 μG and
α = 15◦ –20◦ are in good agreement with the actual distances of 94 AU and 84 AU
of the Voyager 1 and 2 crossings of the TS. This is the reason why this model is
presented here. It is interesting to note that using a multi-fluid approach for in-
terstellar H atoms allows us to estimate the distances for comparable observations
in other directions of the interstellar magnetic field and not just in the directions
measured by the Voyagers (Opher et al. 2009). The detailed discussion and com-
parison with models of other groups is out of the scope of the paper (Izmodenov
and Alexashov 2005).
The bow shock disappears for such a strong magnetic field because the Mach
numbers calculated with respect to the Alfvén and fast magnetosonic waves are
smaller than one.
Figure 2.6a also shows the streamlines of the plasma component. The stag-
nation point is located in the upper half of the zx plane and is displaced from
the z axis. It is important to note that the velocity vector of the plasma passing
through the region of maximum plasma density has a noticeable Vx component.
The secondary interstellar atoms which originate in the vicinity of the HP should
have the properties of the plasma in this region. Figure 2.6e presents the num-
ber density of this interstellar atom component. The maximum density appears
in the region between the HP and the BS. This is the so-called hydrogen wall, as
discussed above. It is seen in the figure that the maximum of the hydrogen wall
is also slightly shifted to the upper half of the xz plane and reflects the behavior
of the plasma distributions. The streamlines of the H atom component are also
shown in Fig. 2.6e. The streamlines were plotted based on the mean velocity field
distribution of the interstellar H atoms. The velocity vector V H determines the
direction of the averaged H atom flow. It can be seen from the figure that in
the heliosphere the velocity vector V H has a noticeable Vx component even very
close to the Sun. The effect of the shift of the direction of interstellar hydrogen
flow relative to the direction of the interstellar helium flow was observed by the
SOHO/SWAN H cell instrument (Lallement et al. 2005).
24 2. Interstellar Hydrogen and Backscattered Lyman-α
The measurements of the deflection of the interstellar H atom flow relative to the
direction of the interstellar helium flow is one of several observational constraints
on the magnitude and direction of the interstellar magnetic fields. Estimates per-
formed by Izmodenov et al. (2005) and Izmodenov and Alexashov (2006) showed
that the models with BCHISM = 2.5 μG, and α = 30 − 45◦ provide levels of de-
viation in agreement with the level obtained from analysis of the SOHO/SWAN
data. However, recent numerical results (Alexashov et al. 2008) showed that the
deflection angle is very sensitive to the charge exchange cross section. Results of
calculations with updated cross section from Lindsay and Stebbing (2005) show a
larger deflection angle as compared with the previous calculations which employed
the cross section from Maher and Tinsley (1977).
Other possible constraints on the direction and magnitude of the interstellar
magnetic field are:
1. Asymmetries of the TS toward Voyager 1 and Voyager 2 direction, which
were discussed before
2. Flows of the solar wind beyond the TS that are directly measured by Voyager
2 (Opher et al. 2009) and estimated from the anisotropy of the energetic
particle fluxes from Voyager 1 (Krimigis et al. 2011)
3. Analyses of the heliospheric ENA fluxes measured by IBEX (Heerikhuisen
and Pogorelov 2011).
More detailed parametric model calculations are required and the work is currently
in progress.
The Baranov–Malama model and more recent models described above assume
immediate assimilation of pickup protons into the solar wind plasma and consider
the mixture of solar wind and pickup protons as a single component.
However, measurements of the pickup proton distribution function on board
the Ulysses and ACE spacecraft show that the pickup proton distribution function
is non-Maxwellian. The observations show also that the velocity distribution func-
tion can be considered as isotropic (fast pitch-angle scattering) except some short
periods in the inner heliosphere when the interplanetary magnetic field is almost
radial. These data point to the absence of thermodynamic equilibrium between
the pickup and solar-origin protons, although their mean velocities are equal. In
the improved model of the solar wind/interstellar medium interaction developed by
Malama et al. (2006), the pickup protons are considered as a separate component
with thermodynamic parameters different from those of the solar wind. Since in the
solar-wind reference frame, the pickup proton distribution function is assumed to
be isotropic, the following angle-averaged distribution function can be introduced:
∗ 1
fpui (r, w) = fpui (r, v) sin θdθdφ . (2.2)
4π
Here, v is the individual velocity of a pickup proton, w = v − V is the pickup
proton velocity in the solar-wind reference frame (V is the solar wind velocity in
the heliocentric coordinate system), w, θ, and φ are the spherical coordinates of
vector w, and fpui (r, v) is the pickup proton distribution function. The kinetic
The Solar Wind/Local Interstellar Medium Interaction 25
∗
equation for fpui (r, w) can be written in the form (Isenberg 1987):
∗ ∗ ∗ ∗
∂fpui ∂fpui 1 ∂ ∂fpui w ∂fpui
+V · = 2 2
w D + divV + S(r, w) . (2.3)
∂t ∂r w ∂w ∂w 3 ∂w
Here, D(r, w) is the diffusion coefficient in the velocity space, while the source
function of pickup protons S(r, w) reflects their birth and loss due to charge ex-
change, photoionization, and the electron impact ionization of H atoms. Malama
et al. (2006) assumed for simplicity that D(r, w) = 0 and the kinetic equation
(2.3) was solved together with the Euler equations written for the sum of all the
charged components and the kinetic equation for the interstellar hydrogen atoms.
The use of Euler equations for the sum of the charged components is justified by
the fact that all the components (solar protons and electrons, as well as pickup pro-
tons) move at the same velocity and that for all the components the distribution
functions are isotropic. In this case, the pressure p is equal to the sum of partial
pressures, while the effective pressure of the pickup protons is determined in terms
∗
of the distribution function fpui , so that:
4π ∗
p = pe + pp + ppui , ppui = mp w4 fpui (r, w)dw
3
To treat the passage of pickup protons through the termination shock we assume
the conservation of their magnetic moment. The analytical derivation of the mag-
netic moment conservation has been done by Alekseev and Kropotkin (1971) for
perpendicular and nearly perpendicular shocks. Later it has been shown numeri-
cally by Terasawa (1979) that the conservation of the magnetic moment exists for
all quasi-perpendicular shocks.
The reflections of pickup protons at the shock potential is not taken into ac-
count. A self-consistent solution of the problem just formulated was obtained by
Malama et al. (2006) for the case D = 0 corresponding to a quiescent solar wind
in which the magnetic field fluctuation level is low.
The thermodynamic non-equilibrium of the pickup and solar protons leads to
a thinning of the inner heliosheath region (Fig. 2.7a), which is attributable to the
decrease in the total pressure of the charged component. As compared with the
calculations performed in accordance with the corresponding equilibrium model,
the TS is displaced by 5 AU away from the Sun in the direction of the oncoming
flow, whereas the HP is by 12 AU closer. In the tail region, the heliocentric distance
of the TS increases by 70 AU.
Recently we expanded the Malama et al. (2006) model by including the pickup
proton component outside the heliopause. Indeed, energetic hydrogen atoms which
originated in the supersonic solar wind and the inner heliosheath can penetrate
through the heliopause and enter into the interstellar plasma. Outside the he-
liopause the atoms can suffer charge exchange with interstellar protons. As a
result new energetic interstellar pickup protons are created. Having an energy of
1 keV or more, these pickup protons have large mean free path as compared with
the thermal interstellar protons and can be considered as collisionless with a non-
Maxwellian velocity distribution. The pickup protons in the outer heliosheath can
charge exchange with interstellar H atoms. Eventually, new ENAs of the outer
26 2. Interstellar Hydrogen and Backscattered Lyman-α
heliosheath origin will be created in this secondary charge exchange. Similar to the
ENAs from the inner heliosheath, the ENAs originating in the outer heliosheath
can penetrate to small heliocentric distances and be measured there.
Figure 2.7b shows the calculated fluxes of H atoms at 1 AU obtained within the
framework of the new non-equilibrium model. The figure presents total fluxes of all
heliospheric ENAs (solid curve), fluxes of ENAs originated in the inner heliosheath
from solar wind protons (black dots), fluxes of the ENAs originated in the inner
heliosheath from the pickup protons (blue triangles), and fluxes of the ENAs orig-
inated in the outer heliosheath from the pickup protons (blue dots). The ENAs
originated in the inner heliosheath from the solar wind protons are the strongest
at energies of ∼ 0.02–0.2 keV. For this energy range the fluxes of the ENAs from
the outer heliosheath (blue dots in Figure) are smaller but still comparable. In
the energy range of ∼ 0.4–1.0 keV the ENAs originating in the inner heliosheath
from the pickup protons (blue triangles in Figure) dominate, while for energies
above 2 keV fluxes of the ENAs from the outer heliosheath (blue dots) become
higher than the other. Therefore, the model suggests that the ENAs with energies
above 1 keV may be of interstellar (outer heliosheath) origin. The ENAs of such
energies have been measured on board of several spacecraft. Upper limits of the
heliospheric ENA fluxes estimated from measurements on board of SOHO/CELIAS
(Hilchenbach et al. 1998; Hilchenbach et al. 2000), Cassini/INCA (Krimigis et al.
2009) and Venus Express (Brandt et al. 2009) are shown in Fig. 2.7b as green dots.
The figure shows that generally there is a good agreement between the model results
and the data.
Here we should also note that the calculations were performed in the frame-
work of a model that has no stochastic mechanisms for pickup proton accelera-
tion (Malama et al. 2006). Izmodenov et al. (2009) presented an extension of the
Malama et al. (2006) model by introducing a non-thermal population of pickup
protons in the interstellar medium. It has been shown that the interstellar pickup
protons form significant fluxes of ENAs due to charge-exchange, and those fluxes
exceed the fluxes of other ENAs at energies above ∼ 1 keV. These models predict
that the ENA fluxes have maxima near the upwind direction of the heliosphere and
minima at the flanks, though, of course, the position of the maxima can slightly
deviate from the upwind direction due to the effects of the interstellar magnetic
field and SW asymmetry (Izmodenov et al. 2009).
The first full sky maps of heliospheric ENAs measured by the Interstellar Bound-
ary Explorer (IBEX) spacecraft (McComas et al. 2009; Fuselier et al. 2009; Funsten
et al. 2009; Schwadron et al. 2009) show results that were entirely unexpected in
the frame of the previously discussed models. The first scan of the whole sky
showed that maxima of ENA fluxes form a long (∼ 250◦ –∼ 300◦ ) and narrow arc-
like feature called ribbon that was not predicted by any model prior to the IBEX
observations. Chalov et al. (2010) proposed a modification of the Malama et al.
(2006), Izmodenov et al. (2009) model that ignore the scattering of newly created
energetic pickup protons, but including the effects of ion transport for the pickup
protons generated in the region outside of the heliopause by charge exchange be-
tween the thermal interstellar protons and heliospheric ENAs. The results of the
The Solar Wind/Local Interstellar Medium Interaction 27
Figure 2.7: (a) Strong discontinuity surfaces calculated taking into account the
pickup proton non-equilibrium (curves 1), and in accordance with the equilibrium
plasma model (curves 2). (b) Fluxes of H atoms at 1 AU calculated with the pickup
proton non-equilibrium model: total fluxes of all heliospheric ENAs (solid curve),
ENAs originated in the inner heliosheath from solar wind protons (black dots),
ENAs originated in the inner heliosheath from pickup protons (blue triangles),
ENAs originated in the outer heliosheath from pickup protons (blue dots). Upper
limits of the heliospheric ENA fluxes measured by SOHO/CELIAS, Cassini/INCA
and Venus Express are shown by the green dots [From Izmodenov et al. (2009)]
model yield a feature qualitatively similar to the IBEX ribbon. In future stud-
ies, the results of simulations will be quantitatively compared to the IBEX ENA
observations.
Note, however, that the more complex kinetic description of the supra-thermal
population of pickup protons has no direct influence on the primary and secondary
populations of the interstellar atoms which are created from the thermal protons
of the interstellar plasma. Nevertheless, an indirect influence is possible because
taking into account the pickup component properly would change the thermal
population too. Therefore, we do not expect that the effects of the non-equilibrium
plasma model would dramatically change the previous conclusion on the properties
of the primary and secondary interstellar atoms, but some effects are possible
mainly due to global dynamic effects (Malama et al. 2006) and need to be checked
in future.
complex 3D kinetic-MHD models of the SW/LIC interaction region. The most im-
portant aspect of the model for the purpose of this paper is that upon entering the
heliosphere, the distribution function of interstellar H atoms is strongly disturbed
in the SW/LIC interaction region. This means that the properties of the interstel-
lar atoms inside the heliosphere are dependent on the structure of the heliospheric
interface, and, therefore, the distribution function of interstellar hydrogen can not
be used directly to determine the local interstellar parameters as it would be if
the population of these atoms were not disturbed in the heliospheric interface. At
the same time, this opens up a possibility to use the backscattered solar Lyman-α
radiation as a diagnostic of the properties of the heliospheric interface.
Figure 2.8: Coordinate system used in the axisymmetric models: z is the axis of
symmetry directed toward the upwind direction, ρ is an axis in cylindrical coor-
dinate system that is perpendicular to the upwind direction; curve 1 is the outer
boundary in the advanced model of hydrogen distribution inside the heliosphere;
curve 2 is a schematic representation of the heliospheric termination shock [From
Katushkina and Izmodenov (2011)]
(1 − μ)GMs r
F(r) = mH (1 − μ)Fg = −mH · ,
r2 r
G is the gravitational constant, Ms the mass of the Sun, mH the mass of a H
atom, and r the radius-vector to a point. μ is the ratio of gravitational attraction
to repulsion from radiation pressure as defined in Eq. 2.1. In general, the parameter
μ depends on time (t), heliolatitude (λ) and the radial component of velocity (wr ).
The kinetic equation for the velocity distribution function of H atoms can be
written as:
∂f (r, w, t) ∂f (r, w, t)
+w· +
∂t ∂r
F(r, t, λ, wr ) ∂f (r, w, t)
· = −β(r, t, λ) · f (r, w, t) (2.4)
mH ∂w
Here f (r, w, t) is velocity distribution function of the hydrogen atoms, w is the
individual velocity of a H atom.
The right hand size of Eq. (2.4) represents losses of atoms due to charge ex-
change (H + H + ↔ H + + H) and photoionization (H + hν = H + + e). Electron
impact ionization is not taken into account. The coefficient β(r, t, λ) is the effective
ionization rate: β(r, t, λ) = βex (r, t, λ) + βph (r, t, λ), where βex and βph are rates of
charge-exchange and photoionization, respectively. It is assumed that these rates
of ionization decrease with distance from the Sun as ∼ 1/r 2 (where r is the helio-
centric distance), since these values are proportional to the number density of the
solar protons and flux of the solar photons, respectively. Therefore:
r 2 r 2
E E
β(r, t, λ) = (βex,E (t, λ) + βph,E (t, λ)) = βE (t, λ) ,
r r
30 2. Interstellar Hydrogen and Backscattered Lyman-α
where rE = 1 AU is the distance from the Earth to the Sun. The subscript E
indicates that the values are measured at 1 AU.
The functions μ(t, λ, wr ), βex,E (t, λ), and βph,E (t, λ) adopted in this model are
based on experimental data. Detailed descriptions of these functions will be given
below.
The kinetic equation (2.4) is a linear partial differential equation, and it can
be solved by the method of characteristics. It is known that the characteristics of
this equation coincide with the trajectories of the atoms. The distribution function
f (r, w, t) changes along a characteristic according to the equation:
The solution of the kinetic equation can then be obtained by integration of the last
equation subject to boundary conditions:
t
f (r, w, t) = fb (r0 , w0 , t0 ) exp − β(r, t, λ)dt , (2.5)
t0
and
Tρ (r) ∼ f (r, w) · (Vρ (r) − wρ )2 dw.
5. The third moment of f (r, w) for the secondary interstellar atoms is not zero:
Kzzz ∼ f (r, w) · (Vz (r) − wz )3 dw = 0.
This means that this distribution function is not symmetric relative to its
maximum.
All these features are illustrated in Figs. 2.9 and 2.10. Figure 2.9 presents
moments of f (r, w) at 90 AU as functions of angle θ. These results were obtained in
the frame of the Baranov–Malama model with the following boundary conditions in
the undisturbed LISM: the number densities of protons and neutral hydrogen atoms
are np,LISM = 0.06cm−3 , nH,LISM = 0.18cm−3 , respectively; the relative LISM/SW
velocity is VLISM = 26.4km/s, and the interstellar temperature is TLISM = 6519 K.
For the inner boundary conditions at the Earth’s orbit the following parameters
of the solar wind were adopted: np,E = 6cm−3 , VE = 441.9km/s, and the Mach
number ME = 4.034. Note that in the 3D case, for example when the interstellar
magnetic field is taken into account, the hydrogen distribution at 90 AU becomes
asymmetric (due to three-dimensional structure of the heliospheric interface) and
the non-Maxwellian properties are even more complex.
Note, also, that the previous attempt to treat the distribution function as non-
Maxwellian has been done by Sherer et al. (1999). However, only the first and
third properties listed above have been taken into account in their model, while
Katushkina and Izmodenov (2010, 2011) studied all these non-Maxwellian features
of the velocity distribution at 90 AU (that hereafter we call the global effects of
the heliospheric interface). They concluded that these global effects are important
and should be included in any theoretical model of H atoms distribution inside the
heliosphere. Therefore, special boundary conditions for the velocity distribution
function at 90 AU are necessary.
32 2. Interstellar Hydrogen and Backscattered Lyman-α
a b c
d e f
Figure 2.9: Number densities (a), bulk velocities (b, c), correlation coefficient Kzρ
(d), third moment Kzzz , and kinetic “temperatures” Tz , Tρ and Tav (e, f ) of the
primary (blue) and secondary (red) interstellar atoms at 90 AU as functions of angle
θ from the upwind direction. Kzρ (d) and Kzzz are presented in dimensionless units;
kinetic temperatures Tz , Tρ and Tav are in Kelvins. Tav is an averaged temperature
defined as 3 · Tav = Tz + Tρ + Tϕ .
Boundary Conditions
In this paper we consider three models of heliospheric H atoms. One of the three
models (Model 3) includes the global effects mentioned above, while the two other
models are more simplified. We will compare the results of the three models to
identify and separate the effects of local and global phenomena on the distribution
of H atoms in the heliosphere. The models differ only by the boundary conditions
at 90 AU. A summary of the models discussed in this section is given in Table 2.2.
Model 1 is the one-component hot model that implies solution of the kinetic
equation (2.4) with the Maxwellian velocity distribution function at the outer
boundary. For the specific calculations performed in the frame of this model we
use parameters of the Maxwellian distribution, i.e., the number density (nH ), bulk
velocity (VH,z ), and temperature (TH ) calculated at 90 AU in the upwind direction
in the frame of the Baranov–Malama model. These values at 90 AU are:
where nH,LISM = 0.18 cm−3 , VH,LISM = 26.4 km/s. These parameters correspond to
the mixture of the primary and secondary interstellar atoms. Note that the effects
of the heliospheric interface are treated very simply, assuming the parameters of
the Maxwellian distribution are different from the parameters in the LISM.
Model 2 is the so-called two-component hot model (Bzowski et al. 2008). In
this model, the kinetic equation (2.4) is solved separately for the primary and
secondary populations of interstellar atoms. The velocity distribution functions
are assumed to be Maxwellian for each of the populations and the parameters
of the distribution functions are calculated in the frame of the Baranov–Malama
model. These parameters for the primary interstellar atoms are:
1 Dρ
+ (Vϕ − wϕ )2 + (Vz − wz )2
Dϕ Dρ Dz − Vzρ 2
Vzρ
−2 · (V ρ − wρ )(V z − wz ) (2.6)
Dρ Dz − Vzρ
2
Figure 2.11: Parameters of the secondary interstellar atoms inside the heliosphere
as functions of heliocentric distance along the upwind direction (a) and crosswind
direction (b). Solid curves correspond to the results of the Baranov–Malama model;
dashed curves correspond to the results of model 2; dashed-dot curves correspond
to the advanced model 3. In these calculations μ = 1.258, βph,E = 1.16 × 10−7 s−1 ,
βex,E = 4.8 × 10−7 s−1 . To calculate the charge-exchange rate at the Earth orbit,
we applied averaged values of the solar wind number density (6 cm−3 ) and velocity
(440 km/s) known from measurements and the charge-exchange cross-section from
Lindsay and Stebbing (2005) [From Katushkina and Izmodenov (2011)]
from local effects (i.e. within the heliosphere), which can be explored in the second
step in details. These local effects may include time-dependent and 3D phenomena
as described in the next section.
At the end of this section we illustrate the importance of the non-Maxwellian
features at the outer boundary by comparing the results of model 2 and model 3
shown in Fig. 2.11. The results of the Baranov–Malama model are shown to verify
precision of model 3.
Figure 2.11 presents the number density and effective kinetic “temperatures”
of the gas along the upwind and crosswind directions. It is seen that there is a
very good agreement between the results of the Baranov–Malama model and the
advanced model of hydrogen distribution within 90 AU, while the two-component
hot model leads to significant discrepancies. Firstly, the two-component hot model
overestimates the number density of hydrogen everywhere in the heliosphere. This
is because the two-component hot model assumes constant parameters of hydrogen
at 90 AU, but due to the heliospheric interface effects the hydrogen parameters
depend on angle θ (e.g. the number density at 90 AU decreases with θ). Second,
36 2. Interstellar Hydrogen and Backscattered Lyman-α
were used to calculate fluxes of mass, momentum and energy. The fluxes with
hourly resolution were used to calculate the averaged fluxes (over one day), which
were adjusted to 1 AU from the Earth orbit. Then these data were averaged over
one Carrington rotation. The charge-exchange ionization rate can be calculated as
follows:
βex,E (t, 0) = np,E (t) · wsw,E (t) · σ(wsw,E ),
The heliolatitudinal dependence for the total ionization rate βtot,E (t, λ) has
been calculated using SOHO/SWAN data. A detailed description of how to get
the total ionization rate βtot,SW AN (t, λ) from SOHO/SWAN is given in Quémerais
et al. (2006). Overall, βtot,E (t, λ) is determined as follows:
βtot,SW AN (t, λ)
βtot,E (t, λ) = βtot,E (t, 0) · .
βtot,SW AN (t, 0)
FB (Fsolar (t), wr ))
μeq (t, wr ) = μeq (t, wr = 0) · ,
FB (Fsolar (t), wr = 0))
μ(t, λ, wr ) = μpole (t, wr ) + cos2 (λ) · (μeq (t, wr ) − μpole (t, wr )),
where
μpole (t, wr ) = FB (Fsolar − ΔFsolar , wr )
and ΔFsolar ≈ 0.05 · 1011 ph/(cm2 s), this value is given from Pryor et al. (1998).
The resulting functions of βtot,E , βex,E , βph,E , and μ are shown in Figs. 2.12, 2.13
and 2.14.
38 2. Interstellar Hydrogen and Backscattered Lyman-α
Figure 2.12: (a) Time dependence of the total ionization rate (βtot,E ), charge-
exchange rate (βex,E ) and photoionization rate (βph,E ) at the Earth’s orbit in the
solar equatorial plane (λ = 0). (b) Time dependence of the radiation pressure
parameter μ for λ = 0 and wr = 0
Figures 2.15 and 2.16 show the number densities, radial components of the bulk
velocities Vr , and the radial kinetic temperatures Tr for the primary and secondary
interstellar atoms in the upwind direction as functions of time. The distributions
are shown for different heliocentric distances. As shown in the figures, the time-
variations of the number densities of the primary and secondary populations behave
qualitatively the same and both are anti-correlated with the time variations of the
ionization rate. For example, in 2002 at solar maximum there is a maximum
of ionization rate as shown in Fig. 2.12a and a minimum of the H atom number
densities. Note that charge exchange provides the main contribution to the total
ionization rate since the charge exchange rate exceeds the photoionization rate
significantly. The local maximum of the number densities in 2004–2005 is due to
the local minimum of the charge exchange rate. There is a monotonic decrease of
the solar activity and therefore ionization rate from 2005 to 2008 that results in a
continuous increase of the number density from 2005 to 2008. There is an abrupt
fall number density in 2009 caused by the increase in the charge exchange rate.
Advanced Model of the Interstellar Hydrogen in the Heliosphere 39
a b
a b
c d
Figure 2.15: Distribution of the number density (a, b), radial velocity (c) and radial
kinetic temperature (d) of the primary interstellar atoms along the upwind
direction as functions of time; different colors correspond to different heliocentric
distances: (1) r = 1 AU, (2) r = 2 AU (3) r = 3 AU, (4) r = 4.4 AU, (5) r = 7.4 AU,
(6) r = 10 AU
It can be seen from the figures that time fluctuations of all parameters are small
at 10 AU and beyond. Therefore, local effects become small for r > 10 AU in the
upwind direction, in agreement with Ruciński and Bzowski (1995).
Let us focus on the bulk velocities of the H atoms in the upwind direction.
Figures 2.15c and 2.16c show that the time-variation of the bulk velocities is very
similar to the time variation of μ(t), shown in Fig. 2.12b. Therefore, fluctuations
of the bulk velocities of the primary and secondary atoms are mainly determined
by the force changing the atom trajectories rather than the ionization processes.
However, a qualitative difference in the behavior of the primary and secondary
H atoms appears when the heliocentric distance increases. In Fig. 2.15c for the
primary population all curves cross each other at two time periods in 1998 and
Advanced Model of the Interstellar Hydrogen in the Heliosphere 41
a b
c d
Figure 2.16: Distribution of the number density (a, b), radial velocity (c) and radial
kinetic temperature (d) of the secondary interstellar atoms along the upwind
direction as functions of time; different colors correspond to different heliocentric
distances: (1) r = 1 AU, (2) r = 2 AU (3) r = 3 AU, (4) r = 4.4 AU, (5) r = 7.4 AU,
(6) r = 10 AU
atoms moves towards larger velocities, and the bulk velocity increases. This effect
was discussed in Lallement et al. (1985a) and Bzowski et al. (1997).
When μ > 1, the effective force decelerates the atoms as they approach the Sun.
For μ ≈ 1.1, the effect of the deceleration due to the effective force compensates
the selection effect. In this case the bulk velocity practically does not change when
the atoms approach the Sun. In the case of μ > 1.1, as it was from 1998 to 2006,
the repulsive force decelerates the atoms and the bulk velocity of the gas becomes
smaller. The smallest bulk velocity was in 2002 when the repulsive force had its
maximum. Also, for μ > 1.1 and some fixed moment of time, the bulk velocity
decreases towards the Sun. In the case of μ < 1 the bulk velocity increases towards
the Sun.
For the secondary interstellar atoms the situation is slightly different. This
population is slower than the primary population, and approaching the Sun the
atoms are more effectively ionized. As a result, the bulk velocity of the secondary
interstellar atoms increases towards the Sun during the entire period considered
with the exception of the 2002 solar maximum when the bulk velocity did not
change with heliocentric distance.
Another qualitative difference between the primary and secondary populations
is in the behavior of the bulk velocity in the period from 2006 to 2009. As it was
mentioned before, the behavior of the bulk velocity for the primary populations
reflects time-variations of μ(t), while for the secondary population the bulk velocity
decreases from 2006 to 2008 and then rapidly increases in 2009. Such a behavior of
the bulk velocity can be explained by the strong decrease in the charge-exchange
rate before 2008. A strong increase of the bulk velocity in 2009 is due to the
increase in the charge exchange rate in 2009 and the corresponding increase of the
selection effect.
Now, let us focus on the kinetic temperature Tr , that is the second moment of
the velocity distribution function calculated for radial velocity component.
Figures 2.15d and 2.16d show that the temperatures of the primary and secondary
populations are qualitatively similar in behavior. For a small heliocentric distance
the time variations of Tr resembles μ(t). This means the radial temperature (as
well as the bulk velocity) is mainly affected by the resulting force. However, the
ionization processes are important as well. For example, the temperature decreases
towards the Sun for both the primary and secondary populations except near solar
maximum. Figure 2.17 presents the number density of both populations of inter-
stellar hydrogen as functions of time and heliocentric distance along the upwind
direction.
Figure 2.18 shows sky-maps of the number densities, radial velocities and kinetic
temperature Tr of the primary and secondary interstellar atoms at 1 AU calculated
for 1996, i.e. for the minimum of solar activity. The upwind direction has the
following heliographic coordinates: heliolongitude αHGI = 178.98◦ ± 0.5◦ and he-
liolatitude βHGI = 5.11◦ ± 0.2◦ . It is seen (Fig. 2.18a, d) that the maxima of the
H atom number densities occur at the upwind heliolongitude. For this upwind he-
liolongitude there are two maxima in the southern and northern hemispheres and
a minimum in the plane of the solar equator. This minimum in number density is
due to the maximum of the ionization rate near the solar equatorial plane. The
distributions of radial velocity, Fig. 2.18b, e, show minima near the upwind direc-
Advanced Model of the Interstellar Hydrogen in the Heliosphere 43
a b
Figure 2.17: Number density of the primary and secondary interstellar atoms in
the upwind direction as functions of time and heliocentric distance
Figure 2.18: Number density (a, d), radial velocity (b, e) and radial kinetic temper-
ature (c, f ) of the secondary (population 3, left column) and primary (population
4, right column) interstellar atoms, calculated at 1 AU as functions of heliolatitude
and heliolongitude in the heliographic coordinate system (HGI). These results are
obtained for solar minimum conditions (year 1996), when the heliolatitudinal de-
pendence of the total ionization rate is very strong. In HGI coordinates, the upwind
direction is at 178.98◦ in heliolongitude and at 5.11◦ of heliolatitude
44 2. Interstellar Hydrogen and Backscattered Lyman-α
tion. Similar minima in the upwind direction are seen in the maps of the radial
kinetic temperature Tr (Fig. 2.18c, f). However, the temperature maps also show
heliolatitudinal effects. The latitudinal effect is more pronounced for the secondary
population.
Here nH (r) is the number density of hydrogen atoms, σν (r, ν) is the differential
scattering cross-section that is proportional to the projection of the hydrogen dis-
tribution function on the line-of-sight, j (r,ν,-Ω) is the atomic emission coefficient
which measures the number of photons emitted by a hydrogen atom per second per
frequency unit and per solid angle. Note that scattered photons travel in the direc-
tion opposite to the line-of-sight direction, i.e. in −Ω. The first term on the right
hand side of Eq. (2.7) is the loss term due to absorption of the scattered photons.
The second term is the local source of emission due to the scattering process.
Equation (2.7) has the following solution in the self-absorption approximation:
∞
4π
I(r, ν, Ω) = 6 nH (r + sΩ) j(r + sΩ, ν, −Ω) e−τν (r+sΩ,r) ds, (2.8)
10 0
Backscattered Lyman-α 45
where s is the coordinate along the line-of-sight, τν (r , r) is the optical thickness
for scattered photons with the frequency ν calculated from the scattered point
r = r + sΩ for an observer located at point r. The atomic emission coefficient j
can be represented by:
j(r , ν, −Ω) = φ(ω) FS (r , νp ) σν (r , ν) . (2.9)
Here φ(ω) is the scattering phase function that gives the relation between the
directions of propagation of the photon before and after the scattering (Brandt
and Chamberlain 1959). FS (r , νp ) is the flux of the solar Lyman-α photons with
the frequency of νp at point r . The solar Lyman-α spectra obtained by Lemaire
et al. (1998) is used to calculate the solar Lyman-α flux at the Earth orbit at a
given frequency ν.
In the case when the line-of-sight is radial, there is a simple relation between the
frequency of the primary solar photon νp and the frequency of the backscattered
photon ν: νp = 2 · ν0 − ν, ν0 is the frequency at line center. Thus, if we know the
hydrogen velocity distribution function in the entire heliosphere, we can calculate
the profiles of the backscattered Lyman-α radiation.
Since we consider two different populations of the interstellar H atoms inside the
heliosphere and neglect multiple scattering effects, it becomes possible to calculate
profiles of the radiation scattered by each populations separately. To do this we
consider photons that were scattered by the primary and secondary interstellar
atoms independently. Optical thickness is calculated for the mixture of the primary
and secondary atoms, because a photon scattered, for example, by the primary
interstellar atom can then be absorbed by atoms of both populations.
We can calculate the following moments of the backscattered radiation profiles:
(a) Intensity measured in [R],
∞
Ilos (r, Ω) = I(r, ν, Ω) dν;
0
a b
Figure 2.20: Contributions to the total number density (a) and radial velocity
(b) of atoms arriving from various directions. Curve 1 in plot (a) corresponds
to point 1 (r1 = 2 AU, θ = 0◦ ), and curve 2 corresponds to point 2 (r2 = 7 AU,
θ = 180◦ ). The contribution to the radial velocity was calculated only for point 2.
Various curves in plot (b) correspond to various models of hydrogen distribution:
the solid line is model 3, the dashed line is model 2 and the dashed-dot line is
model 1. All calculations are performed for a mixture of primary and secondary
interstellar atoms [From Katushkina and Izmodenov (2011)]
upwind line-of-sight and at 7 AU for downwind. From the simple (naive) consider-
ation one could expect that most of the H atoms in the regions close to the Sun
would arrive from upwind. However, Lallement and Bertaux (1990) have shown
(in the case of the hot model) that most of the atoms arrive into the vicinity of the
Sun not exactly from upwind, since the thermal velocities of hydrogen atoms at
the boundary are large enough compared to the bulk velocities. The same is true
for our model. To illustrate this we calculate the function n(ri , θb ) for two points
(i = 1, 2) which correspond to the maximum emissivity region. Point 1 is located
in the upwind direction at 2 AU from the Sun; point 2 is located in the downwind
direction at 7 AU from the Sun. The function n(ri , θb ) represents the contribution
to the total number density at a given point (point 1 for curve 1 in Fig. 2.20a and
point 2 for curve 2) from these particles which cross the outer boundary of 90 AU
at θ = θb . This function n(ri , θb ) is defined as follows:
n(r1 , θb ) = f (r1 , w1 )dw1 . (2.10)
Ω1
Figure 2.21: Schematic picture of penetration of the hydrogen atoms from the outer
boundary to the vicinity of the Sun. Point 1 (r1 ) is located in the upwind direction
at 2 AU from the Sun. In the case of straight atom’s trajectories we consider only
those atoms that arrive to point 1 from the outer boundary at θb . The angle θ1
can be determined if we know r1 and θb [From Katushkina and Izmodenov (2011)]
equal to θ1 , as can be seen from Fig. 2.21. Let us introduce a spherical coordinate
system in velocity space. It means that we describe the velocity vector w by its
modulus w̃ and two spherical angles θ̃ and ϕ̃, i.e. orthogonal coordinates of vector
w can be represented as follows:
wx = w̃ · sin(θ̃) · cos(ϕ̃)
wy = w̃ · sin(θ̃) · sin(ϕ̃)
wz = −w̃ · cos(θ̃)
In spherical coordinates, dw1 = w̃12 sin(θ̃) dw̃1 dθ̃ dϕ̃. For velocities from the
subspace Ω1 , θ̃ = θ1 = const for the chosen value of θb and integration over θ̃ is
not needed. Hence, for our case Eq. (2.10) can be rewritten in the following form:
+∞ 2π
n(r1 , θb ) = f (r1 , w̃1 , θ̃ = θ1 , ϕ̃) w̃12 sin(θ1 ) dw̃1 dϕ̃. (2.11)
0 0
to the axial symmetry of the boundary conditions. Therefore Eq. (2.11) can be
represented as:
+∞ 2π
n(r1 , θb ) = fb (θb , w̃1 , θ1 (θb )) exp(−A(r1 , θb , w̃1 )) w̃12 sin(θ̃1 ) dw̃1 dϕ̃
0 0
+∞
= 2π sin(θ1 ) · fb (θb , w̃1 , θ1 (θb )) exp(−A(r1 , θb , w̃1 )) w̃12 dw̃1
0
= 2π sin(θ1 ) · g(r1 , θb ).
The function fb (θb , w̃1 ) decreases with θb for each given value of w̃1 , because at
90 AU Vz,H Vρ,H and the distribution function has a maximum in the upwind
direction. The loss-function, A(r1 , θb , w̃1 ), increases with θb for point 1 because
the length of the trajectory has a minimal value upwind and exp(−A) decreases
with θb . Hence, the function g(r1 , θb ) decreases with θb , and sin(θ1 ) increases with
θb for θb ∈ [0, π/2]. Therefore, n(r1 , θb ), that is a product of sin(θ1 ) and g(r1 , θb ),
should have a maximum at a certain θb between 0◦ and 90◦ .
Figure 2.20a shows n(ri , θb ) calculated numerically for point 1 in the upwind
direction (curve 1) and for point 2 in the downwind direction (curve 2). The figure
shows that the largest fraction of the interstellar atoms arrive at point 1 from
θ ≈ 15◦ and at point 2 from θ ≈ 55◦ . Therefore, contrary to expectations, most of
the H atoms reach the vicinity of the Sun not exactly from the upwind direction.
Now, it is clear that the excess in the backscattered Lyman-α intensity in models
1 and 2 as compared with model 3 is related to the higher number densities of H
atoms inside the heliosphere in models 1 and 2 arising from the lack of θ-dependent
boundary conditions. In model 3, the number density at 90 AU decreases with θ.
It also becomes clear why the intensities (and line-shifts) of the radiation scat-
tered by the primary interstellar atoms nearly coincide for all models as shown in
Fig. 2.19b. This is because on the one hand, the angle-dependence of the hydro-
gen parameters at 90 AU for the primary interstellar atoms is weaker than for the
secondary interstellar atoms. On the other hand, the distribution function of the
primary interstellar atoms is closer to a Maxwellian than the distribution function
of the secondary interstellar atoms.
From the middle column of Fig. 2.19a, c it can be seen that there are noticeable
differences in the line-shifts of models 2 and 3 in the downwind direction. The
differences are seen for the total radiation as well as for the radiation scattered by
the secondary interstellar atoms. At the same time, there are almost no discrepan-
cies in the line-shifts of model 1 and model 3, despite the fact that model 1 is the
simplest model without any of the effects of the heliospheric interface. Therefore
one would expect some differences.
In order to understand these results we calculated the contribution to the total
radial velocity of hydrogen at point 2 (7 AU in downwind) from the particles that
reach this point from different directions. Similar to n(ri , θb ), the contribution to
the radial velocity can be calculated as follows:
1
Vr (ri , θb ) = +∞ · f (ri , wi ) wi,r dwi
−∞
f (ri , wi )dwi Ωi
Backscattered Lyman-α 51
and
Tρ (r) ∼ f (r, w) · (Vρ (r) − wρ )2 dw
Recall that z-axis is the central axis of the axisymmetric heliosphere and ρ is
the distance from that axis in a cylindrical coordinate system. In other words,
the mean thermal velocities of the H atoms are different in different directions.
Moreover, this difference between Tz and Tρ temperatures increases approaching
to the Sun due to local effects. The large maxima of the line-widths at θ = 90◦
shown in Fig. 2.19b, c are related to the changes in the radial kinetic temperature
Tr of the gas at 90 AU as a function of θ. For example, when an observer looks
toward the upwind direction (i.e. θ = 0◦ ) Tr = Tz , for a line-of-sight with θ = 90◦
Tr = Tρ and for a line-of-sight of θ = 180◦ Tr = Tz again. In model 3 at the
outer boundary, Tρ is higher than Tz for each of the interstellar populations of H
atoms based on the Baranov–Malama model results. This results in the maxima of
Tr at θ = 90◦ and it is reflected in the maxima of the line-widths for the photons
scattered by each population of interstellar H atoms separately.
The differences in the Lyman-α intensities, line-shifts, and line-widths calcu-
lated in models 1–3 are summarized in Table 2.3. It shows the differences (both
absolute and relative) between the model results for the upwind (θ = 0◦ ), crosswind
(θ = 90◦ ), and downwind (θ = 180◦ ) directions. It shows that the one-component
hot model 1 gives differences from 2 % (in the upwind direction) to 24 % (in the
downwind direction) as compared with model 3 in intensities and about 10–18 %
difference in the line-widths. The two-component hot model 2 leads to 8–27 %
discrepancies with model 3 in the intensities, and from 10 % (upwind) to 40 %
(downwind) discrepancies in the line-widths. The differences in the line-shifts of
models 1 and 2 relative to model 3 are also large, especially for the crosswind di-
rection. However, they are related to very small values of the line-shifts and most
probably can not be detected experimentally.
Backscattered Lyman-α 53
1. The ionization processes that change the parameters of H atoms near the Sun.
From our test calculations we have found that solar gravitation and radiation
pressure have a smaller influence on the line-width than ionization.
2. The existence of two populations of the interstellar hydrogen atoms that are
shifted in velocity space. This effect leads to the appearance of the minimum
of Tlos at θ = 90◦ , as discussed before.
4. The spectral shape of the solar Lyman-α line. It will be shown below that
the shape of the solar spectrum has some small but interesting effect on the
line-width as a function of θ.
Figure 2.22 summarizes the results of the test calculations and explores the
effects listed above. Table 2.2 spells out the model numbering scheme used in
Fig. 2.22. For all of the calculations in this section, we assume that the solar
gravitational force is in balance with solar radiation pressure, i.e. μ = 1 and the
trajectories of the H atoms are straight lines. To explore the effects of ionization
54 2. Interstellar Hydrogen and Backscattered Lyman-α
a b
in the vicinity of the Sun, we performed test calculations with a typical ionization
rate of βE = 5.9 × 10−7 s−1 as well as calculations with βE = 0 (models 2b, 3b, 3c).
The effects of the local ionization on the function of Tlos (θ) are clearly illustrated
in the frame of the one-component hot model 1 (curve 1 in Fig. 2.22a) since neither
the two-component nor the non-Maxwellian effects are taken into account in this
model. The line-width of the backscattered Lyman-α profile increases monotoni-
cally from the upwind direction to downwind. This behavior reflects the angular
variation in the effective radial temperature of the H atoms that increases from
upwind to downwind (see Fig. 3.5 in Izmodenov 2006). Such a behavior is due to a
combination of the so-called selection effect that leads to a reduction in the atom
temperature in the upwind direction and in the effect of broadening of the velocity
distribution function toward the downwind direction. The latter effect is due to
the fact that atoms penetrate into the downwind mainly from the sides that makes
the distribution function broader (see Fig. 3.4 in Izmodenov 2006). The selection
effect is the effect when the slower atoms are more ionized as compared with the
fast atoms, because they spend more time before approaching the vicinity of the
Sun (and Earth) and have more time to be ionized. For a more detailed descrip-
tion of these effects and evolution of the H atom velocity distribution inside the
heliosphere, see Izmodenov (2006).
Backscattered Lyman-α 55
Results of the two-component hot model, 2b, that do not account for the ioniza-
tion (curve 2b in Fig. 2.22) illustrate the effect of the two populations of the inter-
stellar H atoms. These two populations have different bulk velocities Vz (at 90 AU
in the upwind direction: Vz,primary ≈ −27 km/s and Vz,secondary ≈ −16 km/s) and
rather small thermal velocities. Therefore, in the upwind and downwind directions
(where Vr = ±Vz ), the line-of-sight projections of the velocity distribution func-
tions of the primary and secondary components overlap only partially in velocity
space. In the crosswind direction Vr = Vρ ≈ 0 for both primary and secondary
interstellar atoms. This means that in this direction the projections of the distri-
bution functions on the radial line-of-sight overlap completely. That is why the
radial temperature of the mixture of primary and secondary interstellar atoms is
smaller in the crosswind direction than in the upwind and downwind directions.
This minimum of the radial temperature of H atoms in the crosswind direction is
reflected in the Lyman-α line-widths as seen from model 2b.
Results of the two-component hot model, curve 2 of Fig. 2.22, where the ion-
ization is taken into account combine both the increase in Tlos from upwind to
downwind due to the ionization effect and the local minimum in the crosswind
direction due to the two populations of H atoms. Hence, a small minimum of Tlos
at θ = 50–60◦ is seen in curve 2.
The line-width obtained from model 3 is presented in Fig. 2.22b. Note that this
model takes into account all known effects of the heliospheric interface, namely: (1)
two populations of interstellar H atoms, (2) θ-dependence of the hydrogen parame-
ters at 90 AU, and (3) non-Maxwellian features of the hydrogen velocity distribution
function at the outer boundary of the model. In particular, the differences between
kinetic temperature’s components, Tz and Tρ , play an important role.
Curves 3b and 3c correspond to models that do not take ionization into account.
Additionally, model 3c assumes that the solar Lyman-α flux does not depend on
frequency. This case is referred to as the “flat” solar spectrum. Note that in all the
other models we use the shape of the solar spectrum from Lemaire et al. (1998).
It can be seen that the effect of minimum in Tlos at θ = 90◦ (due to the two
populations) almost disappears in model 3c, but is still visible. This disappearance
of the minimum is related to the growth of the radial kinetic temperature Tr with
a increase of θ at 90 AU. The effect of the Tr increase with θ compensates the effect
of the minimum in the radial hydrogen temperature in crosswind due to the two
populations of H atoms. That is why the value of the minimum of the line-widths
at θ = 90◦ is much smaller for model 3c as compared with model 2b.
It is interesting to note that for model 3b (which is more realistic than model
3c) the local minimum of Tlos at θ = 90◦ is replaced with a small maximum. This
effect is due to the shape of the solar spectrum. Remember that in model 3c we
use a flat solar spectrum while in model 3b we use the nonuniform solar spectra
from Lemaire et al. (1998).
Let us return to model 3 without any additional assumptions. It is seen (curve 3)
that in this case there is no minimum in Tlos at 50–60◦ at all, but instead a small
maximum at θ = 150◦ . This maximum can be explained by the θ-dependence of
the hydrogen kinetic temperature Tr . The radial temperature of the sum of the
primary and secondary interstellar atoms is shown in Fig. 2.23 (plot a corresponds
to model 2, plot b corresponds to model 3). It is seen that in the case of the two-
56 2. Interstellar Hydrogen and Backscattered Lyman-α
Figure 2.23: Kinetic radial temperature (Tr ) for the sum of the primary and
secondary interstellar atoms in the heliosphere; Plot a corresponds to the two-
component hot model 2, plot b corresponds to the results of the model 3. In these
calculations μ = 1 and βE = 5.9 × 10−7 s−1 [From Katushkina and Izmodenov
(2011)]
was shown that the minimum of the line-widths appears only for the two-component
hot model, and there is no minimum at all for model 3 that takes into account all
effects of the heliospheric interface. It was found that the absence of the minimum
in model 3 is due to the effect of the two components being compensated by the
non-Maxwellian features of the velocity distribution of H atoms at 90 AU after
they passed the heliospheric interface region, namely, by a strong anisotropy of the
kinetic temperatures of H atoms (Tz < Tρ ).
Therefore, the question of why the minimum of the Lyman-α line-width exists
in the experimental data remains unanswered. Possibilities for reproducing the
minimum still exist in the context of model 3. Firstly, the models considered here
do not take into account the effects of latitudinal and solar cycle variations of
the photoionization and charge exchange rates as well as solar radiation pressure.
These local effects may potentially change the result of this paper.
Another option to explain the minimum in line-width would be to change the
boundary conditions at 90 AU, i.e. to make a change in the model of the heliospheric
interface. For example, interstellar magnetic field may play a key role (Izmodenov
et al. 2005). Another possibility is a change in the multi-component nature of both
the heliospheric and interstellar plasmas (Malama et al. 2006; Izmodenov et al.
2009; Chalov et al. 2010). In this non-equilibrium plasma model the interstellar
pickup ions would be treated as a separate kinetic component. The plasma tem-
perature in the vicinity of the heliopause is smaller in that model as compared
with the Baranov–Malama model. Therefore, we could expect a decrease in the
kinetic temperatures of the secondary interstellar atoms. This might result in a
larger velocity space separation of the primary and secondary interstellar atoms at
90 AU. The separation may enhance the effect of the two populations. As a result
one may hope that the observed minimum will appear in the model.
Summary
The interaction between the supersonic flow of partially ionized plasma of the
local interstellar medium and the solar wind produces a complicated flow pattern
consisting of one or two shock waves (the heliospheric termination shock and, pos-
sibly, the bow shock in the LISM), and a contact discontinuity, the heliopause. Due
to the charge exchange process, the region between the two shocks (heliospheric
interface) separating these flows plays the role of a filter for the penetration of the
interstellar hydrogen atoms (and also for O, N, and other species) into the Solar
System.
From a theoretical point of view, the interaction should be considered in the
frame of kinetic-continuum models where the interstellar H atom component is
described in the framework of kinetic theory, since for hydrogen atoms the Knudsen
number with respect to charge exchange is Kn ∼ 1. The first self-consistent model
of the SW/LIC interaction was developed by Baranov and Malama (1993). Since
that time the set of kinetic-continuum models was developed. The modern kinetic-
continuum models take into account the following physical components/effects:
• Ionized interstellar helium component and solar wind alpha particles;
• Anomalous and galactic cosmic rays;
58 2. Interstellar Hydrogen and Backscattered Lyman-α
• Our state-of-the-art model (Katushkina and Izmodenov 2011) has been devel-
oped specifically to model the distribution of H atoms inside the heliosphere
properly for the purpose of understanding the spectral parameters of the
backscattered solar Lyman-α. This model allows us to combine the simplic-
ity of the hot-type models of the H atom distribution inside the heliosphere
with the results of the global models of the SW/LIC interaction.
• Solar cycle variations of the solar wind parameters influences the H atom
distribution and should be taken into account in the analysis of backscattered
Lyman-α.
Acknowledgements
The authors would like to thank ISSI for their support of the working group.
The calculations were performed by using the supercomputers of the Russian
Academy of Sciences and Lomonosov Moscow State University (“Lomonosov” and
“Chebishev”). The work was performed under the Presidium RAS Programm 22.
V.I. and O.K were partially supported by RFBR grants 10-02-93113-CNRS-a, 10-
02-01316-a, 11-02-92605-RS-a, and the Regional Public Fund for Russian science.
M.B. acknowledges support from the Polish ministry for science and higher educa-
tion, grants NS-1260-11-09 and N-N203-513-038.
Bibliography
J.M. Ajello, A.I. Stewart, G.E. Thomas, A. Graps, Solar cycle study of interplanetary
Lyman-alpha variations: Pioneer Venus Orbiter sky background results. Astrophys. J.
317, 964–986 (1987)
D.B. Aleksashov, V. Baranov, E. Barsky, A. Myasnikov, An axisymmetric magnetohydro-
dynamic model for the interaction of the solar wind with the local interstellar medium.
Astron. Lett. 26, 743–749 (2000)
I.I. Alekseev, A.P. Kropotkin, Passage of energetic particles through a magnetohydrody-
namic discontinuity surface. Geomagn. Aeron. 10, 755 (1971)
D.B. Alexashov, V.V. Izmodenov, M. Opher, Effects of the helisopheric and interstellar
magnetic field on the heliospheric interface. 37th COSPAR Scientific Assembly D13-0015-
08, P025-TueWed (poster) (2008)
60 2. Interstellar Hydrogen and Backscattered Lyman-α
W.I. Axford, The interaction of the solar wind with the interstellar medium. In Solar wind,
ed. by C.P. Sonett, P.J. Coleman, J.M. Wilcox. (Scientific and Technical Information
Office, National Aeronautics and Space Administration, Washington), p. 609 format?
(1972)
W.I. Axford, A.J. Dessler, B. Gottlieb, Termination of solar wind and solar magnetic field.
Astrophys. J. 137, 1268–1278 (1963)
V.B. Baranov, Gasdynamics of the solar wind interaction with the interstellar medium.
Space Sci. Rev. 52, 89–120 (1990)
V.B. Baranov, The heliosheath as a special case of stellarsheaths and the hydrogen wall as
a signature of the heliosheath. Planet. Space Sci. 50, 535–539 (2002)
V.B. Baranov, M.K. Ermakov, M.G. Lebedev, A three-component model of solar wind-
interstellar medium interaction: some numerical results. Sov. Astron. Lett. 7, 206–209
(1981)
V.B. Baranov, V.V. Izmodenov, Y.G. Malama, On the distribution function of H atoms in
the problem of the solar wind interaction with the local interstellar medium. J. Geophys.
Res. 103, 9575–9586 (1998)
V.B. Baranov, K.V. Krasnobaev, A.G. Kulikovksy, Sov. Phys. Dokl. 15, 791 (1971)
V.B. Baranov, M.G. Lebedev, Y.G. Malama, The influence of the interface between helio-
sphere and the local interstellar medium on the penetration of the H-atoms to the solar
system. Astrophys. J. 375, 347–351 (1991)
V.B. Baranov, Y.G. Malama, Model of the solar wind interaction with the local interstellar
medium: numerical solution of self-consistent problem. J. Geophys. Res. 98, 15157–15163
(1993)
V.B. Baranov, Y.G. Malama, Axisymmetric self-consistent model of the solar wind interac-
tion with the LISM: basic results and possible ways of development. Space Sci. Rev. 78,
305–316 (1996)
V.B. Baranov, N.A. Zaitsev, On the problem of the solar wind interaction with magnetized
interstellar plasma. Astron. Astrophys. 304, 631–637 (1995)
C.A. Barth, Mariner 6 measurements of the Lyman α sky background. Astrophys. J.Lett.
161, L181–L184 (1970)
J.-L. Bertaux, J.E. Blamont, Evidence for a source of an extraterrestrial hydrogen Lyman
α emission: the interstellar wind. Astron. Astrophys. 11, 200 (1971)
J.-L. Bertaux, R. Lallement, Analysis of interplanetary Lyman α line profile with a hydro-
gen absorption cell: theory of the Doppler angular spectral scanning method. Astron.
Astrophys. 140, 230–242 (1984)
J.-L. Bertaux, A. Ammar, J.E. Blamont, OGO-5 determination of the local interstellar wind
parameters. Space Res. 12, 1559–1567 (1972)
J.-L. Bertaux, J.E. Blamont, N. Tabarie, W.G. Kurt, M.C. Bourgin, A.S. Smirnov, N.N.
Dementeva, Interstellar medium in the vicinity of the sun: a temperature measurement
obtained with the Mars-7 interplanetary probe. Astron. Astrophys. 46, 19–29 (1976)
J.L. Bertaux, J.E. Blamont, E.N. Mironova, V.G. Kurt, M.C. Bourgin, Temperature mea-
surement of interplanetary interstellar hydrogen. Nature 270, 156–158 (1977)
J.-L. Bertaux, E. Quémerais, R. Lallement, E. Kyrölä, W. Schmidt, T. Summanen, J.P.
Goutail, M. Berthé, J. Costa, T. Holzer, First results from the SWAN Lyman α solar
wind mapper on SOHO. Sol. Phys. 175, 737–770 (1997)
J.-L. Bertaux, E. Kyrölä, E. Quémerais, R. Lallement, W. Schmidt, T. Summanen, J. Costa,
T. Makinen, SWAN observations of the solar wind latitude distribution and its evolution
since launch. Space Sci. Rev. 87, 129–132 (1999)
P.W. Blum, H.-J. Fahr, Interaction between interstellar hydrogen and the solar wind. As-
tron. Astrophys. 4, 280–290 (1970)
P.W. Blum, J. Pfleiderer, C. Wulf-Mathies, Neutral gases of interstellar origin in interplan-
etary space. Planet. Space Sci. 23, 93–105 (1975)
P. Blum, P. Gangopadhyay, H.S. Ogawa, D.L. Judge, Solar-driven neutral density waves.
Astron. Astrophys. 272, 549–554 (1993)
J.C. Brandt, J.W. Chamberlain, Interplanetary gas. I. hydrogen radiation in the night sky.
Astrophys. J. 130, 670–682 (1959)
Bibliography 61
S.A. Fuselier, F. Allegrini, H.O. Funsten et al., Width and variation of the ENA flux ribbon
observed by the interstellar boundary explorer. Science 326, 962 (2009)
K. Gringauz, V. Bezrukih, V. Ozerov, R. Ribchinsky, A study of the interplanetary ionized
gas, high-energy electrons and corpuscular radiation from the Sun by means of the three-
electrode trap for charged particles on the second soviet cosmic rocket. Sov. Phys. Doklady
5, 361 (1960)
M. Gruntman, V.V. Izmodenov, Mass transport in the heliosphere by energetic neutral
atoms. J. Geophys. Res. 109, A12108 (2004). doi:10.1029/2004JA010727
M. Gruntman, E.C. Roelof, D.G. Mitchell, H.-J. Fahr, H.O. Funsten, D.J. McComas, En-
ergetic neutral atom imaging of the heliospheric boundary region. J. Geophys. Res. 106,
15767–15782 (2001)
J. Heerikhuisen, N.V. Pogorelov, An estimate of the nearby interstellar magnetic field using
neutral atoms. Astrophys. J. 738, 29 (2011). doi:10.1088/0004-637X/738/1/29
M. Hilchenbach, K.C. Hsieh, D. Hovestadt et al., Detection of 55–80 keV hydrogen atoms
of heliospheric origin by CELIAS/HSTOF on SOHO. Astrophys. J. 503, 916–922 (1998)
M. Hilchenbach, K.C. Hsieh, D. Hovestadt, R. Kallenbach, A. Czechowski, E. Möbius, P.
Bochsler, Energetic neutral hydrogen of heliospheric origin observed with SOHO/CELIAS
at 1 AU. In The Outer Heliosphere: The Next Frontiers, ed. by K. Scherer, H. Fichtner,
H.J. Fahr, E. Marsch. (Pergamon, Elmsford, 2000), pp. 273–276
A.J. Hundhausen, Interplanetary neutral hydrogen and the radius of the heliosphere. Planet.
Space Sci. 16, 783–793 (1968)
P.A. Isenberg, Evolution of interstellar pickup ions in the solar wind. J. Geophys. Res. 92,
1067–1073 (1987)
V.V. Izmodenov, Physics and gasdynamics of the heliospheric interface. Astrophys. Space
Sci. 274, 55–69 (2000)
V.V. Izmodenov, Velocity distribution of interstellar H atoms in the heliospheric interface.
Space Sci. Rev. 97, 385–388 (2001)
V.V. Izmodenov, Early concepts of the heliospheric interface: H atoms. In The Physics
of the Heliospheric Boundaries, ed. by V.V. Izmodenov, R. Kallenbach. ISSI Scientific
Report No. 5 (ESA-ESTEC, Paris, 2006), pp. 45–65
V.V. Izmodenov, Local interstellar parameters as they are inferred from analysis of obser-
vations inside the heliosphere. Space Sci. Rev. 143, 139–150 (2009)
V.V. Izmodenov, D. Alexashov, A model for the tail region of the heliospheric interface.
Astronomy Lett. 29, 58–63 (2003)
V.V. Izmodenov, D. Alexashov, Kinetic vs multi-fluid models of H atoms in the heliospheric
interface. In Solar Wind 11 / SOHO 16 Connecting Sun and Heliosphere conference, ed.
by B. Fleck, T.H. Zurbuchen, H. Lacoste. (ESA SP-592, 2005), p. 355
V.V. Izmodenov, D. Alexashov, Multi-component 3d modeling of the heliospheric inter-
face: effects of interstellar magnetic field. In Physics of the inner heliosheath: Voyager
observations, theory, and future prospects. 5th annual IGPP international astrophysics
conference. AIP Conference Proceedings, vol. 858 (2006), pp. 14–19
V.V. Izmodenov, Y.G. Malama, A.P. Kalinin et al., Hot neutral H in the heliosphere: elastic
H-H, H-p collisions. Astrophys. Space Sci. 274, 71–76 (2000)
V.V. Izmodenov, M. Gruntman, Y.G. Malama, Interstellar hydrogen atom distribution
function in the outer heliosphere. J. Geophys. Res. 106, 10681 (2001)
V.V. Izmodenov, G. Gloeckler, Y.G. Malama, When will Voyager 1 and 2 cross the termi-
nation shock? Geophys. Res. Lett. 30, 3–1 (2003). doi:10.1029/2002GL016127
V.V. Izmodenov, D. Alexashov, A. Myasnikov, Direction of the interstellar H atom inflow in
the heliosphere: role of the interstellar magnetic field. Astron. Astrophys. 437, L35–L38
(2005)
V.V. Izmodenov, Y.G. Malama, M.S. Ruderman, Modeling of the outer heliosphere with
the realistic solar cycle. Adv. Space Res. 41, 318–324 (2008)
V.V. Izmodenov, Y.G. Malama, M.S. Ruderman et al., Kinetic-gasdynamic modeling of the
heliospheric interface. Space Sci. Rev. 146, 329–351 (2009)
J.A. Joselyn, T.E. Holzer, The effect of asymmetric solar wind on the Lyman α sky back-
ground. J. Geophys. Res. 80, 903–907 (1975)
Bibliography 63
O.A. Katushkina, V.V. Izmodenov, Effect of the heliospheric interface on the distribution
of interstellar hydrogen atom inside the heliosphere. Astronomy Lett. 36, 297–300 (2010)
O.A. Katushkina, V.V. Izmodenov, The influence of effects on the heliospheric interface on
parameters of backscattered solar Lα radiation measured at the Earth’s orbit. Cosmic
Res. 50, 141–151 (2012)
S.M. Krimigis, D.G. Mitchell, E.C. Roelof, P.C. Brandt, Energetic neutral atoms (ENA)
from the termination shock/heliosheath? the view from 10AU. Paper presented at Voy-
agers in the heliosheath: observations, models, and plasmas physics Kauai, Hawaii, 9–14
Jan 2009
S.M. Krimigis, E.C. Roelof, R.B. Decker, M.E. Hill, Zero outward flow velocity for plasma
in a heliosheath transition layer. Nature 474, 359–361 (2011)
S. Kumar, A.L. Broadfoot, Evidence from Mariner 10 of solar wind flux depletion at high
ecliptic latitudes. Astron. Astrophys. 69, L5–L8 (1978)
S. Kumar, A.L. Broadfoot, Signatures of solar wind latitudinal structure in interplanetary
Lyman α emissions: Mariner 10 observations. Astrophys. J. 228, 302–311 (1979)
J.E. Kupperian, E.T. Byram, T.A. Chubb, H. Friedman, Far ultraviolet radiation in the
night sky. Planet. Space Sci. 1, 3–6 (1959)
V.G. Kurt, Measurement of scattered Lyman α radiation in the vicinity of the Earth and
in interplanetary space. In Space Research: Transactions of the All-union Conference on
Space Physics, ed. by G.A. Skuridin et al. NASA Technical Translation: NASA TT F-
389, Science Publishing House, Moscow, 10–16 June 1965, p. 769 (Translation published
by NASA, Washington DC, USA, May 1966)
V.G. Kurt, Kosmicheskie Issledovania (in russian) 5(6), 769–775 (1967)
V.G. Kurt, T.A. Germogenova, Scattering of solar Lyman α radiation by galactic hydrogen.
Sov. Astron. 11, 278–282 (1967)
V.G. Kurt, R.A. Syunyaev, Observations and interpretation of the ultraviolet radiation of
the Galaxy. Sov. Astron. 11, 928–931 (1967)
E. Kyrölä, T. Summanen, P. Raback, Solar cycle and interplanetary hydrogen. Astron.
Astrophys. 288, 299–314 (1994)
R. Lallement, The interaction of the heliosphere with interstellar medium. In The Century
of Space Science, ed. by A.M. Bleeker, J. Geiss, M.C.E. Huber. (Kluwer, New York,
2001), pp. 1191–1216
R. Lallement, J.-L. Bertaux, Deceleration of interstellar hydrogen at heliopause crossing
suggested by Lyman-alpha spectral observations. Astron. Astrophys. 231, L3–L6 (1990)
R. Lallement, A.I.F.S. Stewart, Out-of-ecliptic Lyman α observations with Pioneer-Venus:
solar wind anisotropy degree in 1986. Astron. Astrophys. 227, 600–608 (1990)
R. Lallement, J.-L. Bertaux, V.G. Kurt, E.N. Mironova, Observed perturbations of the
velocity distribution of interstellar H atoms in the solar system with Prognoz Lyman α
measurements. Astron. Astrophys. 140, 243–250 (1984)
R. Lallement, J.-L. Bertaux, F. Dalaudier, Interplanetary Lyman α spectral profiles and
intensities for both repulsive and attractive solar force fields predicted absorption pattern
by a hydrogen cell. Astron. Astrophys. 150, 21–32 (1985a)
R. Lallement , J.-L. Bertaux, V.G. Kurt, Solar wind decrease at high heliographic latitudes
detected from Prognoz interplanetary Lyman α mapping. J. Geophys. Res. 90, 1413–1423
(1985b)
R. Lallement, E. Quémerais, J.-L. Bertaux, S. Ferron, D. Koutroumpa, R. Pellinen, De-
flection of the interstellar neutral hydrogen flow across the heliospheric interface. Science
307, 1447–1449 (2005)
R. Lallement, E. Queḿerais, D. Koutroumpa, J.-L. Bertaux, S. Ferron, W. Schmidt, P.
Lamy, The interstellar H flow: updated analysis of SOHO/SWAN Data. AIP Conf. Proc.
1216, 555–558 (2010)
M. Lee, H. Kucharek, E. Möbius et al., An analytical model of interstellar gas in the
heliosphere tailored to interstellar boundary explorer observations. Astrophys. J.Suppl.
198, article id 10 (2012)
P. Lemaire, C. Emerich, W. Curdt et al., Solar H I Lyman alpha full disk profile obtained
with the SUMER/SOHO spectrometer. Astron. Astrophys. 334, 1095–1098 (1998)
64 2. Interstellar Hydrogen and Backscattered Lyman-α
B.G. Lindsay, R.F. Stebbings, Charge transfer cross sections for energetic neutral atom data
analysis. J. Geophys. Res. 110, A12213 (2005)
J. Linsky, B. Wood, The alpha Centauri line of sight: D/H ratio, physical properties of
local insterstellar gas, and measurement of heated hydrogen (the “hydrogen wall”) near
the heliopause. Astrophys. J. 463, 254 (1996)
L.J. Maher, B.A. Tinsley, Atomic hydrogen escape rate due to charge exchange with hot
plasmaspheric ions. J. Geophys. Res. 82, 689–695 (1977)
Y.G. Malama, Monte-Carlo simulation of neutral atoms trajectories in the solar system.
Astrophys. Space Sci. 176, 21–46 (1991)
Y.G. Malama, V.V. Izmodenov, S.V. Chalov, Modeling of the heliospheric interface: multi-
component nature of the heliospheric plasma. Astron. Astrophys. 445, 693–701 (2006)
D.J. McComas, R.W. Ebert, H.A. Elliott et al., Weaker solar wind from the polar coronal
holes and the whole Sun. Geophys. Res. Lett. 35, L18103 (2008)
D.J. McComas, F. Allegrini, P. Bochsler et al., Global observations of the interstellar inter-
action from the interstellar boundary explorer (IBEX). Science v326, 959 (2009)
R.R. Meier, Some optical and kinetic properties of the nearby interstellar gas. Astron.
Astrophys. 55, 211–219 (1977)
D.C. Morton, J.D. Purcell, Observations of the extreme ultraviolet radiation in the night
sky using an atomic hydrogen filter. Planet. Space Sci. 9, 455–458 (1962)
H.-R. Mueller, V. Florinski, J. Heerikhuisen, V.V. Izmodenov, K. Scherer, D. Alexashov,
H.-J. Fahr, Comparing various multi-component global heliospheric models. Astron. As-
trophys. 491, 43–51 (2008)
M. Neugebauer, C.W. Snyder, Solar plasma experiment. Science 138, 1095 (1962)
M. Opher, F.A. Bibi, G. Toth, J.D. Richardson, V.V. Izmodenov, T.I. Gombosi, A strong,
highly-tilted interstellar magnetic field near the Solar System. Nature 462, 1036–1038
(2009)
E.N. Parker, Dynamics of the interplanetary gas and magnetic fields. Astrophys. J.Lett.
128, 664 (1958)
E.N. Parker, The stellar-wind regions. Astrophys. J.Lett. 134, 20–27 (1961)
T.N.L. Patterson, F.S. Johnson, W.B. Hanson, The distribution of interplanetary hydrogen.
Planet. Space Sci. 11, 767–778 (1963)
W.R. Pryor, J.M. Ajello, C.A. Barth et al., The Galileo and Pioneer Venus ultraviolet
spectrometer experiments: solar Lyman-alpha latitude variation at solar maximum from
interplanetary Lyman-alpha observations. Astrophys. J. 394, 363–377 (1992)
W.R. Pryor, J.L. Scott, I.F. Stewart et al., Interplanetary Lyman α observations from
Pioneer Venus over a solar cycle from 1978 to 1992. J. Geophys. Res. 103, 26833–26849
(1998)
W.R. Pryor, J.M. Ajello, D.J. McComas, M. Witte, W.K. Tobiska, Hydrogen atom lifetimes
in the three-dimensional heliosphere over the solar cycle. J. Geophys. Res. 108, 8034
(2003). doi:10.1029/2003JA009878
E. Quémerais, Angle dependent partial frequency redistribution in the interplanetary
medium at Lyman α. Astron. Astrophys. 358, 353–367 (2000)
E. Quémerais, V.V. Izmodenov, Effects of the heliospheric interface on the interplanetary
Lyman α glow seen at 1 AU from the Sun. Astron. Astrophys. 396, 269–281 (2002)
E. Quémerais, R. Lallement, J.-L. Bertaux et al., Interplanetary Lyman α line profiles:
variations with solar activity cycle. Astron. Astrophys. 445, 1135–1142 (2006)
E. Quémerais, V.V. Izmodenov, D. Koutroumpa, Y.G. Malama, Time dependent model of
the interplanetary Lyman α glow: applications to the SWAN data. Astron. Astrophys.
448, 351–359 (2008)
D. Ruciński, M. Bzowski, Modulation of interplanetary hydrogen density distribution during
the solar cycle. Astron. Astrophys. 296, 248–263 (1995)
N.A. Schwadron, M. Bzowski, G.B. Crew et al., Comparison of interstellar boundary ex-
plorer observations with 3D global heliospheric models. Science 326, 966 (2009)
H. Sherer, M. Bzowski, H.-J. Fahr, D. Ruciński, Improved analysis of interplanetary HST
Ly-a spectra using time-dependent modelings. Astron. Astrophys. 342, 601 (1999)
Bibliography 65
I.S. Shklovsky, On hydrogen emission in the night glow. Planet. Space Sci. 1, 63–65 (1959)
E.C. Stone, A.C. Cummings, F.B. McDonald et al., Science 309, 2017–2020 (2005)
B. Strömgren, The physical state of interstellar hydrogen. Astrophys. J. 89, 526–547 (1939)
T. Summanen, R. Lallement, J.-L. Bertaux, E. Kyrölä, Latitudinal distribution of solar
wind as deduced from Lyman α measurements: an improved method. J. Geophys. Res.
98, 13215–13224 (1993)
T. Summanen, The effect of the time and latitude-dependent solar ionisation rate on the
measured Lyman α-intensity. Astron. Astrophys. 314, 663–671 (1996)
T. Terasawa, Energy spectrum and pitch angle distribution of particles reflected by MHD
shock waves fast mode. Planet. Space Sci. 27, 193–201 (1979)
G.E. Thomas, Properties of nearby interstellar hydrogen deduced from Lyman α sky back-
ground measurements. In C.P. Sonett, P.J. Coleman, J.M. Wilcox (eds.), Solar Wind,
Scientific and Technical Information Office, National Aeronautics and Space Administra-
tion., Washington, p. 661 (1972)
G.E. Thomas, R.F. Krassa, OGO-5 measurements of the Lyman α sky background. Astron.
Astrophys. 11, 218 (1971)
M.K. Wallis, Local interstellar medium. Nature 254, 202–203 (1975)
B.Y. Welsh, Warm and hot gas in the local ISM. Space Sci. Rev. 143, 241–252 (2009)
L. Williams, D.T. Hall, H.L. Pauls, G.P. Zank, The heliospheric hydrogen distribution: a
multifluid model. Astrophys. J. 476, 366–384 (1997)
N. Witt, P.W. Blum, J.M. Ajello, Solar wind latitudinal variations deduced from Mariner
10 interplanetary H 1216 A observations. Astron. Astrophys. 73, 272–281 (1979)
N. Witt, P.W. Blum, J.M. Ajello, Polar solar wind and interstellar wind properties from in-
terplanetary Lyman-alpha radiation measurements. Astron. Astrophys. 95, 80–85 (1981)
F.M. Wu, D.L. Judge, Temperature and flow velocity of the interplanetary gases along solar
radii. Astrophys. J. 231, 594–605 (1979)
—3—
Stéphane Ferron
ACRI-ST,
Guyancourt, France
Peter Bochsler
Space Science Center & Department of Physics,
University of New Hampshire, Durham, NH, USA
Physikalisches Institut, University of Bern, Bern, Switzerland
David J. McComas
Southwest Research Institute, San Antonio, TX, USA
University of Texas at San Antonio,
San Antonio, TX, USA
Abstract
The goal of the working group on cross-calibration of past and present ultraviolet
(UV) datasets of the International Space Science Institute (ISSI) in Bern,
Switzerland was to establish a photometric cross-calibration of various UV and
F (r)
v · ∇r f (v, r) + · ∇v f (v, r) = −β f (v, r) , (3.1)
mH
where ∇x is the gradient operation in the x-direction, r and v are position and
velocity vectors of the gas cell element, and mH is the hydrogen atom mass.
Physics of neutral gas in the inner heliosphere 69
Together with the assumption that the gas “at infinity” (in practice: a few
hundreds of AU from the Sun or at solar distances relevant for heliospheric models)
is homogeneous, collisionless, and Maxwellian (Izmodenov et al. 2000), this is the
basis for the classical hot model of the distribution of neutral interstellar gas in
the heliosphere (Thomas 1978; Fahr 1978, 1979; Wu and Judge 1979; Lallement
et al. 1985b). Based on Liouville’s theorem, the solution of this equation (Danby
and Camm 1957) for the distribution function f (v, r, t) for time t, location r and
velocity v can be expressed as:
where W (v, r, t) is the survival probability of an atom that arrives at the time t at
location r with velocity v from a distant location r∞ where its velocity was v∞ .
For now, t is only a formal parameter here. The probability of existence of such an
atom in the distant region of the heliosphere is given by the distribution function
f∞ (v∞ , r∞ ), and the link between the local velocity and position vectors v (t),
r (t) and the corresponding vectors in the so-called source region of the atoms can
be obtained from the solution of the equation of motion of hydrogen atoms in the
heliosphere:
G mH M (1 − μ (vr , t)) r
F (r, t, vr ) = − . (3.3)
r2 r
In this expression, F is the total force acting on the atom with mass mH , G is
the gravitational constant, M the solar mass, vr the radial velocity of the atom at
time t, and μ is the ratio of solar resonant radiation pressure force to solar gravity.
Radiation pressure will be more fully discussed in the section “Radiation Pressure
and Its Variations”.
As can be seen from this description, the distribution of neutral interstellar
hydrogen in the inner heliosphere is determined on one hand by the dynamical
influence of the Sun through the counteracting gravity and radiation pressure forces,
and on the other hand by the ionization losses, collectively denoted β in Eq. 3.1.
Both will be extensively discussed later in this paper. Here we only note that the
ionization processes include charge exchange between the incoming neutral atoms
and solar wind protons, ionization by impact of solar wind electrons, and ionization
by the solar EUV radiation.
There are no important sources of neutral gas in the region of velocity phase
space occupied by neutral interstellar gas, hence the lack of source terms in Eq. 3.1.
Recombination could potentially be considered as such a source, but is not impor-
tant for two reasons: (1) its rate is small in comparison with the ionization rate
(Wachowicz 2006), and (2) the recombined solar wind particles maintain their pre-
reaction velocities, which are equal to solar wind velocity. They do not contribute
to the population of heliospheric atoms capable of scattering the solar FUV radia-
tion which are responsible for the helioglow. Recombination is one of the secondary
sources of the so-called Neutral Solar Wind (Bleszyński et al. 1992; Gruntman 1994;
Bzowski and Ruciński 1996; Ruciński et al. 1998), which is beyond the scope of
this chapter.
70 3. Solar Parameters
The classical hot model is almost analytical (in fact, numerical calculations are
needed only when integrating the local distribution function to yield its moments,
such as density and mean velocity) and thus convenient to use, but it is far from
being perfect because many of its assumptions are not valid.
First, the interstellar gas in the Local Cloud is not fully neutral. The interaction
of its ionized component with the plasma of the solar wind creates a boundary
region of the heliosphere: the heliospheric interface. This interface begins at the
termination shock of the solar wind, where the solar wind becomes subsonic and
eventually turns back at the heliopause. The heliopause can be approximated as a
thin layer separating the solar wind plasma from the interstellar plasma. Beyond
the heliopause there is the outer heliosheath, where the pristine neutral interstellar
gas is altered due to charge exchange interactions with protons from the piled-up
and heated interstellar plasma. The history of the development of modeling of this
region of the heliosphere can be found in Baranov (2006b) and modern views on
this topic have been recently reviewed by Fahr (2004), Baranov (2006a), Izmodenov
and Baranov (2006), and by Izmodenov et al. (2013, this volume). Second, the
solar factors are not stationary or spherically symmetric, as will be demonstrated
in the remaining portion of this chapter.
With these two observations in perspective it can be easily understood that
quantitative interpretation of measuremeints of the helioglow require improvements
in the classical hot model, which were realized quite early in the history of helio-
spheric research.
Lallement et al. (1985b) allowed for latitudinal modulation of the charge ex-
change rate, approximating it with a one-parameter formula: 1 − A sin2 φ. This
enabled them to vary the equator-to-pole ratio of the ionization rates, but required
keeping the width and range of the equatorial region of enhanced ionization fixed.
A different extension of the hot model was proposed by Ruciński and Fahr (1989,
1991) who pointed out that the rate of electron-impact ionization does not scale
as r −2 , even though its effects on the distribution of neutral interstellar hydrogen
in the heliosphere are noticeable only within a few AU from the Sun, where its
density is already strongly reduced. This aspect of the heliospheric physics has
been neglected until Bzowski (2008) and Bzowski et al. (2008) reintroduced it in a
refined, latitude-dependent manner.
The next generation of heliospheric models abandoned the assumption of in-
variability of solar radiation pressure and ionization rate. The first, though very
simplified, model was proposed by Kyrölä et al. (1994), followed by Ruciński and
Bzowski (1995); Bzowski and Ruciński (1995a,b); Bzowski et al. (1997). They
studied variations in density, bulk velocity, and temperature of neutral interstellar
hydrogen near the Sun, as well as variations in the helioglow intensity. Because
of the lack of sufficient observational data at that time, they adopted an analytic
model of the evolution of radiation pressure and ionization rate over the solar cycle.
Another modification to models of the heliosphere was introduced by Scherer
et al. (1999), who addressed the prediction by Baranov et al. (1991); Osterbart and
Fahr (1992); Baranov and Malama (1993) [see also (Malama et al. 2006)] that the
charge exchange processes in the boundary layer of the heliosphere create a new,
so-called secondary, collisionless population of neutral H atoms. They modified
Physics of neutral gas in the inner heliosphere 71
6 × 1011
5 × 1011
4 × 1011
3 × 1011
1950 1960 1970 1980 1990 2000 2010
time [y]
Figure 3.1: Wavelength- and disk-integrated solar Lyman-α flux from the Labora-
tory for Atmospheric and Space Physics (LASP), referred to as the total Lyman-α
flux Itot . The daily time series is a composite of actual measurements from vari-
ous experiments re-scaled to the common calibration of UARS/SOLSTICE, with
the gaps filled by proxies. Color codes: gray: F10.7 proxy, orange: AE-E, pur-
ple: SME, green: MgIIc/w proxy, aquamarine: UARS/SOLSTICE version 18, red:
TIMED/SEE, blue: SORCE/SOLSTICE (based on Woods et al. 2000)
to depend on the source proxy, the difference between the change rates derived
from the F10.7 proxy and the proxies based on FUV measurements of the Sun is
especially pronounced.
As a consequence, the quality of the approximation at these short scales is
reduced even though it may be quite satisfactory at longer time scales, e.g., for
solar rotation period averages, as can be judged from the behavior of the monthly
change rates shown in the lower panel of Fig. 3.2 and as suggested by Dudok de
Wit et al. (2009). For modeling the distribution of neutral interstellar gas in the
heliosphere, the Carrington period of solar rotation is the finest time scale presently
in use,1 so the short time scale proxy imperfections are not a big problem for this
purpose.
−1 × 1010
−2 × 1010
−3 × 1010
1950 1960 1970 1980 1990 2000 2010
time [y]
Monthly change rate in the composite Lyman−α flux
3 × 1010
ΔF/Δt [cm−2 s−1 month−1]
2 × 1010
1 × 1010
0
−1 × 1010
−2 × 1010
−3 × 1010
1950 1960 1970 1980 1990 2000 2010
time [y]
Figure 3.2: Rates of change of the total Lyman-α flux Itot . Shown are the rates
of change per unit time ΔI tot
Δt = Itot (ti+1 )−Itot (ti )
tt+1 −ti of the daily (upper panel) and
monthly (lower panel) composite Lyman-α flux presented in Fig. 3.1, with identical
color coding
should be about aLya = 0.8 during solar minimum, when the active regions are
distributed in latitudinal bands. These suggestions were supported by direct solar
minimum observations of the solar corona by Auchère (2005). Such a ratio was
suggested to be also valid for solar maximum by Pryor et al. (1992), based on
indirect evidence from observations of the Lyman-α helioglow. Thus it seems that
the latitudinal anisotropy of the Lyman-α flux does not change substantially during
the solar cycle, although this conclusion certainly needs further verification.
Bzowski (2008) suggested that the latitude dependence of the disk-integrated
solar Lyman-α flux may be approximated by the formula:
Itot (φ) = Itot (0) aLya sin2 (φ) + cos2 (φ), (3.4)
where φ is heliolatitude, aLya the “flattening” factor and Itot (0) is the equatorial
Lyman-α flux.
Heliolatitude variation of the disk-integrated Lyman-α line profile is, to our
knowledge, unexplored. On one hand, some form of variation should be expected
because, as shown by Tian et al. (2009a,b,c), the line profile depends on the features
on the solar disk that are being observed and the latitude distribution of these
Radiation Pressure and Its Variations 75
features is inhomogeneous and varies during solar cycle. On the other hand, the
apparent lack of strong variability of the disk-integrated flux might suggest that the
spectral variation with heliolatitude is mild. Should the disk-integrated spectral
flux indeed vary with heliolatitude, this would potentially have consequences both
for the photoionization rate and the radiation pressure force. However, recent
investigation of the variation of the solar spectrum with heliolatitude by Kiselman
et al. (2011) seems to have brought a negative result (i.e., no variation).
where ΓR is the energetic width of the second orbital of the atom, corresponding
to a Doppler width of about ±25 m s−1 around the rest wavelength of the Lyman-
α transition, and c is speed of light. An atom whose radial velocity relative to
the Sun is vr = 0 will absorb photons from the very center of the solar line at
λ0 , but if its radial velocity vr is non-zero, then due to the Doppler effect it will
be tuned to a different portion of the solar line profile, namely to the wavelength
λ = λ0 (1 − vr /c). Within about 10−7 s after it absorbs a Lyman-α photon, the
atom will re-emit the photon at an angle ω relative to the impact direction with
the scatter-angle probability p (ω) described by:
cos (ω) 11
p (ω) = + . (3.6)
4 12
Hence, a resonant interaction of the atom with a suitable photon results in a change
of atomic momentum at the moment of absorption by Δp = hv = c h/λ in the
antisolar direction, followed after a time of 10−7 s by another momentum change in
the direction described by Eq. 3.6. However, the typical frequency of interactions
between solar photons and H atoms at 1 AU, which is proportional to the solar
spectral flux, is on the order of 1/500 Hz (Quémerais 2006). Statistically, on time
scales shorter than the time scales of a change in atomic velocity relative to the Sun,
the only dynamical net effect of the interaction of the atom with solar radiation is
the antisolar momentum change.2
Since the interplanetary medium is optically thin within a few AU from the
Sun (Quémerais 2006), the solar spectral flux scales as the inverse square of solar
distance. Consequently, the solar radiation pressure also scales with solar distance
2 One can expect another statistical effect: an increase in the local velocity spread in the
population of neutral H gas in the heliosphere, but, to our knowledge, this effect has not been
studied in the available literature.
76 3. Solar Parameters
as 1/r 2 , thus leading to a partial compensation of solar gravity. Hence the effective
solar force acting on an atom is conveniently expressed by the fraction μ (Itot (t) , vr )
of the solar gravity force. It is proportional to the spectral flux Fλ corresponding
to the Doppler-shifted wavelength λ = λ0 (1 + vr /c), which results from the instan-
taneous radial speed vr of the atom relative to the Sun. Since the spectral flux
varies with time, effectively the μ factor is a function of radial velocity and time,
as expressed in Eq. 3.3.
1.5
gravity compensation factor m
1.0
0.5
0.0
–200 –100 0 100 200
radial velocity [km s–1]
Figure 3.3: Ratio μ of solar radiation pressure force to solar gravity based on
the model specified in Eq. 3.7, shown as a function of radial velocity of a H atom
relative to the Sun for the total solar flux values corresponding to the minimum
(red) and maximum (blue) of solar activity. Thick lines indicate the spectral region
±30 km s−1 around 0 Doppler shift, relevant for the neutral interstellar hydrogen
gas in the heliosphere
2.0
1.8
1.6
m factor
1.4
1.2
1.0
0.8
Figure 3.4: Solar radiation pressure factor μ as a function of the total flux in the
Lyman-α line Itot . The blue line represents the relation from Eq. 3.9 by Emerich
et al. (2005), which connects the total flux Itot with the spectral flux precisely at
line center Fλ0 . The conversion to the μ factor is then performed using Eq. 3.8,
with a = 0.9. The red line shows the relation defined in Eq. 3.10 by Bzowski et al.
(2008). Here, the relation between the total solar flux and spectral flux is adopted
from averaged spectral flux over ±30 km s−1 around the line center
Figure 3.4 shows a comparison of the predicted μ values as a function of the solar
total Lyman-α flux obtained from Eqs. 3.8–3.10. It suggests that if one decides not
to use a model radiation pressure force which is dependent on the radial velocity of
the atom, then the calculation of a good effective μ factor is not a straightforward
task. In fact, it may be appropriate to use different formulae for the upwind and
downwind regions in the heliosphere, as can be inferred from the asymmetry of the
solar Lyman-α line profile seen in Fig. 3.3. Such an approach, to our knowledge,
has never been implemented. Calculating μ as a function of vr is required in order
to model interstellar deuterium in the inner heliosphere (Tarnopolski and Bzowski
2008b).
The μ values obtained from the relations defined in Eqs. 3.9 (Emerich et al.
2005) and 3.10 (Tarnopolski and Bzowski 2008b), calculated from the monthly
values3 of the LASP composite Lyman-α flux (Fig. 3.1) are presented in Fig. 3.5.
Differences between the values obtained from these equations are on the order of
10 %, i.e. on the order of the uncertainty of the total flux.
3 Throughout the text, “monthly” is used as synonymous with “averaged over one Carrington
rotation period”.
Ionization Processes 79
2.0
1.8
1.6
m factor
1.4
1.2
1.0
Figure 3.5: Comparison of the solar radiation pressure factors μ that approximate
the compensation of solar gravity by the resonant radiation pressure force acting on
neutral H atoms. The green line represents monthly values obtained from Eq. 3.9
(Emerich et al. 2005) and the red line shows the monthly values obtained from
Eq. 3.10 (Bzowski et al. 2008). The blue line represents the μ values calculated
from Eq. 3.8 with a = 0.9. The gray dots represent daily values of the μ factor,
calculated from daily values of the composite Lyman-α flux using Fig. 3.8
Ionization Processes
The three main ionization processes of neutral interstellar hydrogen atoms in
the heliosphere are the following:
charge exchange with solar wind charged particles (mostly protons) resulting in a
pickup proton (pPUI ) and an energetic neutral atom (ENA):
H + p → pPUI + HENA ,
H + ν → pPUI + e,
H + e → pPUI + 2e.
a distinct population that can be measured (Möbius et al. 1985; Gloeckler et al.
1993) and analyzed (e.g. Gloeckler and Geiss 2001; Gloeckler et al. 2004; Bzowski
et al. 2008), but this topic is outside the scope of this text.
The charge exchange reaction does not deplete solar wind protons. A proton
from the core of the solar wind distribution function is replaced with a pickup
proton. The concentration of protons per unit volume is not changed, although the
distribution function of solar wind protons is modified.
The two remaining ionization reactions do cause genuine losses for the entire H
population: a H atom enters the reaction and is not simply replaced with another
one at a different velocity. Eventually, a proton–electron pair is created and the
proton is picked up by solar wind.
Charge Exchange
General Formula
The process of resonant charge exchange between H atoms and protons is of
crucial importance for the physics of the heliosphere. It contributes to the pressure
balance between the solar wind and the interstellar gas and enables the energy
and momentum transport across the heliopause. Charge exchange losses of neutral
interstellar gas in the supersonic solar wind, which is discussed in this section,
are only a small piece of a larger picture of the role of charge exchange in the
heliosphere.
The rate of charge exchange between neutral H atoms and solar wind protons
can be regarded as the probability of a charge exchange act within unit time in
a given location in space. For an H atom traveling with velocity v H and a local
proton distribution function fp (vp ), where v p is the velocity vector of an individual
proton, the rate of charge exchange can be calculated from the formula:4
βCX = σCX (|v H − vp |) |vH − v p | fp (v p ) dvp , (3.11)
where v H − v p ≡ v rel is the relative velocity between the H atom and an individual
proton and σCX (|v H − v p |) is the reaction cross section. The integration covers
the entire proton velocity space. This formula can be put into an equivalent form:
βCX = σCX (|v rel |) |vrel | fp (vH − v rel ) dv rel . (3.12)
Depending on the underlying plasma regime and on the velocity of the H atom,
various simplifications can be made. When the kinetic spread of the plasma uT,p
is small compared with the plasma flow velocity vSW :
uT,p |v SW |
2kTp /mp
v p f (v p ) dv p /np
,
4 We adopt a convention where bold-italic characters mean vector quantities, while italics
symbolize scalars.
Ionization Processes 81
where np is the local proton density and vrel becomes vrel ≡ |v H − vSW |. This is
the case for ENAs that travel in the supersonic solar wind at vH ∼ 50 km s−1 or
faster.
For vH vSW , i.e., for atoms from the thermal interstellar H populations in
the supersonic solar wind, vrel
vp . Then the rate of charge exchange between H
atoms and solar wind protons is given by:
This is the baseline formula for charge exchange rate between protons from the
supersonic solar wind and neutral interstellar H atoms. It has been widely used
in the heliospheric physics and will be used in the later part of this chapter. The
accuracy of this approximation is subject of one of the following subsections.
The valid range for the Fite et al. (1962) relationship was claimed to be between 20
and 2,000 eV. The domain of the Maher and Tinsley (1977) expression was taken
to be from 0.005 to 1 keV.
Barnett et al. (1990) fit a form of Chebyshev polynomials which were valid in
a broad energy range. Bzowski (2001b) approximated the data used by Barnett
et al. (1990) but restricted to vrel < 800 km s−1 by the following expression:
3
i
σCX (vrel ) = a0 + ai (ln vrel ) . (3.16)
i=1
4. × 10−15
2. × 10−15
0
10 20 50 100 200 500 1000 2000
collision velocity [km s−1]
Figure 3.6: Cross sections for charge exchange reaction between protons and H
atoms in the energy range most important in the heliospheric physics. The recom-
mended relation from Lindsay and Stebbings (2005) is compared with the formulae
used in the past by Fite et al. (1962), Maher and Tinsley (1977) and Barnett et al.
(1990)
for the cross section expressed in cm2 , valid for collision energies E between 0.005
and 600 keV:
A comparison of the cross sections from the four formulae is presented in Fig. 3.6.
It is important to note that while all four formulae return similar results for
the collision speeds relevant to the supersonic solar wind, the one from Lindsay
and Stebbings (2005) returns a significantly larger cross section for lower energies,
which are relevant for the outer heliosheath. Thus, adoption of the older formulae
may result in a significant underestimation of the coupling strength between the
neutral interstellar gas and the plasma in the outer heliosheath where the secondary
population of interstellar H atoms is formed. This would have marked consequences
for the results of heliospheric modeling, as described by Izmodenov et al. (2012,
this volume).
Since the solar wind speed and density vary with time on time scales from
minutes to centuries, the instantaneous values of charge exchange rate vary on the
same time scales. Theory immediately suggests that to calculate the losses of the
neutral atom population due to charge exchange over a time interval ΔT , one has
to integrate the instantaneous rate given by Eq. 3.14 over this interval. The mean
charge exchange rate βΔT over interval ΔT is calculated from:
1
βΔT = σCX (vSW (t)) np (t) vSW (t) dt. (3.18)
ΔT ΔT
In practice, however, this strict approach is usually not possible to follow be-
cause high-resolution data on solar wind density and speed throughout the helio-
sphere are not available. Hence, a simplified version of the Carrington averaging is
adopted:
βCX Carr = np Carr vSW Carr σCX (vSW Carr ) , (3.19)
Plane”.
84 3. Solar Parameters
1 × 10−6
ionization rate [s−1]
5 × 10−7
1 × 10−7
5 × 10−8
Figure 3.7: Monthly ionization rates in the ecliptic plane from all relevant processes.
Blue marks the charge exchange rate, calculated from Eq. 3.14 using the cross
section from Eq. 3.17 (Lindsay and Stebbings 2005) and monthly averages of solar
wind speed and density shown in Fig. 3.8. Green represents the photoionization
rate (Bochsler et al. 2012), and orange the electron-impact ionization rate (Bzowski
2008). The total ionization rate, being a sum of the three aforementioned rates, is
shown in black. Note that the vertical axis is logarithmic. To better highlight the
secular change in the charge exchange rate after the turn of the past century, an
exploded view of the charge exchange rate for the time interval from 1985 through
2011 is shown in Fig. 3.15
(Bzowski et al. 1997). The modification of the charge exchange rate due to the
proper motion of H atoms can be assessed as follows. The atoms approaching the
Sun, i.e., mostly in the upwind hemisphere, run against solar wind and thus the rela-
tive speed is the sum of the proper velocity of the atoms and of the solar wind speed.
For a typical solar wind speed in the ecliptic plane of ∼ 440 km s−1 , the change in
the charge exchange rate is by (440 + 30) σCX (440 + 30) / (440 σCX (440))
1.035.
Similarly, for the atoms in the downwind hemisphere which recede from the Sun,
the change in charge exchange rate is 0.965 (for σCX defined in Eq. 3.17). Thus,
the error induced by the approximation of stationary atom is about ±3.5 %. It
systematically varies with the offset angle from the upwind direction, transitioning
from an underestimation in the upwind hemisphere to an overestimation in the
downwind hemisphere.
All in all, the errors in monthly charge exchange rates due to the approxima-
tions presented are of similar magnitude. Since they are independent sources of
error, they can be added in quadrature and total about ∼ 5.5 %. This should be
contrasted with the uncertainty of ∼ 10 % related to the uncertainty of the cross
section alone.
Another approximation frequently made in the heliospheric modeling is that the
charge exchange rate decreases with the square of solar distance. This assumption
Ionization Processes 85
originates from the inverse-quadratical reduction of solar wind density with dis-
tance, with approximately constant velocity. We will assess now how good this
approximation is.
The solar wind expands almost radially (i.e., its non-radial velocity components
are very small in comparison with the radial component). It expands with basically
constant speed between the outer boundary of the acceleration region near the Sun
(located inside a few solar radii) and approximately 10 AU, where the slowdown
effects of mass loading due to the ionization of neutral interstellar gas become
measurable (Fahr and Ruciński 1999, 2001, 2002; Lee et al. 2009; Richardson et al.
1995, 2008b). The overall slowdown continues up to a few AU upstream from
the termination shock. At that point, the flow speed has already been reduced
by about 67 km s−1 relative to the speed at 1 AU (Richardson et al. 2008b). It is
additionally slowed down by the component of protons reflected at the termination
shock (Liewer et al. 1993; Richardson et al. 2008a). The effect of pickup ions
on distant solar wind varies with the angle from the inflow direction (Fahr and
Ruciński 1999), but the strongest effect is expected in the upwind hemisphere,
where Voyager measurements were made.
Outside the acceleration region, the solar wind flux initially falls off with solar
distance as 1/r 2 . This relation, stemming directly from the continuity equation, is
not significantly altered by the interaction of the solar wind plasma with neutral
interstellar gas. The main reaction is charge exchange—a reaction that does not
result in a change in the local proton density, only in a shift of the reaction product
into the pickup ion region of the total distribution function. Only the two secondary
reactions, photoionization and ionization by electron impact, actually inject new
protons into the pickup ion region in phase space.
As a result, the adjusted6 solar wind density, treated as a sum of the core so-
lar wind protons and pickup protons, increases very slowly with solar distance.
This was approximately assessed by Lee et al. (2009), who give the following for-
mula for the absolute density at a distance r of pickup ions nPUI,ph (r) created by
photoionization
2
rE nH,TS βph,E
nPUI,ph (r) = , (3.20)
r vSW,E
where βph,E is the photoionization rate at rE = 1 AU, vSW,E the solar wind speed
at rE , and nH,TS the neutral interstellar H density at the termination shock.
For the H density at the termination shock located at 90 AU equal to 0.087 cm−3
(Bzowski et al. 2008), the photoionization rate is 10−7 s−1 (see below), and the
typical solar wind speed is 440 km s−1 . The density of pickup ions at 90 AU from the
Sun is then equal to 3.3 10−5 cm−3 , which scales to 0.27 cm−3 when quadratically
adjusted to 1 AU. For a typical solar wind density at 1 AU equal to 5 cm−3 , this
yields a ∼ 5 % excess of the total solar wind density with respect to the pure 1/r 2
drop.
Inserting this modified adjusted density into Eq. 3.14, along with the reduced
solar wind speed of 440 − 67 = 373 km s−1 , one obtains an adjusted charge ex-
change rate of 3.7 10−7 s−1 , which is ∼ 3.5 % less than the rate calculated for the
6 Throughout this chapter, we refer to various quantities as “adjusted” meaning that we take
their magnitudes scaled by r2 , i.e., multiplied by the square of solar distance expressed in AU.
86 3. Solar Parameters
unenhanced and undecelerated solar wind. Hence, the modification of the charge
exchange rate in the outer supersonic solar wind from the 1/r 2 decrease is on the
order of the uncertainty related to the averaging and approximating the charge
exchange rate or less.
The remaining issue is whether or not using Eq. 3.14 to calculate the charge ex-
change rate of neutral H with pickup ions is justified. Pickup ions in the supersonic
solar wind make a special case in the calculation of the charge exchange rate. Even
though their bulk speed is equal to the solar wind expansion speed, the width of
their distribution function is comparable to the magnitude of the expansion speed
(Vasyliunas and Siscoe 1976; Möbius et al. 1988) and consequently the full version
of the formula for the charge exchange rate, defined in Eq. 3.12, should in principle
be used.
The simplification of using Eq. 3.14 for pickup ions seems acceptable because
it is not expected to introduce major errors. Close to the Sun (within ∼ 10 AU),
their abundance in the solar wind is small and consequently their contribution to
the total charge exchange rate is negligible. Outside of this region, their content
in the solar wind increases, but the total charge exchange rate becomes small.
Consequently, the contribution of charge exchange operating in this region of space
to the total losses of neutral interstellar gas observed from the inner heliosphere, es-
pecially via helioglow measurements, is also small. Another issue in the calculation
of charge exchange rate may be departures of the solar wind flow from purely ra-
dial expansion at larger heliocentric distances, which could result in a change of the
latitudinal structure with solar distance, as suggested by Fahr and Scherer (2004).
Detailed studies to support this qualitative discussion are missing in the literature,
as far as we know, except for some insight provided by Bzowski and Tarnopol-
ski (2006) on the radial profiles of survival probabilities of ENAs approaching
the Sun.
In our opinion, consistently taking all these effects into account would require
using a comprehensive, multi-fluid, three-dimensional and time-dependent model
of solar wind. Such a model should include the core solar wind, electrons, and
pickup ions as well as the solar wind magnetic field and be able to address both
the large-scale behavior of the solar wind flow and small-scale disturbances such
as turbulence. As boundary conditions, it should take actually and continuously
measured solar wind parameters as close to the corona as practical. To our knowl-
edge, such a model is now in development (see Usmanov et al. 2011 and references
therein).
Photoionization
From Solar Spectrum to Photoionization Rate
Photoionization is a secondary ionization factor of neutral interstellar H, but
its significance has recently increased because of the decrease in the average solar
wind flux observed since the last solar maximum (see upper panel in Fig. 3.8).
This has resulted in a decrease in the intensity of the dominant charge exchange
ionization rate.
Ionization Processes 87
12
density [cm−3]
10
4
1985 1990 1995 2000 2005 2010
time [y]
550
500
450
400
350
1985 1990 1995 2000 2005 2010
time [y]
6 × 108
flux [cm−2 s−1]
6 × 108
6 × 108
6 × 108
1985 1990 1995 2000 2005 2010
time [y]
The rate of photoionization βph (t) at a time t can be calculated from the
formula:
λion
1
cross section [Mb]
0.01
10−4
10−6
0 20 40 60 80
wavelength [nm]
Figure 3.9: Cross section for hydrogen photoionization based on Eq. 3.22, adapted
from Verner et al. (1996)
where σph (λ) is the cross section for photoionization for wavelength λ and Fλ (λ, t)
is the solar spectral flux at the time t and wavelength λ. λion is the wavelength for
the threshold ionization energy. In the case of hydrogen, the spectral range of the
radiation capable of knocking out electrons from H atoms is entirely in the EUV
range. The cross section for photoionization of H can be expressed by the following
formula (Verner et al. 1996):
−2.963
−10 9.36664
σHph (λ) = 6.82297 × 10 √ +1 (λ − 2, 884.69) 2 λ2.0185 , (3.22)
λ
where the cross section is expressed in megabarns (Mb), wavelength in nm, and
λ ≤ 91.18 nm. The cross section is shown in Fig. 3.9, and the importance of various
portions of the spectrum for the photoionization is illustrated in Fig. 3.10. It shows
the integrand function from Eq. 3.21 for day of year 122 in 2001 (Bochsler et al.
2012).
It is clear from Fig. 3.10 that the most important portion of the spectrum for
the photoionization of hydrogen is the one immediately above the photoioniza-
tion threshold. The photoionization rate varies because the solar EUV spectral
flux varies. Direct measurements of the solar EUV spectrum in the entire rele-
vant energy range have been available on a 2-hourly basis since 2002, when the
TIMED/SEE experiment (Woods et al. 2005) was launched. Before then, the
spectral coverage was intermittent and one had to resort to indirect methods to
estimate the solar EUV flux.
Basically, these methods can be grouped into two classes: (1) direct integration
using Eq. 3.21 with the spectrum Fλ (λ, t) calculated from proxies, and (2) using
correlation formulae between selected proxies and photoionization rates obtained
for the times when the spectrum measurements are available. Since measurements
covering only a portion of the spectrum with a relatively low spectral resolution are
also available (CELIAS/SEM onboard SOHO; Hovestadt et al. 1995; Judge et al.
1998), a variant of method (2) would require finding a correlation between these
partial direct measurements and the photoionization rate. Neither of the methods
Ionization Processes 89
0.00010
0.00005
0.00000
0 20 40 60 80
wavelength [nm]
Figure 3.10: Integrand function from Eq. 3.21 for the flux obtained from
TIMED/SEE for 2010 DOY 122. The photoionization cross section used is de-
fined in Eq. 3.22 and shown in Fig. 3.9
Using this formula, Bochsler et al. (2012) calculated the Carrington averages of
photoionization rates for the entire interval for which the SEM data were available.
For the times when SEM data was unavailable, but the MgIIc/w index from LASP
was, they used another correlation formula:
For epochs where the MgIIc/w index was unavailable, Bochsler et al. (2012) used
the following correlation formula with the F10.7 flux expressed in sfu units (i.e.,
10−22 W m−2 Hz−1 ):
It is worth pointing out that the exponent at the F10.7 flux is close to 1/2, not 1,
as is frequently adopted. The results of the model with color coding of the sources
used is shown in the lower panel of Fig. 3.11 for the time interval since 1990 until
present.
The photoionization rate obtained in the way described above can be compared
with the rate from the SOLAR2000 model, used extensively in previous studies
(e.g. Bzowski et al. 2009). The comparison is shown in the upper panel of Fig. 3.11.
The two models agree to about 10–15 %, with the direct integration model giving
Ionization Processes 91
1.4 × 10−7
1.2 × 10−7
1.× 10−7
8.× 10−7
1950 1960 1970 1980 1990 2000 2010
time [y]
SOL2000 vs direct integration relative difference
0.10
0.05
0.00
−0.05
−0.10
−0.15
−0.20
−0.25
1950 1960 1970 1980 1990 2000 2010
time [y]
Monthly photoionization rates of hydrogen:
SOLAR2000 vs present model
1.6 × 10−7
photoionization rate [Hz]
1.4 × 10−7
1.2 × 10−7
1.× 10−7
8.× 10−7
almost always higher values. Such an accuracy is basically equal to the present
accuracy of the EUV measurements, especially in the low-energy portion of the
spectrum which contributes most to the ionization.
It seems that the cause of these small discrepancies is the difference in the way
the photoionization rates are calculated by SOLAR2000 and by Bochsler et al.
(2012). SOLAR2000 first calculates the solar spectrum from their sophisticated
system of proxies and then integrates the spectra to yield the ionization rate. The
quality of the derived spectrum seems to worsen when going backwards in time.
This is understandable because it is gradually based on fewer available measure-
ments of solar proxies. Bochsler et al. (2012) used measured solar spectra for the
time intervals when they were available and the absolute calibration was most cred-
ible. For other times they used proxies and correlation formulae specially developed
to connect a proxy measurement with the photoionization rate of a given species.
In this sense, technically, it is not a hierarchy of proxies, where one proxy model
is based on another proxy model. Rather, it is a collection of proxy models, which
all are based on one reliable series of direct measurements.
The lower panel of Fig. 3.11 presents the best estimate of the hydrogen pho-
toionization rate obtained by Bochsler et al. (2012) from their collection of proxies
(the colored line) with the rate calculated solely from the F10.7 proxy (black line)
and with the rate presented by SOLAR2000 (gray line). Particularly interesting
is the comparison of the model from the F10.7 proxy with the results of direct
integration of the solar spectrum (red vs black line), and with the SOLAR2000
results. While the F10.7 proxy model exactly tracks the direct rate, showing only
some departures above or below the red line, SOLAR2000 is consistently below.
Also outside the interval when the direct TIMED spectra could be used, the
F10.7 proxy model tracks quite well the best proxy model of Bochsler et al. (2012),
which leads us to believe that the system developed by these authors is self-
consistent and reliable, unless the relation of the F10.7 radio flux to the solar EUV
spectrum changed between 1948 and 1979, when the Mg IIc/w proxy became avail-
able. We are not aware of any such change described in the literature.
On the other hand one has to remember that the model by Bochsler et al.
(2012) is only able to calculate Carrington period averages of the ionization rate
of hydrogen and a few other selected species (He, O, Ne), while SOLAR2000’s
ambition is to provide an estimate for an arbitrarily selected day within its validity
interval. In fact, it allows one to calculate the ionization rate of any species since
it produces an approximate solar spectrum at a resolution of 1 nm. Given all the
challenges it has to face, it seems to be doing it remarkably well.
be expected and that some north-south asymmetry cannot be ruled out. In the
absence of a complete model and sufficient data we surmise that a latitude varia-
tion of hydrogen photoionization rate approximated by a formula similar to Eq. 3.4
may be tentatively adopted. The subject certainly needs further studies.
Electron Ionization
The significance of the electron-impact ionization reaction for the distribution
of neutral interstellar gas in the inner heliosphere was pointed out by Ruciński
and Fahr (1989), who developed a model of the electron ionization rate based on
the local electron temperature and density. Further insight into the problem of
electron-impact ionization of neutral interstellar hydrogen inside the heliosphere
can be found in Bzowski et al. (2008) and Bzowski (2008).
The ionization rate in the electron-impact reaction at a location described by
the radius-vector r can be calculated from the following formula (Owocki et al.
1983):
∞
8π
βel (r) = 2 σel (E) fe (E, r) E dE, (3.26)
me
Eion
where σel is the energy-dependent reaction cross section, E the collision energy
and Eion the ionization threshold energy. For hydrogen, Eion equals ∼ 13.6 eV.
Ionization occurs for the H atom—electron collisions with energies exceeding the
limiting energy Eion . Practically, almost all of the energy of the electrons in the
solar wind at a few AU from the Sun is in thermal motions.
The kinetic energy of an electron moving at a typical solar wind expansion speed
of 440 km s−1 is about 0.5 eV, which is much less than the ionization threshold.
Since the temperature of the electron fluid at 1 AU is on the order of ∼ 105 –
∼ 106 K, the thermal speeds of the solar wind electrons are on the order of a
few thousand of km s−1 , which strongly exceeds the expansion speed. Thus, the
expansion speed of the electron fluid can be neglected in the calculation of the
electron-impact ionization rate.
The formula for the cross section for electron ionization was proposed by Lotz
(1967b) and simplified for H by Lotz (1967a):
NA a q ln E
i i Pi 1 − bi exp −ci PEi − 1
σel (E) = , (3.27)
i=1
EPi
where NA is the number of electrons per ion and the summation goes over the
partial cross sections for knocking out all individual electrons from the ion. Pi
is the ionization potential for a given charge state of the ion, E is the impacting
electron energy, ai , bi , ci are parameters specific to a given ion and its charge state,
and qi is the statistical weight. For hydrogen, there is only one electron to be
knocked out and Eq. 3.27 takes the form:
E ln (E/13.6)
σel (E) = 4.0 × 10−14 1 − 0.60 exp −0.56 −1 , (3.28)
13.6 13.6E
94 3. Solar Parameters
6. × 10−17
5. × 10−17
cross section [cm2]
4. × 10−17
3. × 10−17
2. × 10−17
1. × 10−17
0
50 100 500 1000 5000
collision energy [eV]
where kB is the Boltzmann constant, me the electron mass, E the kinetic energy
of electron, Te the temperature of the electron fluid, and r the radius-vector of the
7 Note that the electron distribution function can be approximated by the kappa function—
originally by Maksimovic et al. (1997) and recently by Le Chat et al. (2010, 2011)—which naturally
covers both the core and halo components.
Ionization Processes 95
location in space. The contribution of the halo population to the net rate is on
the level of a few percent and is an increasing function of the heliocentric distance
(Maksimovic et al. 2005; Štveràk et al. 2009). Estimates by Bzowski et al. (2008)
show that at 1 AU, the ionization rate due to the core population of the solar
wind electrons is equal to about 0.4 × 10−7 s−1 and to the halo population less
than 0.04 × 10−7 s−1 , respectively. The amplitude of fluctuations in the electron
ionization rate may reach one order of magnitude, which is much more than the
long-time variations related to variations in solar activity. On the other hand, the
electron data from WIND (Salem et al. 2003) imply an in-ecliptic solar minimum
(1995) electron ionization rate of ∼ 0.68 × 10−7 s−1 and a solar maximum (2000)
rate of ∼ 0.73 × 10−7 s−1 . Thus, assuming a constant electron ionization rate over
the solar cycle is a reasonable approximation.
Observations done with Ulysses (Phillips et al. 1995a; Issautier et al. 1998; Le
Chat et al. 2011) suggest that the electron ionization rate is a 3D, time dependent
function of the solar cycle phase. Both the temperature magnitude and the cooling
rate differ between the fast and slow solar wind. Bzowski (2008) adopted the
following radial profiles of the core Tc and halo Th temperatures and the halo-to-
core density ratios ξhc = nh /nc for the slow solar wind (after Scime et al. 1994):
Tc = 1.3 · 105 r −0.85
Th = 9.2 · 105 r −0.38 (3.31)
ξhc = 0.06 r −0.25
and for the fast solar wind (after Issautier et al. 1998; Maksimovic et al. 2000):
Tc = 7.5 · 104 r −0.64
Th /Tc = 13.57 (3.32)
ξhc = 0.03.
In both solar wind regimes, the core nc and halo nh densities are calculated
from the equations:
1 + 2ξα
nc = np
1 + ξhc
nh = ξhc nc , (3.33)
with the solar wind alpha abundance ξα = 0.04, adopted to be identical in both
fast and slow wind regimes. np in this expression is the proton density. We adopt it
as velocity-independent because it introduces a relatively small modification to the
electron density, even though Kasper et al. (2012) show variations of this quantity
with solar activity as a function of solar wind speed.
Ruciński and Fahr (1989) inserted the formulae from Eqs. 3.27, 3.29, and 3.30
to the integrand function in Eq. 3.26 and obtained a formula for each of the terms
i contributing to the total electron-impact cross section, which we present here in
a slightly modified form:
⎛ ⎞
Γ 0, c + Pi exp [c ] b P
ai qi 8 me ⎝ Pi i kB Te i i i
βel,i (Te ) = ne Γ 0, − ⎠,
me Pi π kB Te kB Te Pi + ci kB Te
(3.34)
96 3. Solar Parameters
5 × 10−8
2 × 10−8
1 × 10−8
5 × 10−9
2 × 10−9
1 × 10−9
0.5 1.0 2.0 5.0 10.0 20.0
heliocentric distance [AU]
Figure 3.13: Normalized radial profiles r 2 βel (r) of the equatorial (red) and polar
(blue) rates of electron-impact ionization of neutral interstellar H atoms in the
supersonic solar wind, defined by Eqs. 3.35 and 3.36, respectively (adapted from
Bzowski 2008). Since those rates are calculated based on measurements performed
between 0.3 and 5.5 AU from the Sun, their validity is limited to this distance range
electron ionization rate in this solar wind regime is poorly known. Bzowski (2008)
approximated the electron ionization in the solar wind of intermediate velocity as
a weighted mean value between the rates specific for the fast and slow wind.
The differences in the electron fluid parameters and their variation with solar
distance, as well as the differences between the electron fluids in the fast and slow
wind result in a pronounced latitude anisotropy of the electron ionization rate
throughout the solar cycle. An exception is a brief interval at solar maximum,
when the solar wind becomes almost spherically symmetric.
The significance of the electron impact ionization in the overall balance of the
contributing ionization reactions increases towards the Sun. Eventually, the role
of electron ionization and its anisotropy is greater in the downwind hemisphere,
where neutral interstellar gas has already passed the Sun (i.e., the streamlines of
the H gas flow have passed their perihelia).
Because of the fast cooling of the electron fluid with solar distance, electron-
impact ionization almost negligibly affects the distribution of neutral interstellar
hydrogen in the upwind hemisphere at the distances beyond ∼ 2 AU (Tarnopolski
and Bzowski 2008b). This distance seems to be close to the distance to the Max-
imum Emissivity Region of the heliospheric Lyman-α backscatter glow.8 Hence,
the limited validity range of the formulae specified in Eqs. 3.35 and 3.36 seems to
not hinder modeling of the effective ionization rate of heliospheric neutral atoms
at larger distances from the Sun within the supersonic solar wind.
8 The Maximum Emissivity Region is by definition the region where the maximum of the
20 spacecraft have been available, obtained using various observations and data
processing techniques.
The highly supersonic solar wind in the ecliptic plane consists of a sequence
of various interleaved components: a “genuine” slow solar wind, a fast solar wind,
solar wind plasma from stream-stream interaction regions, and (intermittently)
interplanetary coronal mass ejections (CME). The Mach number of the flow at
1 AU varies from ∼ 3 to ∼ 10. The balance between the populations varies with
solar activity.
Historically, measurements of the solar wind speed obtained from various ex-
periments generally agreed among themselves with an accuracy of ∼ 5 %, but
systematic differences between density values from different experiments up to
∼ 30 % existed (for extensive discussion, see the OMNIWeb documentation at
http://omniweb.gsfc.nasa.gov/html/omni2 doc.html). Hence any study of a long-
term behavior of the solar wind required intercalibration of the results from dif-
ferent experiments. Such an initiative brought the OMNI data collection [King
and Papitashvili available at at http://omniweb.gsfc.nasa.gov/]; see also King and
Papitashvili (2005) where historical and present measurements of the solar wind
density, velocity, temperature, alpha abundance, and magnetic field vector were
brought to a common calibration.
Originally, the OMNI collection was created in the 1970s by the National Space
Science Data Center (NSSDC) at the NASA Goddard Space Flight Center. In 2003,
a successor database OMNI-2 was made available, which has been maintained until
present. The OMNI-2 data collection was the basis for the present study of solar
wind parameters in the ecliptic and is used in this chapter to construct the charge
exchange ionization rates.
The early period of the OMNI databases (1963–1971) includes data from mul-
tiple spacecraft (Bonetti et al. 1969; Neugebauer 1970), the middle-period data
are mostly from IMP-8 and span from 1971 to 1994, and the later periods, from
1994 until present, include mostly data from IMP-8, or the WIND Solar Wind
Experiment (SWE) (Kasper 2002), and ACE/SWEPAM (McComas et al. 1998a).
Since there was no overlap between the early and middle period, the data from
the early period in the OMNI-2 database are adopted unchanged from the original
OMNI collection. The data obtained from various spacecraft during the early
period were extensively intercalibrated, but no intercalibration with the middle
and recent periods was possible. Still, owing to the overlap of the data between the
middle and recent periods, it was possible to perform an intercalibration between
the data from these two periods.
It is important to mention the significance of the correlation work that the
OMNI team performed on the data from the IMP-8 spacecraft and early WIND and
ACE measurements. IMP-8 operated for 28 years from 1973 to 2001 and provided
a bridge between the early and present observations, enabling the presentation of
a more or less homogeneous series of solar wind parameters shown further down in
this chapter.
Most of the solar wind plasma data used in the OMNI collection were ob-
tained from the MIT Faraday Cups (Bridge et al. 1965; Lyon et al. 1967, 1968;
Lazarus and Paularena 1998) and LANL electrostatic analyzers (Bame et al. 1971,
1978a,b; Hundhausen et al. 1967; Ogilvie et al. 1968; Feldman et al. 1973; Asbridge
et al. 1976; McComas et al. 1998a). In the OMNI-2 series, King and Papitashvili
Evolution of Solar Wind in the Ecliptic Plane 99
density [cm−3] 15
10
0
2055 2056 2057 2058 2059 2060 2061 2062
Carrington rotation number
600
speed [km s−1]
500
400
300
200
2055 2056 2057 2058 2059 2060 2061 2062
Carrington rotation number
Figure 3.14: Daily averages of solar wind density adjusted to 1 AU (upper panel)
and speed (lower panel) calculated from hourly averages obtained from the OMNI-
2 database (King and Papitashvili 2005) for an example time interval covering
7 full Carrington rotation periods, presented to illustrate the complex structure
of solar wind parameter evolution at different time scales and the approximate
anticorrelation of density and speed. The interval shown covers 7 full Carrington
rotations in 2008 and is placed in the context of long-term solar wind evolution
shown in Fig. 3.8
7. × 10−7
6. × 10−7
5. × 10−7
4. × 10−7
3. × 10−7
1985 1990 1995 2000 2005 2010
time [y]
Figure 3.15: Carrington rotation averages of the rate of charge exchange between
solar wind protons and neutral interstellar H atoms, calculated from monthly av-
erages of solar wind speed and adjusted density using Eq. 3.19. The thin vertical
lines mark the time interval expanded to daily time resolution in Fig. 3.14
102 3. Solar Parameters
the solar wind and the situations when the range of the slow solar wind significantly
differs from ∼ 45◦ .
Despite these deficiencies, Eq. 3.37 was successfully used by a number of authors
(e.g. Lallement et al. 1985a; Pryor et al. 1998, 2003) to qualitatively infer the
solar wind structure. The conclusions were similar to those obtained from the IPS
analysis: during solar minimum the solar wind is latitudinally structured, with a
band of enhanced flux at the equator and two polar caps of a rarefied and fast
wind. During solar maximum the ordered structure changes and the polar caps
become almost fully covered with the slow wind.
However, it became clear that the simple model given by Eq. 3.37 is not fully
adequate to describe reality and a need for more observations became evident.
Bertaux et al. (1995) proposed an experiment to study solar wind anisotropies
using the technique of analysis of the heliospheric Lyman-α backscatter glow, which
was implemented in the French/Finnish project SWAN9 onboard the Solar and
Heliospheric Observatory mission (SOHO).
Already shortly after the launch of SOHO it became evident that the early
conclusions on the evolution of the heliospheric groove, and thus the solar wind,
with the solar activity cycle were confirmed (Bertaux et al. 1996, 1997, 1999; Kyrölä
et al. 1998), but the formula used to describe the latitude profile of the ionization
rate needed modification. Thus in the latter work, Eq. 3.37 was modified to describe
separately the northern and southern hemispheres:
βCX (φ) = β0 Θ (φ) 1 − AN sin2 φ + Θ (−φ) 1 − AS sin2 φ + B (φ) , (3.38)
where Θ is the Heavyside step function, AN , AS are the separate anisotropy param-
eters for the northern and southern hemispheres, and B (φ) is used to narrow the
width of the equatorial band of enhanced solar wind flux. An even more sophisti-
cated approach was proposed by Summanen (1996), who suggested to approximate
the equatorial band in the latitudinal profile of the total ionization rate by:
βCX (φ, t) = β0 1 − A (t) sin2 (c φ) (3.39)
for heliolatitudes −40◦ ≤ φ ≤ 40◦ , where c = 9/4 limits the equatorial band to
±40◦ , and by:
2
t − P/2
βCX (φ, t) = β0 exp − (3.40)
0.2P
outside the ±40◦ equatorial band, where P is the solar cycle length and t, time.
In this formula, there was no north-south asymmetry allowed, but it was possible
to homogeneously reproduce the variations of the anisotropy parameters with the
solar activity cycle.
The north-south anisotropies in the ionization rate and their evolution with
solar activity were evident on one hand, and on the other hand the first fast latitude
scan by Ulysses (Phillips et al. 1995a) suggested that the profile of solar wind
parameters can be approximated by an equatorial plateau (with a “rough surface”
of the gusty slow wind) standing out from a flat “foot” of the fast polar wind.
where ωp was obtained as 2π/main period of the composite Lyman-α flux (cf
Fig. 3.1) and the free parameters φ0 and φ1 are fit separately for the northern
and southern hemispheres. The validity range of this approximation is limited to
the time interval from 1990 to 2002. Based on this model of ionization, it was
possible to infer the evolution of solar wind speed and density as a function of time
and latitude. This topic is covered later in this chapter.
The model by Bzowski et al. (2003) and Bzowski et al. (2008) was able to more
correctly reproduce the latitudinal span of the slow wind region and its evolu-
tion with solar activity, but was inadequate to correctly reproduce the boundaries
themselves. Observations from SWAN showed that especially during the transi-
tion phases of solar activity, the photometric latitudinal profiles of the groove are
complex and variable in time.
To address this problem, it turned out to be necessary to use an approach orig-
inally proposed by Summanen et al. (1993) and to model the ionization rate as
a multi-step function, with different levels in fixed—though arbitrarily selected—
latitudinal bands. Such an approach to the modeling of the ionization rate was
numerically implemented already by Lallement et al. (1985a), but these authors
filled the latitudinal “slots” with values obtained from the analytical models dis-
cussed earlier in this chapter.
In the refined approach, presented by Quémerais et al. (2006) and Lallement
et al. (2010), the ionization rates in the latitudinal bands are free parameters fit
to the maps of heliospheric backscatter glow, without any assumptions on the
relations between the neighboring bands. In contrast to the approach exercised by
Bzowski et al. (2003), Quémerais et al. (2006) and Lallement et al. (2010) used
all the data available, cleaned only by appropriate masking to cut off the known
extraheliospheric “chaff”.
As a result, a time series of the ionization rate profiles at a resolution of ∼ 10◦
in heliolatitude and ∼ 2 days in time was obtained from fitting the model to
the filtered full-sky maps and subsequent scaling to the in-ecliptic ionization rates
obtained from in situ measurements. An illustration of results of this analysis is
shown by Lallement et al. (2010). They support the general picture of the evolution
of the solar wind structure with solar activity, but they point out that the width
of the equatorial band of enhanced ionization rate was wider during the extended
solar minimum in 2005 through 2009 than during the former minimum.
On the other hand, inspection of the heliolatitude vs time map of the ionization
rate in Lallement et al. (2010) reveals some morphological features that are not
supported by Ulysses or IPS observations. These discrepancies were one of the
reasons to take a closer look at the calibration of the heliospheric FUV observations.
80
60
heliolatitude [deg]
40
20
0
−20
−40
−60
−80
1990 1992 1994 1996 1998 2000 2002 2004 2006 2008 2010
time [y]
Figure 3.16: Illustration of the heliolatitude track of Ulysses (red) and the Earth
(green) during the time span of Ulysses mission. The pale blue line is the F10.7
solar radio flux, superimposed to correlate variations in solar activity with Ulysses
heliolatitudes during its more than three orbits in a polar plane almost perpendic-
ular to the inflow direction of neutral interstellar gas. Adapted from Sokól et al.
(2012)
plane, it performed a Jupiter gravity-assist maneuver that cast it away from the
ecliptic plane on an elliptical nearly polar orbit. The orbit has aphelion ∼ 5.5 AU
and perihelion ∼ 1.4 AU and is nearly perpendicular both to the ecliptic plane and
solar equator. It is also nearly perpendicular to the inflow direction of interstellar
gas. The period of the orbit is about 6 years. The heliolatitude track of Ulysses is
shown in Fig. 3.16.
The spacecraft was launched at solar maximum and its radial cruise was com-
pleted just when solar activity was beginning to decrease. The first dive towards
the south solar pole was carried out during an interval of decreasing solar activity,
followed by the first so-called fast latitude scan, when the spacecraft coasted from
the south to the north solar pole, almost covering the full span of heliolatitudes
during just about a year at solar minimum activity conditions.
This fast latitude scan preceded the first observations of the helioglow by SWAN
by about a year. This provided an opportunity to calibrate the model of ionization
rates obtained from the analysis of the heliospheric Lyman-α glow. Ulysses contin-
ued on its polar orbit, performing its first full slow latitude scan during an interval
of increasing solar activity. The second fast latitude scan occurred in a totally
different solar cycle phase, namely during solar maximum. This scan also took
about a year and was performed during an interval of dynamically variable solar
wind structure. In contrast to the first slow scan, the subsequent slow latitude scan
occurred during decreasing solar activity. Afterwards, the spacecraft performed its
last fast latitude scan, again during minimum solar activity.
The geometry and timing of Ulysses trajectory produced a unique dataset of
direct in situ measurements of the solar wind plasma parameters, obtained from
SWOOPS (Bame et al. 1992) and SWICS (Gloeckler et al. 1992) experiments. The
Latitudinal structure and evolution of the solar wind 109
discoveries and findings from the plasma measurements were presented in dozens
of papers (e.g. Phillips et al. 1995a,b; Marsden and Smith 1997, McComas et al.
1998a, McComas et al. 1999, 2000a,b, 2002a,b, 2003, 2006, 2008). An additional
benefit from this unique mission is the use of the Ulysses flight spare plasma instru-
ment with only minor modifications on the ACE mission (McComas et al. 1998a),
which facilitates intercalibrating the Ulysses measurements with the OMNI time
series.
The Ulysses solar wind dataset is unique and invaluable because it is the first
and only direct in situ measurement of the solar wind parameters outside the eclip-
tic plane. The evolution of solar wind speed, adjusted density, and adjusted flux
during the previous and current solar minima and during the previous maximum
are compiled in Fig. 3.17, where the parameter values are averaged over 10-degree
bins in heliolatitude. “Adjustment” throughout this text means scaling to 1 AU
assuming an average dropoff with heliocentric distance as 1/r 2 .
It can be seen that the heliolatitude structure during the two minima is basically
similar, featuring an equatorial enhancement in density with the associated reduc-
tion in velocity (the slow wind region), and that during solar maximum the slow
wind expands to all heliolatitudes (see also Fig. 3.18 and discussion of time scales in
the variation of the solar wind structure). However, the region of slow wind seems
to reach farther in heliolatitude during the last solar minimum than during the min-
imum of 1995, which, interestingly, is much less conspicuous in density. Thus the
findings from the Ulysses in situ measurements and SWAN remote-sensing obser-
vations by Lallement et al. (2010), discussed in the section “Historical Perspective:
Insight from Heliospheric Backscatter Glow”, qualitatively agree in this respect.
A striking feature seen in Fig. 3.17 is a strong reduction in flux with heliolati-
tude, reported by McComas et al. (2008) and Ebert et al. (2009). The reduction is
visible as a continuous trend from the 1995 minimum through the 2002 maximum
until present. Another interesting trend is a variation of ∼ 1 km s−1 deg−1 in the
fast polar solar wind, discovered by McComas et al. (2000b) and expanded upon
by Ebert et al. (2009).
800
speed [km s−1]
700
600
500
400
−80 −60 −40 −20 0 20 40 60 80
heliolatitude [deg]
SW density for Ulysses fast scans
10
8
[cm−3]
6
density
3.0 × 108
2.5 × 108
2.0 × 108
1.5 × 108
1.0 × 108
−80 −60 −40 −20 0 20 40 60 80
heliolatitude [deg]
Figure 3.17: Solar wind speed (upper panel), adjusted density (middle panel), and
adjusted flux (lower panel) as a function of heliolatitude for the first (blue), sec-
ond (red) and third (green) Ulysses fast latitude scans, obtained from SWOOPS
(Bame et al. 1992). The parameters are averaged over 10-degree heliolatitude bins.
Adapted from Sokól et al. (2012)
Latitudinal structure and evolution of the solar wind 111
700
speed [km s-1]
600
500
400
300
-80 -60 -40 -20 0 20 40 60 80
heliolatitude [deg]
700
speed [km s-1]
600
500
400
300
-80 -60 -40 -20 0 20 40 60 80
heliolatitude [deg]
700
speed [km s-1]
600
500
400
300
-80 -60 -40 -20 0 20 40 60 80
heliolatitude [deg]
Figure 3.18: Solar wind speed profiles from Ulysses measurements and IPS obser-
vations for the three fast Ulysses scans. Red: Ulysses, blue: IPS during the year of
beginning of a Ulysses fast scan, gray: IPS during the year of the end of Ulysses
fast scan. Top panel: the first fast scan during solar minimum, middle panel: the
second fast scan during solar maximum, bottom panel: the third fast scan during
minimum. Adapted from Sokól et al. (2012)
112 3. Solar Parameters
relies on the absolute calibration of the OMNI dataset both in speed and density
in the ecliptic plane. Out of ecliptic, the baseline is the absolute calibration of
Ulysses SWOOPS measurements and IPS. The prime source of information on
the heliolatitude structure of the solar wind are IPS observations, interpreted by
tomography modeling, that generally agree quite well with the Ulysses SWOOPS
in situ measurements out of the ecliptic and with the OMNI measurements in the
ecliptic. Up to now, no continuous measurements of solar wind density as a function
of ecliptic latitude have been available.
Tests revealed that an appropriate balance between the latitudinal resolution
of the coverage and the fidelity of the results is obtained at a subdivision of the
heliolatitudes’ range into 10-degree bins. Concerning the time resolution, the most
welcome would be Carrington rotation averages, identical with the resolution of
the photoionization rate and Lyman-α flux. Regrettably, such a high resolution
seems to be hard to achieve because (1) the time coverage in the data from IPS has
gaps that typically occur during ∼ 4 months at the beginning of each year, which
would induce an artificial 1-year periodicity in the data, and (2) the fast latitude
scans by Ulysses were about 12 months long, hence differentiating between time
and latitude effects in its measurements is challenging. Thus a reliable latitude
structure of the solar wind can only be obtained on a time scale of 1 year and this
is the time resolution of the model that Sokól et al. (2012) developed.
Concerning the global mapping of solar wind parameters from the ecliptic, it
has to be pointed out that the accuracy of measurements of solar wind parameters
decreases with latitude because of geometry. The polar values are the most un-
certain (and possibly biased) because the signal in the polar lines of sight is only
partly formed in the polar region of space, which is illustrated in Fig. 3.19.
In the following section, we will present a procedure to retrieve the solar wind
speed evolution in time and heliolatitude, plus two procedures for retrieval of solar
wind density. One, by Sokól et al. (2012), is based on a correlation between the solar
wind speed and density that was established from the three Ulysses fast latitude
scans and must be regarded as an interim solution, to be used until the other one,
based on the SWAN Lyman-α helioglow observations, will be available.
LOS
Earth
Sun
Figure 3.19: Illustration of the geometry of line of sight (LOS) in a remote sensing
observation of the solar polar region. The observer is close to the Earth in the
ecliptic plane and aims its instrument (e.g., a radio telescope antenna or a Lyman-α
photometer) at a target so that the line of sight crosses a cone with a small opening
angle centered at the pole. The signal is collected from the full length of the line
of sight, but the contributions from various parts are different and depend on the
observations technique. In the case of IPS observations, the strongest contribution
to the signal is from the point nearest to the Sun along the line of sight, marked with
P, because the source function of the scintillation signal decreases with the square
of solar distance. In the case of helioglow observations, the maximum of the signal
comes from the so-called Maximum Emissivity Region, which is located within 1.5–
5 AU from the Sun, so it is important to carefully select the solar elongation of the
line of sight. Note that the angular area of the polar region is quite small, so with
an observations program that maps the entire sky only a small region in the maps
indeed includes the signal from polar regions. Adapted from Sokól et al. (2012)
The line of sight integration effect is deconvolved using the Computer Assisted
Tomography (CAT) method developed by the STEL group (Kojima et al. 1998,
2007). The LOS’s are projected on the source surface at 2.5 solar radii (R ), which
is used as a reference surface in time sequence tomography.
The heliolatitude coverage by IPS observations is not uniform and is strongly
correlated with the Sun’s position in the sky, which changes during the year. The
coverage is also constrained by the target distribution in the sky, with relatively
few of them near the solar poles. Additionally, the observations of the south pole
are of lower quality than those of the north pole because of the low elevation of
the Sun during winter in Japan. The original latitude coverage was improved by
adding the new antenna in 1994 and by optimization of the choice of the targets.
In the following, we briefly present the approach and results from Sokól et al.
(2012). For the analysis, they took the solar wind speed from 1990 to 2011, without
114 3. Solar Parameters
year 2010, when the number of observations was not sufficient to retrieve reliable
solar wind speeds. The solar wind speed observations were mapped at the source
surface on a grid of 11×360×180 records per year, which corresponds to a series of
Carrington rotations. The data were organized in heliolatitude from 89.5◦ North
to 89.5◦ South.
A comparison of the tomography-derived solar wind speed with the in situ
measurements by Ulysses performed by Sokól et al. (2012) showed that the accuracy
of the tomographic results depends on the number of IPS observations available
for a given rotation. The intervals with a small number of data points clearly tend
to underestimate the speed. Consequently, the Carrington rotations with the total
number of points less than 30 000 were removed from the data. Small numbers of
available observations typically happen at the beginning and at the end of the year
and at the edges of data gaps. The selection of data by the total number of points
per rotation constrained the coverage mainly to the summer and autumn months,
when all latitudes are well sampled.
The selected subset of data was split into years, and within each yearly subset
into 19 heliolatitude bins, equally spaced from −90◦ to 90◦ . The speeds averaged
over bins and over year for the latitudinal bins yield the yearly latitudinal profiles
of solar wind speed, shown in Figs. 3.20 and 3.21. They cover half of solar cycle 22
and the full solar cycle 23. In the analysis a two-step calculation was used.
First, Carrington rotation averaged values per bin were computed. Next, the
yearly averages were calculated from the monthly ones. It is worth noting that the
solar wind speed profiles for individual Carrington rotations during a given year
typically have very similar shapes to the related yearly profile, which suggests that
the latitude structure is stable during a year and changes only on a time scale
comparable with solar activity variations.
The results confirm that the solar wind speed is bimodal during solar minimum,
slow at latitudes close to the solar equator (and thus the ecliptic plane) and fast at
the poles. The latitude structure evolves with the solar activity cycle and becomes
flatter when the activity is increasing. The structure is approximately homogeneous
in latitude only during a short time interval during the peak of solar maximum,
when the solar wind is slow at all latitudes (see the panel for 2000 in Fig. 3.20).
Shortly after the activity maximum, the bimodal structure reappears and the fast
wind at the poles is observed again, but switchovers from the slow to fast wind at
the poles may still occur during the high activity period, as shown in the panels
for 2001 and 2003 in Figs. 3.20 and 3.21.
During the descending and ascending phases of solar activity there is a wide
band of slow solar wind that straddles the equator and extends to midlatitudes; the
fast wind is restricted to the polar caps and upper midlatitudes. At solar minimum,
the structure is sharp and stable during a few years straddling the turn of solar
cycles, with high speed at the poles and at midlatitudes and a rapid decrease at
the equator.
Thus, apart from short time intervals at solar maximum, the solar wind struc-
ture close the poles is almost flat, with a steady fast speed value typical of wide
polar coronal holes, in perfect agreement with the measurements from Ulysses
(Phillips et al. 1995b; McComas et al. 2000b, 2006).
Latitudinal structure and evolution of the solar wind 115
]
−1
−1
−1
700 700 700
speed [km s
speed [km s
speed [km s
600 600 600
500 500 500
400 400 400
300 300 300
−90 −60 −30 0 30 60 90 −90 −60 −30 0 30 60 90 −90 −60 −30 0 30 60 90
heliolatitude [deg] heliolatitude [deg] heliolatitude [deg]
1993.53 1994.67 1995.52
800 800 800
]
]
−1
−1
−1
700 700 700
speed [km s
speed [km s
speed [km s
600 600 600
500 500 500
400 400 400
300 300 300
−90 −60 −30 0 30 60 90 −90 −60 −30 0 30 60 90 −90 −60 −30 0 30 60 90
heliolatitude [deg] heliolatitude [deg] heliolatitude [deg]
1996.56 1997.59 1998.57
800 800 800
]
]
−1
−1
−1
700 700 700
speed [km s
speed [km s
speed [km s
600 600 600
500 500 500
400 400 400
300 300 300
−90 −60 −30 0 30 60 90 −90 −60 −30 0 30 60 90 −90 −60 −30 0 30 60 90
heliolatitude [deg] heliolatitude [deg] heliolatitude [deg]
1999.53 2000.55 2001.52
800 800 800
]
]
−1
−1
−1
700 700 700
speed [km s
speed [km s
speed [km s
600 600 600
500 500 500
400 400 400
300 300 300
−90 −60 −30 0 30 60 90 −90 −60 −30 0 30 60 90 −90 −60 −30 0 30 60 90
heliolatitude [deg] heliolatitude [deg] heliolatitude [deg]
Figure 3.20: Heliolatitude profiles of yearly averaged solar wind speed for 1990–2001
obtained by Sokól et al. (2012) from the Computer Assisted Tomography analysis
of interplanetary scintillation observations (Tokumaru et al. 2010). Adapted from
Sokól et al. (2012)
To further verify the results obtained from the IPS analysis, we compared them
with the data from the three Ulysses fast latitude scans and the OMNI measure-
ments in the ecliptic plane. The Ulysses velocity profiles used for this comparison
were constructed from subsets of hourly averages available from the NSSDC, split
into identical heliolatitude bins as those used for the IPS data analysis and aver-
aged. They are shown in Fig. 3.18 as red lines. Since the acquisition of the Ulysses
profiles took one year each and the fast scans straddle the turn of the years, we
show the IPS results for the years straddling the fast latitude scans. They are
presented in blue and gray in Fig. 3.18.
The fast scans were performed at the perihelion half of the Ulysses elliptical
orbit, with the perihelion close to the solar equator plane. Hence, the angular
speed of its motion was highest close to the equator and traversing the 10-degree
bin took it less than one solar rotation period. The apparent bin-to-bin fluctuation
is an effect of incomplete Carrington longitude coverage of the bimodal solar wind,
with slow wind interleaved with fast wind streams.
Near the poles, the angular speed was slower and it took more than 1 Carrington
rotation to scan the 10-degree bin. Still, the longitudinal coverage was uneven.
116 3. Solar Parameters
]
−1
−1
−1
700 700 700
speed [km s
speed [km s
speed [km s
600 600 600
500 500 500
400 400 400
300 300 300
−90 −60 −30 0 30 60 90 −90 −60 −30 0 30 60 90 −90 −60 −30 0 30 60 90
heliolatitude [deg] heliolatitude [deg] heliolatitude [deg]
2005.53 2006.64 2007.51
800 800 800
]
]
−1
−1
−1
700 700 700
speed [km s
speed [km s
speed [km s
600 600 600
500 500 500
400 400 400
300 300 300
−90 −60 −30 0 30 60 90 −90 −60 −30 0 30 60 90 −90 −60 −30 0 30 60 90
heliolatitude [deg] heliolatitude [deg] heliolatitude [deg]
2008.55 2009.48 2011.48
800 800 800
]
]
−1
−1
−1
700 700 700
speed [km s
speed [km s
speed [km s
600 600 600
500 500 500
400 400 400
300 300 300
−90 −60 −30 0 30 60 90 −90 −60 −30 0 30 60 90 −90 −60 −30 0 30 60 90
heliolatitude [deg] heliolatitude [deg] heliolatitude [deg]
Figure 3.21: Heliolatitude profiles of yearly averaged solar wind speed for 2002–
2011 obtained by Sokól et al. (2012) from the Computer Assisted Tomography
analysis of interplanetary scintillation observations (Tokumaru et al. 2010). Note
that data for 2010 are missing because of a non-sufficient number of observations
available. Adapted from Sokól et al. (2012)
Thus, during solar maximum, when the gusty slow wind engulfed the whole space,
the “sawtooth effect” expanded into the full latitude span. By contrast, during the
low-activity scans the solar wind speed at high latitude was stable, which resulted
in the lack of the small-scale latitude variations in the Carrington rotation averages
at high latitudes, even though the issue of uneven longitudinal coverage persisted.
The solar wind speed profiles obtained from IPS and Ulysses observations are
very similar, but some systematic differences do exist. On one hand, it seems that
∼ 50 km s−1 is a typical difference between Ulysses and IPS values in the polar
regions, with the northern region usually in better agreement than the southern.
On the other hand, sometimes the agreement is almost perfect.
The difference between the blue and gray lines in the top and middle panels of
Fig. 3.18 is a measure of true variation of the latitudinal profile of solar wind speed
during one year. Ulysses was moving from south to north during the fast latitude
scans, so the south limb of the profile from Ulysses ought to be closer to the south
limb of the blue profile obtained from the IPS analysis, while the north limb of the
Ulysses profile should agree better with the north limb of the gray IPS profile.
Such behavior is observed in the second panel of Fig. 3.18, which corresponds
to the solar maximum in 2001. In our opinion, it is a very interesting observation
because (1) it shows how rapidly the latitude structure of the solar wind varies
during solar maximum, and (2) it confirms both Ulysses and IPS results regarding
solar wind structure.
Latitudinal structure and evolution of the solar wind 117
500
450
400
350
1990 1995 2000 2005 2010
time [y]
Figure 3.22: Yearly averages of solar wind speed from the IPS analysis (blue) and
in-situ measurements collected in the OMNI-2 database (green), compared with
the average solar wind speed measured by Ulysses during its passage within 5◦
from the solar equator plane during the three fast latitude scans (red diamonds)
and three slow scans (red triangles). Adapted from Sokól et al. (2012)
Originally, the interpretation of the speed profiles obtained from Ulysses was
not clear. It was pondered whether the north-hemisphere increase in the solar
wind speed was a long-standing feature of the solar wind or was just due to a time-
variability of the wind at the north pole. Similarly, it was pondered whether the
IPS analysis is able to reliably reproduce the solar wind profiles given the fact that
some of the profiles obtained approximately at the time of the fast scan seemed to
disagree with the in situ data.
The analysis by Sokól et al. (2012) suggests that the yearly-averaged velocity
profiles obtained from the tomography analysis of IPS observations agree with the
in-situ observations from Ulysses even at times when the solar wind is restructuring
rapidly at the peak of solar activity. The IPS data are in a very good agreement
with the OMNI data collected in the ecliptic plane (see Fig. 3.22). Up to 2004, the
agreement is almost perfect, afterwards small differences appear. The agreement is
better than with the in-ecliptic Ulysses measurements from the fast latitude scans.
This, in our opinion, is because the measurements of solar wind parameters in the
ecliptic plane obtained during the fast latitude scans from Ulysses are challenging
to directly compare with the OMNI and IPS measurements. The reason is that
Ulysses was passing through the ecliptic 10-degree latitude bin in a time equal to
about half of a Carrington rotation and thus a reliable longitude averaging of the
solar wind parameters could not be obtained. A detailed discussion of the non-
trivial issue of comparing the OMNI, Ulysses, and IPS solar wind speed close to
the ecliptic plane is provided by Sokól et al. (2012).
The IPS solar wind speed profiles provide a solar wind latitude structure that
can be adopted as an interim solution. They agree well both with the OMNI time
series in the ecliptic and with the Ulysses measurements out of the ecliptic for the
time intervals when they can be directly compared.
118 3. Solar Parameters
speed [km s ]
−1
700 700 700
600 600 600
500 500 500
400 400 400
300 300 300
−90 −60 −30 0 30 60 90 −90 −60 −30 0 30 60 90 −90 −60 −30 0 30 60 90
heliolatitude [deg] heliolatitude [deg] heliolatitude [deg]
1993.53 1994.67 1995.52
800 800 800
speed [km s ]
speed [km s ]
−1
700 700 700
600 600 600
500 500 500
400 400 400
300 300 300
−90 −60 −30 0 30 60 90 −90 −60 −30 0 30 60 90 −90 −60 −30 0 30 60 90
heliolatitude [deg] heliolatitude [deg] heliolatitude [deg]
1996.56 1997.59 1998.57
800 800 800
speed [km s ]
speed [km s ]
−1
Figure 3.23: Latitude profiles of yearly-averaged solar wind speed for 1990–2001
obtained from the interim procedure described in the text. The red lines show the
speed profiles obtained from the Ulysses fast scans. The profile for 2010 is a linear
interpolation between the profiles for 2009 and 2011 (Sokól et al. 2012)
For further analysis, Sokól et al. (2012) smoothed the yearly speed profiles as-
suming that the variation of the solar wind speed at high latitude close to the poles
is linear and the variation outside the polar caps can be approximated by a series
of smoothly-transitioning parabolae. The resulting smoothed yearly heliolatitude
profiles of solar wind speeds are shown in Figs. 3.23 and 3.24. The gap in the ob-
servations in 2010 was filled by linear interpolation between the profiles from 2009
and 2011. The formula used to smooth the profiles along with the numerical values
of their parameters are listed in Sokól et al. (2012). These smoothed profiles will be
used in the remainder of this paper to obtain the solar wind density and flux profiles.
speed km s 1
speed km s 1
700 700 700
600 600 600
500 500 500
400 400 400
300 300 300
90 60 30 0 30 60 90 90 60 30 0 30 60 90 90 60 30 0 30 60 90
heliolatitude deg heliolatitude [deg] heliolatitude deg
2005.53 2006.64 2007.51
speed km s 1
speed km s 1
700 700 700
600 600 600
500 500 500
400 400 400
300 300 300
90 60 30 0 30 60 90 90 60 30 0 30 60 90 90 60 30 0 30 60 90
heliolatitude deg heliolatitude deg heliolatitude deg
speed km s 1
speed km s 1
700 700 700
600 600 600
500 500 500
400 400 400
300 300 300
90 60 30 0 30 60 90 90 60 30 0 30 60 90 90 60 30 0 30 60 90
heliolatitude deg heliolatitude deg heliolatitude deg
2011.48
800
speed km s 1
700
600
500
400
300
90 60 30 0 30 60 90
heliolatitude deg
Figure 3.24: Latitude profiles of yearly-averaged solar wind speed for 2002–2011
obtained from the interim procedure described in the text. The red lines show the
speed profiles obtained from the Ulysses fast scans. The profile for 2010 is a linear
interpolation between the profiles for 2009 and 2011 (Sokól et al. 2012)
8
−3]
6
density [cm
Figure 3.25: Anticorrelation between the solar wind density and speed obtained
from Ulysses fast latitude scans. Blue corresponds to the first scan at solar min-
imum (see Fig. 3.16), red to the second scan (performed during solar maximum
conditions), and green to the third scan excecuted during solar minimum. The
dots represent speeds and densities averaged over the 10-degree heliolatitude bins,
the blue and green lines are the linear correlations specified in Eq. 3.43. The dotted
red line is the density—speed relation proposed for the transition interval close to
the solar maximum of 2002, calculated as the mean of the correlation relations
obtained from the first and third latitude scans. Adapted from Sokól et al. (2012)
The relation between density and speed for the second scan, which occurred
during solar maximum, does not seem to be linear, but in this case the spatial and
temporal effects seem to be convolved (as discussed earlier in this section). There-
fore, Sokól et al. (2012) proposed to use an arithmetic mean of the relations for the
first and third scans: asecond = (afirst + athird) /2 and bsecond = (bfirst + bthird) /2.
This relation is shown in Fig. 3.25 as the red broken line. A comparison of the
density values actually measured during the second latitude scan and calculated
from the correlation formula is shown in Fig. 3.26.
The interval of applicability of the latter formula is from ∼1998 until 2002. The
formula from the first scan is applicable to the interval before 1998 and the formula
from the third scan for the interval after 2002.
Sokól et al. (2012) calculated the interim yearly profiles of solar wind density as
a function of heliolatitude by applying Eq. 3.43 to the speed profiles presented in
the preceding subsection. Since this is an interim and very approximate solution,
which by its nature is not very accurate, the authors did not attempt to further
adjust it at the equator to the corresponding OMNI densities. The results are
shown in Figs. 3.27 and 3.28. The accuracy of the results in the polar regions is in
the range of 20–40 %.
With the density and speed profiles on hand, one can easily calculate the flux:
8
−3]
density [cm
Figure 3.26: Comparison of solar wind density averaged over 10-degree heliolatitude
bins actually measured by Ulysses (red) and calculated from the correlation formula
in Eq. 3.43 (blue) (Sokol et al., Solar Physics, 2012)
dynamic pressure:
1 2
pdyn (φj , ti ) = mp n (φj , ti ) (v (φj , ti )) , (3.45)
2
and charge exchange rate (in the stationary H atom approximation; Eq. 3.14). They
are collectively shown in Fig. 3.29.
To obtain the solar wind parameters (speed, density, flux, dynamic pressure,
charge exchange rate and total ionization rate) at monthly resolution, Sokól et al.
(2012) replaced the equatorial bin directly with the Carrington rotation averaged
series from OMNI, linearly interpolated to halves of Carrington rotations. The
±10◦ bins were replaced with values linearly interpolated between the ±20 deg
bins and the equatorial bin.
The pole values were calculated from the the parabolic interpolation between
the ±70 and ±80◦ bins, because due to the problems discussed earlier in this chapter
direct measurements over the poles are not available. The remaining latitudinal
bins were linearly interpolated in time between the yearly profiles. As a result
of such a treatment, all available information on the equatorial bin of the solar
wind was utilized. Away from the equatorial bin, where such an information is
not available, there is a smooth transition into the latitude region where the low
time-resolution model is used.
Figure 3.29 shows contour maps of solar wind flux (upper), charge exchange
rate (middle), and dynamic pressure (lower) as a function of time (horizontal axis)
and heliolatitude (vertical axis). The magnitude of the quantity being shown is
indicated by color code and isocontours.
The latitudinal profile of a given quantity at a desired moment in time can be
retrieved by taking a vertical strip of the map from a location corresponding to the
122 3. Solar Parameters
density [cm−3]
density [cm−3]
−3
8 8 8
6 6 6
4 4 4
2 2 2
−90 −60 −30 0 30 60 90 −90 −60 −30 0 30 60 90 −90 −60 −30 0 30 60 90
heliolatitude [deg] heliolatitude [deg] heliolatitude [deg]
1993.53 1994.67 1995.52
10 10 10
density [cm ]
density [cm−3]
density [cm−3]
−3
8 8 8
6 6 6
4 4 4
2 2 2
−90 −60 −30 0 30 60 90 −90 −60 −30 0 30 60 90 −90 −60 −30 0 30 60 90
heliolatitude [deg] heliolatitude [deg] heliolatitude [deg]
1996.56 1997.59 1998.57
10 10 10
density [cm ]
density [cm−3]
density [cm−3]
−3
8 8 8
6 6 6
4 4 4
2 2 2
−90 −60 −30 0 30 60 90 −90 −60 −30 0 30 60 90 −90 −60 −30 0 30 60 90
heliolatitude [deg] heliolatitude [deg] heliolatitude [deg]
1999.53 2000.55 2001.52
10 10 10
density [cm−3]
density [cm ]
density [cm ]
−3
−3
8 8 8
6 6 6
4 4 4
2 2 2
−90 −60 −30 0 30 60 90 −90 −60 −30 0 30 60 90 −90 −60 −30 0 30 60 90
heliolatitude [deg] heliolatitude [deg] heliolatitude [deg]
Figure 3.27: Latitude profiles of yearly averaged solar wind density for 1990–2001
obtained from the interim procedure. The red lines show the Ulysses fast scan
profiles (Sokól et al. 2012)
desired time on the horizontal axis. The evolution of a given quantity in time at
a selected heliolatitude can be retrieved taking a horizontal strip for the latitude
indicated at the vertical axis. The magnitude of this quantity is color-coded and
the color code is given in the color bar next to the panels. The isocontours indicate
regions in (time vs heliolatitude) space where the quantity has a specific value.
As discussed by Sokól et al. (2012), the structure of solar wind flux and charge
exchange painted in Fig. 3.29 shows a clear variation with solar activity level. Dur-
ing the epochs of two solar minima presented in the figure, the flux and charge
exchange rate were almost spherically symmetric, as indicated by the light bands
spanning the whole latitude range in 1991 and in 2001.
By contrast, the flux and charge exchange structure during prolonged intervals
of lower activity is quite different. The flux and charge exchange rate feature clear
maxima at the equatorial latitude. One can observe some north-south asymmetry
in the latitudinal range of this equatorial enhancement, as well as a small (< 1 year)
shift in phase between the northern and southern hemispheres.
The absolute magnitude of the flux and charge exchange rate dropped after
the solar maximum of activity in 2001, which can be seen as a general dimming
of the maps in their right-hand halves. This weakening exists at all heliolatitudes.
Latitudinal structure and evolution of the solar wind 123
density [cm−3]
density [cm−3]
density [cm−3]
8 8 8
6 6 6
4 4 4
2 2 2
−90 −60 −30 0 30 60 90 −90 −60 −30 0 30 60 90 −90 −60 −30 0 30 60 90
heliolatitude [deg] heliolatitude [deg] heliolatitude [deg]
2005.53 2006.64 2007.51
10 10 10
density [cm−3]
density [cm−3]
density [cm−3]
8 8 8
6 6 6
4 4 4
2 2 2
−90 −60 −30 0 30 60 90 −90 −60 −30 0 30 60 90 −90 −60 −30 0 30 60 90
heliolatitude [deg] heliolatitude [deg] heliolatitude [deg]
2008.55 2009.48 2010.5
10 10 10
density [cm−3]
density [cm−3]
density [cm−3]
8 8 8
6 6 6
4 4 4
2 2 2
−90 −60 −30 0 30 60 90 −90 −60 −30 0 30 60 90 −90 −60 −30 0 30 60 90
heliolatitude [deg] heliolatitude [deg] heliolatitude [deg]
2011.48
10
density [cm−3]
8
6
4
2
−90 −60 −30 0 30 60 90
heliolatitude [deg]
Figure 3.28: Latitude profiles of yearly averaged solar wind density for 2002–2011
obtained from the interim procedure. The red lines show the Ulysses fast scan
profiles (Sokól et al. 2012)
Interestingly, the behavior of the solar wind ram pressure is different. This quantity
is much less structured in heliolatitude than radiation pressure and charge exchange
are. This is true at all times, practically regardless of the solar activity level. The
solar wind ram pressure also weakened, similarl to the flux and charge exchange
rate. However, the drop in pressure began earlier than the drop in the other
quantities, namely just before the last solar maximum, i.e., about 1998.
Summing up this section, we have the structure of the solar wind speed from
the smoothed IPS profiles (Figs. 3.23 and 3.24). The density structure is obtained
(Figs. 3.27 and 3.28) from the density-speed correlation from Ulysses (Eq. 3.43,
Fig. 3.25). From these, one calculates the solar wind flux, dynamic pressure, and—
from Eq. 3.19—charge exchange rate between solar wind protons and neutral H
atoms (Fig. 3.29).
124 3. Solar Parameters
60 4.
3.75
40
heliolatitude [deg]
3.5
20 3.25
3.
0
2.75
−20 2.5
−40 2.25
2.
−60 1.75
−80 1.5
1.25
1991 1993 1995 1997 1999 2001 2003 2005 2007 2009 2011 1.
time [y]
Results: charge exchange rate [10−7s−1]
8.
80
7.5
60 7.
40 6.5
heliolatitude [deg]
6.
20 5.5
0 5.
4.5
−20
4.
−40 3.5
−60 3.
2.5
−80
2.
1991 1993 1995 1997 1999 2001 2003 2005 2007 2009 2011 1.5
time [y]
Results: dynamic pressure [nPa]
2.
80 1.9
1.8
60
1.7
40
heliolatitude [deg]
1.6
1.5
20
1.4
0 1.3
1.2
−20 1.1
−40 1.
0.9
−60 0.8
−80 0.7
0.6
1991 1993 1995 1997 1999 2001 2003 2005 2007 2009 2011 0.5
time [y]
Figure 3.29: Contour maps of solar wind flux (in 108 cm−2 s−1 ), charge exchange
(in 10−7 s−1 ) and dynamic pressure (in nPa) shown as a function of time (horizontal
axis) and heliolatitude (vertical axis). The magnitudes are color-coded, with the
color code shown in the color bars accompanying the panels (Sokól et al. 2012)
Latitudinal structure and evolution of the solar wind 125
where βtot (φj , ti ) is obtained from the inversion of a SWAN map and the photoion-
ization rate βHph (0, ti ), here assumed to be spherically symmetric, is from one of
the formulae specified in Eqs. 3.23, 3.24, or 3.25.
Since the total ionization rate data from SWAN for the equator does not always
agree with the rate derived from direct in situ measurements, we calculate the
contrasts κβ,SWAN (φj , ti ) of the SWAN-derived ionization rates as a function of
heliolatitude using:
βSWAN (φj , ti )
κβ,SWAN (φj , ti ) = , (3.47)
βSWAN (0, ti )
where βSWAN (φj , ti ) is an ionization rate at φj th heliolatitude and βSWAN (0, ti )
is an ionization rate at the equator. In this way we can build yearly contrasts for
the 19 heliolatitude bins and for the years since 1996. Multiplying the contrasts
κβ,SWAN (φj , ti ) with the monthly averages of the total equatorial ionization rates
(see Fig. 3.7), we obtain latitudinal profiles of the total ionization rate that agree
with the baseline values for the equator.
We can calculate the absolute charge exchange ionization rate for all heliolat-
itudes by subtracting the photoionization rate from the total rate from Eq. 3.46.
Considering the formula for charge exchange rate given in Eq. 3.14, with the solar
wind velocity profile obtained from IPS observations v (φj , ti ) we can now calculate
the total solar wind flux as a function of heliolatitude in the following way:
With the flux and solar wind speed profile on hand, it is straightforward to calculate
the profile of density:
In that way we can obtain a model of the evolution of the solar wind speed and
density as a function of time and heliolatitude. This model can be used to calculate
the evolution of the ionization rate for H atoms traveling at arbitrary speeds and
can be used as an input to global models of the heliosphere to calculate the flux
and dynamic pressure of the solar wind. This calculation will be performed once
126 3. Solar Parameters
the intercalibration of the heliospheric EUV measurements has been completed and
the final inversion of the SWAN photometric observations for the ionization rate
profiles is performed.
It is worthwhile to mention here that the proposed method involves a time-
latency because the inflowing hydrogen gas needs some time to adapt its density
distribution to the changing ionization rate and radiation pressure. This effect was
discussed by Bzowski et al. (2002), who showed that the latency is a function of the
offset angle of the line of sight from the upwind direction. Its magnitude depends
on the velocity and temperature of the gas, but it is almost constant on a level
between 0.5 and 1.1 years for the offset angles between 0◦ and 90◦ . For higher offset
angles, between 90◦ and 140◦ , it increases by a factor of 1.5–2. This suggests that
a time dependent model of the distribution of neutral interstellar hydrogen in the
inner heliosphere should be used for modeling.
Summary
Intercalibration of heliospheric UV and EUV measurements requires a common
basis of heliospheric ionization processes. The ionization processes affect the dis-
tribution of neutral interstellar gas in the heliosphere and thus influence both the
spectrum and intensity distribution of the heliospheric backscatter glow.
In this chapter we presented a review of the solar factors affecting the distri-
bution of neutral interstellar gas in the heliosphere. We discussed the radiation
pressure, solar EUV ionizing radiation, and the solar wind parameters. We re-
viewed the history of measurements of these factors and developed a model of
time and heliolatitude evolution of solar wind speed and density based on data
available from in situ measurements of the solar wind parameters, from remote
sensing interplanetary scintillation observations of the solar wind speed structure,
and from correlation between the solar wind density and speed, inferred from in
situ measurements.
The results of this model are used as input in the global heliospheric models,
discussed by Izmodenov et al. (2012, this volume). The results of the global he-
liospheric modeling can in turn be used to fine tune the absolute calibration of
heliospheric EUV measurements and, in a new iteration, to further refine the solar
wind evolution model.
The refined model will cover at least the two recent solar cycles and will thus
provide a common homogeneous basis for interpretation of the present and past
heliospheric experiments. It will also be used to interpret the observations of En-
ergetic Neutral Atoms by the Interstellar Boundary Explorer (IBEX) (McComas
et al. 2009a,b) and hopefully other past and future heliospheric measurements.
Acknowledgements
M.B. and J.S. are obliged to Marty Snow for sharing his insight into the realm of EUV mea-
surements. The authors acknowledge the use of NASA/GSFC’s Space Physics Data Facility’s
ftp service for Ulysses/SWOOPS and TIMED/SEE data, SOHO/CELIAS/SEM (http://
www.usc.edu/dept/space science/semdatafolder/long/daily avg/), and OMNI2 data col-
lection (ftp://nssdcftp.gsfc.nasa.gov/spacecraft data/omni/). The F10.7 solar radio flux
Bibliography 127
was provided by the NOAA and Pentincton Solar Radio Monitoring Programme operated
jointly by the National Research Council and the Canadian Space Agency (ftp://ftp.ngdc.
noaa.gov/STP/SOLAR DATA/SOLAR RADIO/FLUX/Penticton Adjusted/ and ftp://
ftp.geolab.nrcan.gc.ca/data/solar flux/daily flux values/). The composite Lyman-α flux
and MgIIc/w were obtained from LASP, accessed through the LISIRD Web page at
(http://lasp.colorado.edu/lisird/lya/). The SOLAR2000 Research Grade historical ir-
radiances are provided courtesy of W. Kent Tobiska and SpaceWx.com. These historical
irradiances have been developed with funding from the NASA UARS, TIMED, and SOHO
missions.
M.B. and J.S. were supported by the Polish Ministry for Science and Higher Education
grants NS-1260-11-09 and N-N203-513-038. P.B. acknowledges support from NASA SR&T
Grant NX09AW32G. Contributions from D.M. were supported by NASA’s IBEX Explorer
mission. Support from the International Space Science Institute in Bern, Switzerland, is
gratefully acknowledged.
Bibliography
P.O. Amblard, S. Moussaoui, T. Dudok de Wit, J. Aboudarham, M. Kretzschmar,
J. Lilensten, F. Auchère, The euv sun as the superposition of elementary suns. Astron.
Astrophys. 487, L13–L16 (2008). doi:10.1051/0004-6361:200809588
G. Artzner, J.C. Vial, P. Lemaire, P. Gouttebroze, J. Leibacher, Simultaneous time-resolved
observations of the H L-alpha Mg K 2795 A, and Ca K solar lines. Astrophys. J. Lett.
224, L83–L85 (1978). doi:10.1086/182765
K. Asai, M. Kojima, M. Tokumaru, A. Yokobe, B.V. Jackson, P.L. Hick, P.K. Manoharan,
Heliospheric tomography using interplanetary scintillation observations. iii - correlation
between speed and electron density fluctuations in the solar wind. J. Geophys. Res. 103,
1991–2001 (1998). doi:10.1029/97JA02750
J.R. Asbridge, S.J. Bame, W.C. Feldman, M.D. Montgomery, Helium and hydro-
gen velocity differences in the solar wind. J. Geophys. Res. 81, 2719–2727 (1976).
doi:10.1029/JA08li016p02719
F. Auchère, Effect of the H I Ly α chromospheric flux anisotropy on the total inten-
sity of the resonantly scattered coronal radiation. Astrophys. J. 622, 737–743 (2005).
doi:10.1086/427903
F. Auchère, J.W. Cook, J.S. Newmark, D.R. McMullin, R. von Steiger, M. Witte,
Model of the all-sky He II 30.4 nm solar flux. Adv. Space Res. 35, 388–392 (2005a).
doi:10.1016/j.asr.2005.02.036
F. Auchère, D.R. McMullin, J.W. Cook et al., A model for solar euv flux helium photoion-
ization throughout the 3-dimensional heliosphere, in Proceedings of the Solar Wind 11
/ SOHO 16 “Connecting the Sun and Heliosphere” Conference, Whistler, Canada, June
2005. S.P. ESA-592, ed. by B. Fleck, T.H. Zurbuchen, H. Lacoste (2005b), pp. 327–329
S.J. Bame, E.W. Hones Jr., S.-L. Akasofu, M.D. Montgomery, J.R. Asbridge, Geomagnetic
storm particles in the high-latitude magnetotail. J. Geophys. Res. 76, 7566–7583 (1971).
doi: 10.1029/JA076i031p07566
S.J. Bame, J.R. Asbridge, H.E. Felthauser, J.P. Glore, H.L. Hawk, J. Chavez, ISEE-C solar
wind plasma experiment. IEEE Trans. Geosci. Electron. 16, 160–162 (1978a)
S.J. Bame, J.R. Asbridge, H.E. Felthauser, J.P. Glore, G. Paschmann, P. Hemmerich,
K. Lehmann, H. Rosenbauer, ISEE-1 and ISEE-2 fast plasma experiment and the ISEE-1
solar wind experiment. IEEE Trans. Geosci. Electron. 16, 216–220 (1978b)
S.J. Bame, D.J. McComas, B.L. Barraclough, J.L. Phillips, K.J. Sofaly, J.C. Chavez,
B.E. Goldstein, R.K. Sakurai, The Ulysses solar wind plasma experiment. Astron. Astro-
phys. Supp. 92, 237–265 (1992)
V.B. Baranov, Kinetic and hydrodynamic approaches in space plasma, in The Physics of
the Heliospheric Boundaries, ed. by V.V. Ismodenov, R. Kallenbach, Vol. SR-005 of ISSI
Science Report, pp. 1–26, ESA Publications Division, EXTEC ISBN 1608-280X (2006a)
128 3. Solar Parameters
V.B. Baranov, Early concepts of the heliospheric interface: plasma, in The Physics of the
Heliospheric Boundaries, ed. by V.V. Ismodenov, R. Kallenbach, vol. SR-005 of ISSI
Science Report, pp. 27–44 (2006b)
V.B. Baranov, Y.G. Malama, Model of the solar wind interaction with the local interstellar
medium—numerical solution of self-consistent problem. J. Geophys. Res. 98, 15157–15163
(1993). doi:10.1029/93JA01171
V.B. Baranov, M.G. Lebedev, Y.G. Malama, The influence of the interface between the
heliosphere and local interstellar medium on the penetration of H atoms to the solar
system. Astrophys. J. 375, 347–351 (1991). doi:10.1086/170194
C.F. Barnett, H.T. Hunter, M.I. Kirkpatrick, I. Alvarez, C. Cisneros, R.A. Phaneuf, Atomic
data for fusion, in Volume 1: Collisions of H, H2 , He, and Li Atoms and Ions with Atoms
and Moleucles. vol. ORNL-6086/V1 (Oak Ridge National Laboratories, Oak Ridge, Ten-
nessee, 1990)
J.-L. Bertaux, J.E. Blamont, Evidence for a source of an extraterrestrial hydrogen Lyman
alpha emission. Astron. Astrophys. 11, 200–217 (1971)
J.-L. Bertaux, E. Kyrölä, E. Quémerais, R. Pellinen, R. Lallement, W. Schmidt, M. Berthé,
E. Dimarellis, J.P. Goutail, C. Taulemasse, C. Bernard, G. Leppelmeier, T. Summanen,
H. Hannula, H. Huomo, V. Kehlä, S. Korpela, K. Leppälä, E. Strömmer, J. Torsti, K.
Viherkanto, J.-F. Hochedez, G. Chretiennot, R. Peyroux, T. Holzer, SWAN: a study
of solar wind anisotropies on SOHO with Lyman alpha sky mapping. Sol. Phys. 162,
403–439 (1995). doi:10.1007/BF00733435
J.-L. Bertaux, E. Quémerais, R. Lallement, Observations of a sky Lyman α groove related
to enhanced solar wind mass flux in the neutral sheet. Geophys. Res. Lett. 23, 3675–3678
(1996). doi:10.1029/96GL03475
J.-L. Bertaux, E. Quémerais, R. Lallement, E. Kyrölä, W. Schmidt, T. Summanen,
T. Mäkinen, The first 1.5 years of observation from SWAN Lyman-alpha solar wind map-
per on SOHO, in Fifth SOHO Workshop: The Corona and Solar Wind near Minimum
Activity, Oslo, Norway, June 1997, ed. by A. Wilson. S.P. ESA-404, p. 29 (1997)
J.-L. Bertaux, E. Kyrölä, E. Quémerais, R. Lallement, W. Schmidt, J. Costa, T. Mäkinen,
SWAN observations of the solar wind latitude distribution and its evolution since launch.
Space Sci. Rev. 87, 129–132 (1999). doi:10.1023/A:1005178402842
J.-L. Bertaux, E. Quémerais, R. Lallement, E. Lamassoure, W. Schmidt, E. Kyrölä, Mon-
itoring solar activity on the far side of the sun from sky reflected Lyman α radiation.
Geophys. Res. Lett. 27, 1331–1334 (2000). doi:10.1029/1999GL003722
S. Bleszyński, S. Grzȩdzielski, D. Ruciński, J. Jakimiec, Expected fluxes of about 1 keV neu-
tral H atoms in interplanetary space—comparison with the uv background and possibility
of detection. Planet. Space Sci. 40, 1525–1533 (1992). doi:10.1016/0032-0633(92)90049-T
P. Bochsler, M. Bzowski, L. Didkovsky, H. Kucharek, J.M. Sokól, T.N. Woods, Ionization
rates (preliminary), (2012) in preparation
A. Bonetti, G. Moreno, S. Cantarano, A. Egidi, R. Marconero, F. Palutan, G. Pizella, Solar
wind observations with satellite ESRO HEOS-1 in December 1969. Nuovo Cimento B
Series 46, 307–323 (1969). doi:10.1007/BF02711013
R.M. Bonnet, P. Lemaire, J.C. Vial, G. Artzner, P. Gouttebroze, A. Jouchoux,
A. Vidal-Madjar, J.W. Leibacher, A. Skumanich, The LPSP instrument on OSO 8.
ii—in-flight performance and preliminary results. Astrophys. J. 221, 1032–1053 (1978).
doi:10.1086/156109
J.C. Brandt, R.G. Roosen, R.S. Harrington, Interplanetary gas. xvii. an astrometric deter-
mination of solar wind velocities from orientations of ionic comet tails. Astrophys. J. 177,
277–284 (1972). doi:10.1086/151706
J.C. Brandt, R.S. Harrington, R.G. Roosen, Interplanetary gas. xx. does the radial solar
wind speed increase with latitude. Astrophys. J. 196, 877–878 (1975). doi:10.1086/153478
M. Brasken, E. Kyrölä, Resonance scattering of Lyman alpha from interstellar hydrogen.
Astron. Astrophys. 332, 732–738 (1998)
J.S. Bridge, A. Egidi, A.J. Lazarus, E. Lyon, L. Jacobson, Preliminary results of plasma mea-
surements on IMP-A, in Space Research. V:969–978, ed. by D.G. King-Hele, P. Muller,
G. Righini (North Holland, Amsterdam, 1965)
Bibliography 129
M. Bzowksi, Time dependent radiation pressure and time dependent 2d ionisation rate for
heliospheric modelling, in The Outer Heliosphere: The Next Frontiers, Cospar Colloquia
Series, vol. 11, ed. by K. Scherer, H. Fichtner, H.-J. Fahr, E. Marsch (Pergamon Press,
Amsterdam, 2001a), pp. 69–72
M. Bzowski, A model of charge exchange of interstellar hydrogen on a time-dependent, 2d
solar wind. Space Sci. Rev. 97, 379–383 (2001b). doi:10.1023/A:1011814125384
M. Bzowski, Response of the groove in heliospheric Lyman-α glow to latitude-dependent ion-
ization rate. Astron. Astrophys. 408, 1155–1164 (2003). doi:10.1051/0004-6361:20031023
M. Bzowski, Survival probability and energy modification of hydrogen energetic neutral
atoms on their way from the termination shock to earth orbit. Astron. Astrophys. 488,
1057–1068 (2008). doi:10.1051/0004-6361:200809393
M. Bzowski, D. Ruciński, Solar cycle modulation of the interstellar hydrogen density distri-
bution in the heliosphere. Space Sci. Rev. 72, 467–470 (1995a). doi:10.1007/BF00768821
M. Bzowski, D. Ruciński, Variability of the neutral hydrogen density distribution due
to solar cycle related effects. Adv. Space Res. 16, 131–134 (1995b). doi:10.1016/0273-
1177(95)00325-9
M. Bzowski, D. Ruciński, Neutral solar wind evolution during solar cycle, in Solar Wind
Eight, ed. by D. Winterhalter, J.T. Gosling, S.R. Habbal, W.S. Kurth, M. Neugebauer.
AIP Conference Proceedings, vol. 382 (American Institute of Physics, Woodbury, New
York, 1996), pp. 650–654. doi:10.1063/1.51452
M. Bzowski, S. Tarnopolski, Neutral atom transport from the termination shock to
1 au, in Physics of the Inner Heliosheath, ed. by J. Heerikhuisen, V. Florinski,
G.P. Zank, N.V. Pogorelov. AIP Conference Series, vol. 858, pp. 251–256 (2006).
doi:10.1063/1.2359335
M. Bzowski, H.-J. Fahr, D. Ruciński, H. Scherer, Variation of bulk velocity and temperature
anisotropy of neutral heliospheric hydrogen during the solar cycle. Astron. Astrophys.
326, 396–411 (1997)
M. Bzwoski, T. Summanen, D. Ruciński, E. Kyrölä, Response of interplanetary glow to
global variations of hydrogen ionization rate and solar Lyman-α flux. J. Geophys. Res.
107, ssh2-1 (2002). doi:10.1029/2001JA000141
M. Bzowski, T. Mäkinen, E. Kyrölä, T. Summanen, E. Quémerais, Latitudinal structure
and north-south asymmetry of the solar wind from Lyman-α remote sensing by SWAN.
Astron. Astrophys. 408, 1165–1177 (2003). doi:10.1051/0004-6361:20031022
M. Bzowski, E. Möbius, S. Tarnopolski, V. Izmodenov, G. Gloeckler, Density of neutral
interstellar hydrogen at the termination shock from Ulysses pickup ion observations.
Astron. Astrophys. 491, 7–19 (2008). doi:10.1051/0004-6361:20078810
M. Bzowski, E. Möbius, S. Tarnopolski, V. Izomdenov, G. Gloeckler, Neutral H density at
the termination shock: a consolidation of recent results. Space Sci. Rev. 143, 177–190
(2009)
S. Chabrillat, G. Kockarts, Simple parameterization of the absorption of the solar Lyman-
alpha line. Geophys. Res. Lett. 24, 2659–2662 (1997)
W.A. Coles, S. Maagoe, Solar-wind velocity from IPS observations. J. Geophys. Res. 77,
5622–5624 (1972). doi: 10.1029/JA077i028p05622
W.A. Coles, B.J. Rickett, IPS observations of the solar wind speed out of the ecliptic. J.
Geophys. Res. 81, 4797–4799 (1976)
J.W. Cook, G.E. Brueckner, M.E. van Hoosier, Variability of the solar flux in the far ultra-
violet 1175–2100 Å. J. Geophys. Res. 85, 2257–2268 (1980)
J.W. Cook, R.R. Meier, G.E. Brueckner, M.E. van Hoosier, Latitudinal anisotropy of the
solar far ultraviolet flux—effect on the Lyman alpha sky background. Astron. Astrophys.
97, 394–397 (1981)
A.E. Covington, Micro-wave solar noise observations during the partial eclipse of November
23, 1946. Nature 159, 405–406 (1947). doi:10.1038/159405a0
J.M.A. Danby, J.L. Camm, Statistical dynamics and accretion. Monthly Not. Royal Astron.
Soc. 117, 150 (1957)
130 3. Solar Parameters
G. de Toma, Evolution of coronal holes and implications for high-speed solar wind during the
minimum between cycles 23 and 24. Solar Phys. 274, 195–217 (2011). doi:10.1007/s11207-
010-9677-2
P.A. Dennison, A. Hewish, The solar wind outside the plane of the ecliptic. Nature 213,
343–346 (1967). doi:10.1038/213343a0
T. Dudok de Wit, J. Lilensten, J. Aboudarham, P.-O. Amblard, M. Kretzschmar, Retrieving
the solar euv spectrum from a reduced set of spectral lines. Ann. Geophys. 23, 3055 (2005)
T. Dudok de Wit, M. Kretzschmar, J. Aboudarham, P.-O. Amblard, F. Auchère, J. Lilen-
sten, Which solar euv indices are best for reconstructing the solar euv irradiance? Adv.
Space Res. 42, 903–911 (2008). doi:10.1016/j.asr.2007.04.019
T. Dudok de Wit, M. Kretzschmar, J. Lilensten, T. Woods, Finding the best proxies for the
solar uv irradiance. Geophys. Res. Lett. 36, L10 (2009). doi:10.1029/2009/GL037825
R.W. Ebert, D.J. McComas, H.A. Elliott, R.J. Forsyth, J.T. Gosling, Bulk properites
of the slow and fast solar wind and interplanetary coronal mass ejections measured
by Ulysses: three polar orbits of observations. J. Geophys. Res. 114, A1109 (2009).
doi:10.1029/2008JA013631
C. Emerich, P. Lemaire, J.-C. Vial, W. Curdt, U. Schüle, K. Wilhelm, A new relation be-
tween the central spectral solar H I Lyman α irradiance and the line irradiance measured
by sumer/SOHO during cycle 23. Icarus 178, 429–433 (2005)
H.-J. Fahr, Non-thermal solar wind heating by supra-thermal ions. Solar Phys. 30, 193–206
(1973)
H.-J. Fahr, Change of interstellar gas parameters in stellar wind dominated atmospheres:
solar case. Astron. Astrophys. 66, 103–117 (1978)
H.-J. Fahr, Interstellar hydrogen subject to a net repulsive solar force field. Astron.
Astrophys. 77, 101–109 (1979)
H.-J. Fahr, The 3d heliosphere: three decades of growing knowledge. Adv. Space Res. 32,
3–13 (2004)
H.-J. Fahr, D. Ruciński, Neutral interestellar gas atoms reducing the solar wind number
and fractionally neutralizing the solar wind. Astron. Astrophys. 350, 1071–1078 (1999)
H.-J. Fahr, D. Ruciński, Modification of properties and dynamics of distant solar wind
due to its interaction with neutral interstellar gas. Space Sci. Rev. 97, 407–412 (2001).
doi:10.1023/A:1011874311272
H.-J. Fahr, D. Ruciński, Heliospheric pick-up ions influencing thermodynamics and dynam-
ics of the distant solar wind. Nonlinear Proc. Geophys. 9, 377–386 (2002)
H.-J. Fahr, K. Scherer Perturbations of the solar wind flow by radial and latitudinal pick-up
ion pressure gradients. Ann. Geophys. 22, 2229–2238 (2004)
H.-J. Fahr, H. Fichtner, K. Scherer, Theoretical aspects of energetic neutral atoms as mes-
sengers from distant plasma sites with emphasis on the heliosphere. Rev. Geophys. 45,
RG4003 (2007). doi:10.1029/2006RG000214
W.C. Feldman, J.R. Asbridge, S.J. Bame, M.D. Montgomery, Double ion streams in the
solar wind. J. Geophys. Res. 78, 2017–2027 (1973). doi:10.1029/JA078i013p02017
W.L. Fite, A.C.S. Smith, R.F. Stebbins, Charge transfer in collisions involving symmetric
and asymmetric resonance. Proc. R. Soc. London Ser. A 268, 527 (1962)
L. Floyd, D.K. Prinz, P.C. Crane, L.C. Herring, Solar uv irradiance variation during cycles
22 and 23. Adv. Space Res. 29, 1957–1962 (2002)
L. Floyd, G. Rottman, M. Deland, J. Pap, 11 years of solar uv irradiance measurements from
UARS, in Solar Variability as an Input to the Earth’s Environment, ed. by A. Wilson.
ESA SP-535, pp. 195–203 (2003)
L. Floyd, J. Newmark, J. Cook, L. Herring, D. McMullin, Solar euv and uv spec-
tral irradiances and solar indices. J. Atmos. Sol. Terr. Phys. 67, 3–15 (2005).
doi:10.1016/j.jastp.2004.07.013
P.C. Frisch, M. Bzowski, E. Grün, V. Izmodenov, H. Krüger, J.L. Linsky, D.J. McComas,
E. Möbius, S. Redfield, N. Schwadron, R. Shelton, J.D. Slavin, B.E. Wood, The galactic
environment of the sun: interstellar material inside and outside of the heliosphere. Space
Sci. Rev. 146, 235–273 (2009). doi:10.1007/s11214-009-9502-0
Bibliography 131
P.C. Frisch, S. Redfield, J.D. Slavin, The interstellar medium surrounding the sun. Ann.
Rev. Astron. Astrophys. 49, 237–279 (2011). doi:10.1146/annurev-astro-081710-102613
K. Fujiki, M. Kojima, M. Tokumaru, T. Ohmi, A. Yokobe, K. Hayashi, Solar cycle de-
pendence of high-latitude solar wind, in Solar Wind Ten, ed. by M. Velli, R. Bruno,
F. Malara, B. Bucci. American Institute of Physics Conference Series, vol. 679 (American
Institute of Physics, Woodbury, New York, 2003a), pp. 141–143. doi:10.1063/1.1618561
K. Fujiki, M. Kojima, M. Tokumaru, T. Ohmi, A. Yokobe, K. Hayashi, D.J. McComas,
H.A. Elliott, Solar wind velocity structure around the solar maximum observed by in-
terplanetary scintillation, in Solar Wind Ten, ed. by M. Velli, R. Bruno, F. Malara, B.
Bucci. American Institute of Physics Conference Series, vol 679 (American Institute of
Physics, Woodbury, New York, 2003b) pp. 226–229. doi:10.1063/1.1618583
K. Fujiki, M. Kojima, M. Tokumaru, T. Ohmi, A. Yokobe, K. Hayashi, D.J. McComas,
H.A. Elliott, How did the solar wind structure change around the solar maximum?
from interplanetary scintillation observation. Ann. Geophys. 21, 1257–1261 (2003c).
doi:10.5194/angeo-21-1257-2003
G. Gloeckler, J. Geiss, Heliospheric and interstellar phenomena deduced from pickup ion
observations. Space Sci. Rev. 97, 169–181 (2001)
G. Gloeckler, J. Geiss, H. Balsiger, P. Bedini, J.C. Cain, J. Fisher, L.A. Fisk, A.B. Galvin, F.
Gliem, D.C. Hamilton, The solar wind ion composition spectrometer. Astron. Astrophys.
Supp. 92, 267–289 (1992)
G. Gloeckler, J. Geiss, H. Balsinger, L.A. Fisk, A.B. Galvin, F.M. Ipavich, K.W. Ogilvie,
R. von Steiger, B. Wilken, Detection of interstellar pickup hydrogen in the solar system.
Science 261, 70–73 (1993)
G. Gloeckler, E. Möbius, J. Geiss, M. Bzowksi, S. Chalov, H.-J. Fahr, D.R. McMullin,
H. Noda, M. Oka, D. Ruciński, R. Skoug, T. Terasawa, R. von Steiger, A. Yamazaki, T.
Zurbuchen, Observations of the helium focusing cone with pickup ions. Astron. Astrophys.
426, 845–854 (2004)
K. Gringauz, V. Bezrukih, V. Ozerov, R. Ribchinsky, A study of the interplanetary ionized
gas, high-energy electrons and corpuscular radiation from the sun by means of hte three
electrode trap for charged particles on the second soviet cosmic rocket. Sov. Phys. Doklady
5, 361 (1960)
M.A. Gruntman, Neutral solar wind properties: advance warning of major geomagnetic
storms. J. Geophys. Res. 99, 19213–19227 (1994)
J.K. Harmon, Scintillation studies of density microstructure in the solar wind plasma. Dis-
sertation, University of California, San Diego, 1975
K.L. Harvey, F. Recely, Polar coronal holes during cycles 22 and 23. Solar Phys. 211, 31–52
(2002)
K. Hayashi, M. Kojima, M. Tokumaru, K. Fujiki, MHD tomography using interplanetary
scintillation measurement. J. Geophys. Res. 108, 1102 (2003). doi:10.1029/2002JA009567
D.F. Heath, B.M. Schlesinger, The Mg 280-nm doublet as a monitor of changes in solar ultra-
violet irradiance. J. Geophys. Res. 91, 8672–8682 (1986). doi:10.1029/JD091iD08p08672
A. Hewish, M.D. Symonds, Radio investigation of the solar plasma. Planet. Space Sci. 17,
313 (1969). doi:10.1016/0032-0633(69)90064-6
A. Hewish, P.F. Scott, D. Wills, Interplanetary scintillation of small diameter radio sources.
Nature 203, 1214–1217 (1964). doi:10.1038/2031214a0
H.E. Hinteregger, K. Fukui, B.R. Gilson, Observational, reference and model data on
solar euv, from measurements on AE-E. Geophys. Res. Lett. 8, 1147–1150 (1981).
doi:10.1029/GL008i011p01147
Z. Houminer, Radio source scintillation—evidence of plasma streams corotating about the
sun. Nature 231, 165 (1971)
D. Hovestadt, M. Hilchenbach, A. Bürgi, B. Klecker, P. Laeverenz, M. Scholer,
H. Grünwaldt, W.I. Axford, S. Livi, E. Marsch, B. Wilken, H.P. Winterhoff, F.M. Ipavich,
P. Bedini, M.A. Coplan, A.B. Galvin, G. Gloeckler, P. Bochsler, H. Balsiger, J. Fischer,
J. Geiss, R. Kallenbach, P. Wurz, K.-U. Reiche, F. Gliem, D.L. Judge, H.S. Ogawa, K.C.
Hsieh, E. Möbius, M.A. Lee, G.G. Managadze, M.I. Verigin, M. Neugebauer, CELIAS—
132 3. Solar Parameters
charge, element and isotope analysis system for SOHO. Solar Phys. 162, 441–481 (1995).
doi:10.1007/BF00733436
A.J. Hundhausen, J.R. Asbridge, S.J.B.H.E. Gilbert, I.B. Strong, Vela 3 satellite observa-
tions of solar wind ions. J. Geophys. Res. 72, 1979 (1967). doi:10.1029/JZ072i007p01979
T. Isobe, E.D. Feigelson, M.G. Akritas, G.J. Babu, Linear regression in astronomy. Astro-
phys. J. 364, 104–113 (1990). doi:10.1086/169390
K. Issautier, Diagnostics of the solar wind plasma, in Turbulence in Space Plasmas, ed.
by P. Cargill, L. Vlahos. Lecture Notes in Physics, vol. 778 (Springer, Berlin, 2009), pp.
223–246
K. Issautier, N. Meyer-Vernet, M. Moncuquet, S. Hoang, Solar wind radial and latitudi-
nal structure—electron density and core temperature from Ulysses thermal noise spec-
troscopy. J. Geophys. Res. 103, 1969–1979 (1998)
K. Issautier, R.M. Skoug, J.T. Gosling, S.P. Gary, D.J. McComas, Solar wind plasma pa-
rameters on Ulysses: detailed comparison between the urap and swoops experiments. J.
Geophys. Res. 106, 15665–15676 (2001). doi:10.1029/2000JA000412
V.V. Izmodenov, V.B. Baranov, Modern multi-component models of the heliospheric inter-
face, in The Physics of the Heliospheric Boundaries, ed. by V.V. Izmodenov, R. Kallen-
bach. ISSI Scientific Report Series, SR-005, pp. 67–136 (2006)
V.V. Izmodenov, Y.G. Malama, A.P. Kalinin, M. Gruntman, R. Lallement, I.P. Rodionova,
Hot neutral H in the heliosphere: elastic H-H, H-p collisions. Astrophys. Space Sci. 274,
71–76 (2000). doi:10.1023/A:1026531519864
V.V. Izmodenov, D.B. Alexashov, S.V. Chalov, O.A. Katushkina, Y.G. Malama,
E.A. Provornikova, Kinetic-gasdynamic modeling of the heliospheric interface: global
structure, interstellar atoms and heliospheric enas. Space Sci. Rev. 146, 329–351 (2009).
doi:10.1007/s11214-009-9528-3
V.V. Izmodenov, O.A. Katushkina, E. Quémerais, M. Bzowski, Distribution of interstellar
H atoms in the heliosphere and backscattered solar Lyman-α, in Cross-Calibration of Far
uv Spectra of Solar System Objects and the Heliosphere, ed. by E. Quémerais, M. Snow,
R.M. Bonnet. ISSI Scientific Report Series, SR-013 (2013) (this volume)
B.V. Jackson, P.L. Hick, M. Kojima, A. Yokobe, Heliospheric tomography using interplan-
etary scintillation observations. Adv. Space Res. 20, 23–26 (1997). doi:10.1016/S0273-
1177(97)00474-2
B.V. Jackson, P.L. Hick, M. Kojima, A. Yokobe, Heliospheric tomography using interplan-
etary scintillation observations. i. combined nagoya and cambridge data. J. Geophys. Res.
103, 12049–12067 (1998)
D.L. Judge, D.R. McMullin, H.S. Ogawa, D. Hovestadt, B. Klecker, M. Hilchenbach,
E. Möbius, L.R. Canfield, R.E. Vest, R. Watts, C. Tarrio, M. Kuehne, P. Wurz, First solar
euv irradiances obtained from SOHO by the CELIAS/SEM. Solar Phys. 177, 161–173
(1998)
J.C. Kasper, Solar wind plasma: kinetic properties and micro-instabilities. Dissertation,
Massachusetts Institute of Technology, Cambridge, 2002
J.C. Kasper, A.J. Lazarus, J.T. Steinberg, K.W. Ogilvie, A. Szabo, Physics-based tests to
identify the accuracy of solar wind ion measurements: a case study with the wind faraday
cups. J. Geophys. Res. 111, A03105 (2006). doi:10.1029/2005JA011442
J.C. Kasper, M.L. Stevens, K.E. Korreck, B.A. Maruca, K.K. Kiefer, N.A. Schwadron, S.T.
Lepri, Evolution of the relationships between helium abundance, minor ion charge state,
and solar wind speed over the solar cycle. Astrophys. J. 745, 162 (2012). doi:10.1088/0004-
637X/745/2/162
O.A. Katushkina, V.V. Izmodenov, Effect of the heliospheric interface on the distribution
of interstellar hydrogen atom inside the heliosphere. Astron. Lett. 36, 297–306 (2010).
doi:10.1134/S1063773710040080
J.H. King, N.E. Papitashvili, Solar wind spatial scales in and comparisons of hourly wind
and acd plasma and magnetic field data. J. Geophys. Res. 110, 2104–2111 (2005).
doi:10.1029/2004JA010649
D. Kiselman, T. Pereira, B. Gustafsson, M. Asplund, J. Meléndez, K. Langhans, Is the solar
spectrum latitude dependent? an investigation with SST/TRIPPEL. Astron. Astrophys.
535, A18 (2011)
Bibliography 133
M. Kojima, T. Kakinuma, Solar cycle evolution of solar wind speed structure between 1973
and 1985 observed with the interplanetary scintillation method. J. Geophys. Res. 92,
7269–7279 (1987). doi:10.1029/JA092iA07p07269
M. Kojima, T. Kakinuma, Solar cycle dependence of global distribution of solar wind speed.
Space Sci. Rev. 53, 173–222 (1990). doi:10.1007/BF00212754
M. Kojima, M. Tokumaru, H. Watanabe, A. Yokobe, K. Asai, B.V. Jackson, P.L. Hick,
Heliospheric tomography using interplanetary scintillation observations. 2. latitude and
heliocentric distance dependence of solar wind structure at 0.1-1 au. J. Geophys. Res.
103, 1981–1989 (1998)
M. Kojima, K. Fujiki, T. Ohmi, M. Tokumaru, A. Yokobe, K. Hakamada, The highest solar
wind velocity in a polar region estimated from IPS tomography analysis. Space Sci. Rev.
87, 237–239 (1999). doi:10.1023/A:1005108820106
M. Kojima, K. Fujiki, T. Ohmi, M. Tokumaru, A. Yokobe, K. Hakamada, Latitudinal veloc-
ity structures up to the solar poles estimated from interplanetary scintillation tomography
analysis. J. Geophys. Res. 106, 15677–15686 (2001)
M. Kojima, M. Tokumaru, K. Fujiki, K. Hayashi, B.V. Jackson, IPS tomographic observa-
tions of 3d solar wind structure. Astron. Astrophys. Trans. 26, 467–476 (2007)
M. Kretzschmar, J. Lilensten, J. Aboudarham, Retrieving the solar euv spectral
irradiance from the observation of 6 lines. Adv. Space Res. 37, 341–346 (2006).
doi:10.1016/j.asr.2005.02.029
S. Kumar, A.L. Broadfoot, Evidence from mariner 10 of solar wind flux depletion at high
ecliptic latitudes. Astron. Astrophys. 69, L5-L8 (1978).
S. Kumar, A.L. Broadfoot, Signatures of solar wind latitudinal structure in interplanetary
Lyman-α emissions: mariner 10 observations. Astrophys. J. 228, 302–311 (1979)
E. Kyrölä, T. Summanen, P. Råback, Solar cycle and interplanetary hydrogen. Astron.
Astrophys. 288, 299–314 (1994)
E. Kyrölä, T. Summanen, T. Mäkinen, E. Quémerais, J.-L. Bertaux, R. Lallement, J. Costa,
Preliminary retrieval of solar wind anisotropies / SOHO observations. J. Geophys. Res.
103, 14523–14538 (1998)
R. Lallement, A.I. Stewart, Out-of-ecliptic lyman-alpha observations with Pioneer-Venus:
solar wind anisotropy degree in 1986. Astron. Astrophys. 227, 600–608 (1990)
R. Lallement, J.-L. Bertaux, F. Dalaudier, Interplanetary lyman α spectral profiles and
intensities for both repulsive and attractive solar force fields: predicted absorption pattern
by a hydrogen cell. Astron. Astrophys. 150, 21–32 (1985a)
R. Lallement, J.-L. Bertaux, V.G. Kurt, Solar wind decrease at high heliographic lati-
tudes detected from prognoz interplanetary lyman alpha mapping. J. Geophys. Res. 90,
1413–1420 (1985b)
R. Lallement, T.E. Holzer, R.H. Munro, Solar wind expansion in a polar coronal hole: infer-
ences from coronal white light and interplanetary lyman alpha observations. J. Geophys.
Res. 91, 6751–6759 (1986)
R. Lallement, E. Quémerais, P. Lamy, J.L. Bertaux, S. Ferron, W. Schmidt, The solar
wind as seen by SOHO/SWAN since 1996: comparison with SOHO/LASCO C2 coronal
densities. In Proceedings of SOHO 23 Workshop, ed. by S. Cranmer, T. Hoeksma, J.
Kohl. ASP Conference Series, vol. 428 (2010), pp. 253–258
A.J. Lazarus, K. Paularena, A comparison of solar wind parameters from experiments on
the IMP 8 and Wind spacecraft. In Measurement Techniques in Space Plasmas, ed. by
E. Borovsky, F. Pfaff, T. Young. AGU Geophysical Monograph Series, vol. 102 (1998),
pp. 85–90
G. Le Chat, K. Issautier, N. Meyer-Vernet, I. Zouganelis, M. Moncuquet, S. Hoang, Quasi-
thermal noise spectroscopy: preliminary comparison between kappa and sum of two
Maxwellian distributions, in Twelfth International Solar Wind Conference, vol. 1216,
pp. 316–319 (2010). doi:10.1063/1.3395864
G. Le Chat, K. Issautier, N. Meyer-Vernet, S. Hoang, Large-scale variation of solar wind
electron properties from quasi-thermal noise spectroscopy: Ulysses measurements. Solar
Phys. 271, 141–148 (2011). doi:10.1007/s11207-011-9797-3
134 3. Solar Parameters
J.L. Lean, H.P. Warren, J.T. Mariska, J. Bishop, A new model of solar euv irradiance
variability 2. comparisons with empirical models and observations and implications for
space weather. J. Geophys. Res. 108, 1059 (2003). doi:10.1029/2001JA009238
J.L. Lean, T.N. Woods, F.G. Eparvier, R.R. Meier, D.J. Strickland, J.T. Correira,
J.S. Evans, Solar extreme ultraviolet irradiance: present, past, and future. J. Geophys.
Res. 116, A01102 (2011). doi:10.1029/2010JA015901
M.A. Lee, H.J. Fahr, H. Kucharek, E. Möbius, C. Prested, N.A. Schwadron, P. Wu,
Physical processes in the outer heliosphere. Space Sci. Rev. 146, 275–294 (2009).
doi:10.1007/s11214-009-9522-9
P. Lemaire, J. Charra, A. Jouchoux, A. Vidal-Madjar, G.E. Artzner, J.C. Vial, R.M. Bonnet,
A. Skumanich, Calibrated full disk solar H I Lyman-alpha and Lyman-beta profiles.
Astrophys. J. Lett. 223, L55–L58 (1978). doi:10.1086/182727
P. Lemaire, C. Emerich, W. Curdt, U. Schühle, K. Wilhelm, Solar HI Lyman α full disk
profile obtained with the SUMER/SOHO spectrometer. Astron. Astrophys. 334, 1095–
1098 (1998)
P.L. Lemaire, C. Emerich, J.-C. Vial, W. Curdt, U. Schühle, K. Wilhelm, Variation of the
full sun hydrogen lyman α and β profiles with the activity cycle, in ESSP A-508: From
solar min to max: half a solar cycle with SOHO, 2002, pp. 219–222
P. Lemaire, C. Emerich, J.-C. Vial, W. Curdt, U. Schühle, K. Wilhelm, Variation of the full
sun hydrogen Lyman profiles through solar cycle 23. Adv. Space Res. 35, 384–387 (2005)
P.C. Liewer, B.E. Goldstein, N. Omidi, Hybrid simulations of the effects of interstellar pickup
hydrogen on the solar wind termination shock. J. Geophys. Res. 981, 15211–15220 (1993).
doi:10.1029/93JA01172
B.G. Lindsay, R.F. Stebbings, Charge transfer cross sections for energetic neutral atom data
analysis. J. Geophys. Res. 110, A12213 (2005). doi:10.1029/2005JA011298
W. Lotz, An empirical formula for the electron-impact ionization cross-section. Zeitschrift
f. Phys. 206, 205–211 (1967a)
W. Lotz, Electron-impact ionization cross-sections and ionization rate coefficients for atoms
and ions. Ap. J. Suppl. 14, 207–238 (1967b)
E.F. Lyon, H.S. Bridge, J.H. Binsack, Explorer 35 plasma measurements in the vicinity of
the moon. J. Geophys. Res. 72, 6113–6117 (1967). doi:10.1029/JZ072i023p06113
E.F. Lyon, A. Egidi, G. Pizella, H.S. Bridge, J.S. Binsack, R. Baker, R. Butler, Plasma
measurements on Explorer 33 (I) interplanetary region. Space Research, VIII, 99 (1968)
L.J. Maher, B.A. Tinsley, Atomic hydrogen escape rate due to charge exchange with hot
plasmaspheric ions. J. Geophys. Res. 82, 689–695 (1977)
M. Maksimovic, V. Pierrard, P. Riley, Ulysses distributions fitted with Kappa functions.
Geophys. Res. Lett. 24, 1151–1154 (1997). doi:10.1029/97GL00992
M. Maksimovic, S.P. Gary, R.M. Skoug, Solar wind electron suprathermal strength and
temperature gradients: Ulysses observations. J. Geophys. Res. 105, 18337–18350 (2000)
M. Maksimovic, I. Zouganelis, J.-Y. Chaufray, K. Issautier, E.E. Scime, J.E. Littleton, E.
Marsch, D.J. McComas, C. Salem, R.P. Lin, H. Elliott, Radial evolution of the electron
distribution functions in the fast solar wind between 0.3 and 1.5 AU. J. Geophys. Res.
110, A9104 (2005). doi:10.1029/2005JA011119
Y. Malama, V.V. Izmodenov, S.V. Chalov, Modeling of the heliospheric interface: multi-
component nature of the heliospheric plasma. Astron. Astrophys. 445, 693–701 (2006)
P.K. Manoharan, Three-dimensional structure of the solar wind: Variation of density with
the solar cycle. Sol. Phys. 148, 153–167 (1993). doi:10.1007/BF00675541
R.G. Marsden, E.J. Smith, Ulysses: a summary of the first high-latitude survey. Adv. Space
Res. 19, (6)825–(6)834 (1997)
D.J. McComas, S. J. Bame, P. Barker, W. C. Feldman, J. L. Phillips, P. Riley, J.W. Griffee,
Solar wind electron proton alpha monitor (SWEPAM) for the Advanced Composition
Explorer. Space Sci. Rev. 86, 563–612 (1998a)
D.J. McComas, S.J. Bame, B.L. Barraclough, W.C. Feldman, W.C. Funsten, J.T. Gosling,
P. Riley, R. Skoug, Ulysses’ return to the slow solar wind. Geophys. Res. Lett. 25(1), 1-4
(1998)
Bibliography 135
D.J. McComas, H.O. Funsten, J.T. Gosling, W.R. Pryor, Ulysses measurements of variations
in the solar wind – interstellar hydrogen charge exchange rate. Geophys. Res. Lett. 26,
2701–2704 (1999)
D.J. McComas, B.L. Barraclough, H.O. Funsten, J.T. Gosling, E. Santiago-Muñoz, B.E.
Goldstein, M. Neugebauer, P. Riley, A. Balogh, Solar wind observations over Ulysses first
full polar orbit. J. Geophys. Res. 105, 10419–10433 (2000a)
D.J. McComas, J.T. Gosling, R.M. Skoug, Ulysses observations of the irregularly structured
mid-latitude solar wind during the approach to solar maximum. Geophys. Res. Lett. 27,
2437–2440 (2000b)
D.J. McComas, H.A. Elliot, R. von Steiger, Solar wind from high-latitude coronal holes at
solar maximum. Geophys. Res. Lett. 29, 1314 (2002a). doi:10.1029/2001GL013940
D.J. McComas, H.A. Elliott, J.T. Gosling, D.B. Reisenfeld, R.M. Skoug, B.E. Goldstein, M.
Neugebauer, A. Balogh, Ulysses second fast-latitude scan: Complexity near solar maxi-
mum and the reformation of polar coronal holes. Geophys. Res. Lett. 29, 1290 (2002b).
doi:10.1029/2001GL014164
D.J. McComas, H.A. Elliot, N.A. Schwadron, J.T. Gosling, R.M. Skoug, B.E. Goldstein,
The three-dimensional solar wind around solar maximum, Geophys. Res. Lett. 30, 24–1,
(2003). doi:10.1029/2003GL017136
D.J. McComas, F. Allegrini, L. Bartolone, P. Bochsler, M. Bzowski, M. Collier, H. Fahr, H.
Fichtner, P. Frisch, H. Funsten, S. Fuselier, G. Gloeckler, M. Gruntman, V. Izmodenov, P.
Knappenberger, M. Lee, S. Livi, D. Mitchell, E. Möbius, T. Moore, S. Pope, D. Reisenfeld,
E. Roelof, H. Runge, J. Scherrer, N. Schwadron, R. Tyler, M. Wieser, M. Witte, P. Wurz,
G. Zank, The Interstellar Boundary Explorer (IBEX): Update at the end of phase B, in
Physics of the Inner Heliosheath, ed. by J. Heerikhuisen, V. Florinski, G.P. Zank, N.V.
Pogorelov. American Institute of Physics Conference Series, vol. 858 (American Institute
of Physics, Woodbury, New York, 2006), pp. 241–250
D.J. McComas, R.W. Ebert, H.A. Elliot, B.E. Goldstein, J.T. Gosling, N.A. Schwadron,
R.M. Skoug, Weaker solar wind from the polar coronal holes and the whole sun, Geophys.
Res. Lett. 35, L18103 (2008). doi:10.1029/2008GL034896
D.J. McComas, F. Allegrini, P. Bochsler, M. Bzowski, E.R. Christian, G.B. Crew, R. De-
Majistre, H. Fahr, H. Fichtner, P.C. Frisch, H.O. Funsten, S.A. Fuselier, G. Gloeckler,
M. Gruntman, J. Heerikhuisen, V. Izmodenov, P. Janzen, P. Knappenberger, S. Krimigis,
H. Kucharek, M. Lee, G. Livadiotis, S. Livi, R.J. MacDowall, D. Mitchell, E. Möbius,
T. Moore, N.V. Pogorelov, D. Reisenfeld, E. Roelof, L. Saul, N.A. Schwadron, P.W.
Valek, R. Vanderspek, P. Wurz, G.P. Zank, Global observations of the interstellar inter-
action from the Interstellar Boundary Explorer (IBEX). Science 326, 959–962 (2009a).
doi:10.1126/science.1180906
D.J. McComas, F. Allegrini, P. Bochsler, M. Bzowski, M. Collier, H. Fahr, H. Fichtner,
P. Frisch, H.O. Funsten, S.A. Fuselier, G. Gloeckler, M. Gruntman, V. Izmodenov, P.
Knappenberger, M. Lee, S. Livi, D. Mitchell, E. Möbius, T. Moore, S. Pope, D. Reisen-
feld, E. Roelof, J. Scherrer, N. Schwadron, R. Tyler, M. Wieser, M. Witte, P. Wurz,
G. Zank, IBEX - Interstellar Boundary Explorer, Space Sci. Rev. 146, 11–33 (2009b).
doi:10.1007/s11214-009-9499-4
E. Möbius, D. Hovestadt, B. Klecker, M. Scholer, G. Gloeckler, Direct observation of He+
pick-up ions of interstellar origin in the solar wind. Nature 318, 426–429 (1985)
E. Möbius, B. Klecker, D. Hovestadt, M. Scholer, Interaction of interstellar pick-up ions
with the solar wind. Astrophys. Space Sci. 144, 487–505 (1988)
M. Neugebauer, Initial deceleration of solar wind positive ions in the earth’s bow shock. J.
Geophys. Res. 75, 717–733 (1970)
M. Neugebauer, C.W. Snyder, Solar plasma experiment. Science 138, 1095–1097 (1962).
doi:10.1029/JA075i004p00717
H.S. Ogawa, C.Y.R. Wu, P. Gangopadhyay, D.L. Judge, Solar photoionization as a loss
mechanism of neutral interstellar hydrogen in interplanetary space. J. Geophys. Res.
100, 3455–3462 (1995)
K.W. Ogilvie, L.F. Burlaga, T.D. Wilkerson, Plasma observations on Explorer 34.
J. Geophys. Res. 73, 6809–6824 (1968). doi:10.1029/JA073i021p06809
136 3. Solar Parameters
D. Ruciński, H.-J. Fahr, Nonthermal ions of interstellar origin at different solar wind con-
ditions. Ann. Geophys. 9, 102–110 (1991)
D. Ruciński, M. Bzowski, H.-J. Fahr, Minor helium components co-moving with the solar
wind. Astron. Astrophys. 334, 337–354 (1998)
C. Salem, J.-M. Bosqued, D.E. Larson, A. Mangeney, M. Maksimovic, C. Perche, R.P.
Lin, J.-L. Bougeret, Determination of accurate solar wind electron parameters using
particle detectors and radio wave receivers. J. Geophys. Res. 106, 21701–21717 (2001).
doi:10.1029/2001JA900031
C. Salem, S. Hoang, K. Issautier, M. Maksimovic, C. Perche, WIND-Ulysses in-situ thermal
noise measurements of solar wind electron density and core temperature at solar maximum
and minimum. Adv. Space Res. 32, 491–496 (2003). doi:10.1016/S0273-1177(03)00354-5
H. Scherer, M. Bzowski, H.-J. Fahr, D. Ruciński, Improved analysis of interplanetary HST-
HLyα spectra using time-dependent modelings. Astron. Astrophys. 342, 601–609 (1999)
H. Scherer, H.-J. Fahr, M. Bzowski, D. Ruciński, The influence of fluctuations of the solar
emission line profile on the Doppler shift of interplanetary H Lyα lines observed by the
Hubble-Space-Telescope. Astrophys. Space Sci. 274, 133–141 (2000)
E.E. Scime, S.J. Bame, W.C. Feldman, S.P. Gary, J.L. Phillips, Regulation of the solar wind
electron heat flux from 1 to 5 au JGR 99, 23401–23410 (1994)
G.J. Smith, L.K. Johnson, R.S. Gao, K.A. Smith, R.F. Stebbings, Absolute differential cross
sections for electron capture and loss by kilo-electron-volt hydrogen atoms. Phys. Rev. A
44, 5647–5652 (1991)
J.M. Sokól, M. Bzowski, M. Tokumaru, K. Fujiki, D.J. McComas, Heliolatitude and time
variations of solar wind structure from in-situ measurements and interplanetary scintil-
lation observations. Solar Phys. (2012). doi: 10.1007/s11207-012-9993-9.
Š. Štverák, P.M. Trávnı́ček, M. Maksimovic, E. Marsch, A.N. Fazakerley, E.E. Scime, Elec-
tron temperature anisotropy constraints in the solar wind. J. Geophys. Res. 113, A03103
(2008). doi:10.1029/2007JA012733
Š. Štveràk, M. Maksimovic, P.M. Trávnı́ček, E. Marsch, A.N. Fazakerley, E.E. Scime,
Radial evolution of nonthermal electron populations in the low-latitude solar wind:
Helios, Cluster, and Ulysses observations. J. Geophys. Res. 114, A05104 (2009).
doi:10.1029/2008JA013883
T. Summanen, The effect of the time and latitude-dependent solar ionisation rate on the
measured Lyman-α-intensity. Astron. Astrophys. 314, 663–671 (1996)
T. Summanen, R. Lallement, J.-L. Bertaux, E. Kyrölä, Latitudinal distribution of solar
wind as deduced from Lyman alpha measurements: an improved method. J. Geophys.
Res. 98, 13215–13224 (1993)
K.F. Tapping, Recent solar radio astronomy at centimeter wavelengths - the tem-
poral variability of the 10.7-cm flux. J. Geophys. Res. 92, 829–838 (1987).
doi:10.1029/JD092iD01p00829
S.T. Tarnopolski, Expected distribution of interstellar deuterium in the heliosphere.
Dissertation. Space Research Centre PAS, 2007
S. Tarnopolski, M. Bzowski, Detectability of neutral interstellar deuterium by a forthcoming
SMEX mission IBEX. Astron. Astrophys. 483, L35-L38 (2008a). doi:10.1051/0004-
6361:200809593
S. Tarnopolski, M. Bzowski, Neutral interstellar hydrogen in the inner heliosphere under
the influence of wavelength-dependent solar radiation pressure. Astron. Astrophys. 493,
207–216 (2008b). doi:10.1051/0004-6361:20077058
G.E. Thomas, The interstellar wind and its influence on the interplanetary environment.
Ann. Rev. Earth Planet. Sci. 6, 173–204 (1978)
H. Tian, W. Curdt, E. Marsch, U. Schühle, Hydrogen Lyman-α and Lyman-β spectral radi-
ance profiles in the quiet sun. Astron. Astrophys. 504, 239–248 (2009a). doi:10.1051/0004-
6361/200811445
H. Tian, W. Curdt, L. Teriaca, E. Landi, E. Marsch, Solar transition region above sunspots.
Astron. Astrophys. 505, 307–318 (2009b). doi:10.1051/0004-6361/200912114
H. Tian, L. Teriaca, W. Curdt, J.-C. Vial, Hydrogen Ly α and Ly β radiances and profiles
in polar coronal holes. Astrophys. J. Lett. 703, L152–L156 (2009c). doi:10.1088/0004-
637X/703/2/L152
138 3. Solar Parameters
W.K. Tobiska, T. Woods, F. Eparvier, R. Viereck, L.E. Floyd, D. Bouwer, G. Rottman, O.R.
White, The SOLAR2000 empirical solar irradiance model and forecast tool. J. Atmos.
Sol. Terr. Phys. 62, 1233–1250 (2000)
M. Tokumaru, M. Kojima, K. Fujiki, K. Hayashi, Non-dipolar solar wind structure
observed in the cycle 23/24 minimum. Geophys. Res. Lett. 360, L09101 (2009).
doi:10.1029/2009GL037461
M. Tokumaru, M. Kojima, K. Fujiki, Solar cycle evolution of the solar wind speed distribu-
tion from 1985 to 2008. J. Geophys. Res. 115, A04102 (2010). doi:10.1029/2009JA014628
A.V. Usmanov, W.H. Matthaeus, B.A. Breech, M.L. Goldstein, Solar wind modeling
with turbulence transport and heating. Astrophys. J. 727, 84 (2011). doi:10.1088/0004-
637X/727/2/84
V. Vasyliunas, G. Siscoe, On the flux and the energy spectrum of interstellar ions in the
solar wind. J. Geophys. Res. 81, 1247–1252 (1976)
D.A. Verner, G.J. Ferland, T.K. Korista, D.G. Yakovlev, Atomic data for astrophysics. ii.
new fits for photoionization cross-sections of atoms and ions. Astrophys. J. 465, 487–498
(1996)
I.S. Veselovsky, A.V. Dmitriev, A.V. Suvorova, Algebra and statistics of the solar wind.
Cosmic Res. 48, 113–128 (2010). doi:10.1134/S0010952510020012
A. Vidal-Madjar, Evolution of the solar Lyman alpha flux during four consecutive years.
Solar Phys. 40, 69–86 (1975)
A. Vidal-Madjar, B. Phissamay, The solar L α flux near solar minimum. Solar Phys. 66,
259–271 (1980)
R.A. Viereck, L.C. Puga, The NOAA Mg II core-to-wing solar index: construction of a
20-year time series of chromospheric variability from multiple satellites. J. Geophys. Res.
104, 9995–10006 (1999). doi:10.1029/1998JA900163
M.E. Wachowicz, Global model of distribution of ionization states of heavy ions from solar
plasma in the heliosphere (in Polish). Dissertation, Space Research Centre PAS, 2006
H.P. Warren, NRLEUV 2, A new model of solar euv irradiance variability. Adv. Space Res.
37, 359–365 (2006). doi:10.1016/j.asr.2005.10.028
H.P. Warren, J.T. Mariska, J.L. Lean, A new reference spectrum for the euv irradiance
of the quiet sun 1. emission measure formulation. J. Geophys. Res. 103, 12077–12090
(1998a). doi:10.1029/98JA00810
H.P. Warren, J.T. Mariska, J.L. Lean, A new reference spectrum for the euv irradiance of
the quiet sun 2. comparisons with observations and previous models. J. Geophys. Res.
103, 12091–12102 (1998b). doi:10.1029/98JA00811
H.P. Warren, J.T. Mariska, K. Wilhelm, High-resolution observations of the solar hydrogen
Lyman lines in the quiet sun with the SUMER instrument on SOHO. Astrophys. J. Suppl
119, 105–120 (1998c). doi:10.1086/313151
K.-P. Wenzel, R.G. Marsden, D.E. Page, E.J. Smith, Ulysses: the first high-latitude helio-
spheric mission. Adv. Space Res. 9, 25–29 (1989). doi:10.1016/0273-1177(89) 90089-6
T.N. Woods, G.J. Rottman, O.R. White, J. Fontenla, E.H. Avrett, The solar Ly-alpha line
profile. Astrophys. J. 442, 898–906 (1995). doi:10.1086/175492
T.N. Woods, D.K. Prinz, G.J. Rottman, J. London, P.C. Crane, R.P. Cebula, E. Hilsenrath,
G.E. Brueckner, M.D. Andrews, O.R. White, M.E. VanHoosier, L.E. Floyd, L.C. Her-
ring, B.G. Knapp, C.K. Pankratz, P.A. Reiser, Validation of the UARS solar ultraviolet
irradiances: comparison with the ATLAS 1 and 2 measurements. J. Geophys. Res. 101,
9541–9570 (1996). doi:10.1029/96JD00225
T.N. Woods, W.K. Tobiska, G.J. Rottman, J.R. Worden, Improved solar Lyman irradiance
modeling from 1979 through 1999 based on UARS observations. J. Geophys. Res. 105,
27195–27215 (2000)
T.N. Woods, F.G. Eparvier, S.M. Bailey, P.C. Chamberlin, J. Lean, G.J. Rottman, S.C.
Solomon, W.K. Tobiska, D.L. Woodraska, Solar euv experiment (SEE): mission overview
and first results. J. Geophys. Res. 110, A01312 (2005). doi:10.1029/2004JA010765
F.M. Wu, D.L. Judge, Temperature and velocity of the interplanetary gases along solar
radii. Astrophys. J. 231, 594–605 (1979)
Part II
Interplanetary Hydrogen
—4—
Eric Quémerais∗
LATMOS-IPSL, Université Versailles-Saint Quentin, Guyancourt, France
Bill R. Sandel
Lunar and Planetary Laboratory, University of Arizona, Tucson, AZ, USA
Vladislav V. Izmodenov
Lomonosov Moscow State University, School of Mechanics and Mathematics
Institute for Problems in Mechanics, Russian Academy of Sciences,
Moscow, Russia
Space Research Institute, Russian Academy of Sciences, Moscow, Russia
G. Randall Gladstone
Southwest Research Institute, San Antonio, TX, USA
Abstract
This chapter compares results of models of the interplanetary background, such
as the one presented in Chap. 1, to different datasets obtained in the outer helio-
sphere (Voyager-UVS, Alice New-Horizons) and in the inner heliosphere (SWAN-
SOHO, STIS-HST). The aim of this work is to combine these datasets and the
models and to derive calibration factors that give a coherent picture of the various
instruments and the interplanetary background. These datasets do not overlap and
the models are used to bridge the gaps in distance or in time. In the case of Voy-
ager 1 and 2 UVS instruments, the calibration factors derived here are significantly
different from the values published by Hall (Ultraviolet resonance radiation and
the structure of the heliosphere. Dissertation, University of Arizona, 1992).
Introduction
The aim of this section is to show that the interplanetary (IP) ultraviolet (UV)
background can be used to calibrate UV instruments at a given wavelength, i.e.
H Lyman-α at 121.566 nm. This emission can also be used to follow the temporal
evolution of the sensitivity of the instrument at the same wavelength. Most UV
instruments studying planetary atmospheres, exospheres or cometary atmospheres
are also able to measure the interplanetary background. To give an example of
how this can be done, we will use different datasets obtained at various positions
in the heliosphere and compare them with models. The models are described by
Izmodenov et al. (2013, this volume). Our aim is also to show how accurate the
determination of the absolute calibration can be. There are various uncertainties
that need to be taken into account in this work.
In the following chapter, we will briefly present the models of the interplanetary
background used here. They are described in greater length in Quémerais et al.
(2008). Following that, we will compare the models with the data obtained by the
two UVS instruments aboard Voyager 1 and Voyager 2 (Broadfoot et al. 1977). We
will show that from this model comparison we can derive a new absolute calibration
of these instruments at Lyman-α.
In the following section, we will show that this new calibration is compatible
with the calibration of two other UV instruments: the UV spectrometer ALICE
on the New Horizons mission and the Solar Wind ANisotropies (SWAN) photome-
ter on the Solar Heliospheric Observatory (SOHO). Both instruments have also
observed the interplanetary background but at times and regions of the helio-
sphere that are different from the Voyager observations. The SWAN calibration
has been recently revised and is significantly different from the values published
by Quémerais and Bertaux (2002). This revision is based on a comparison of the
SWAN measurements with Hubble Space Telescope measurements (Clarke et al.
1998; Quémerais et al. 2003).
is small and the single scattering approximation can be used to compute local
emissivities. However, this simplification does not work for calculating intensities.
To model background intensities, it is necessary to perform an integration of the
emissivity along the line of sight. The integration has to be performed until ex-
tinction on the line of sight is significant, which means that the optically thin
approximation is not valid anymore. Beyond five astronomical units (AU) from
the Sun, multiple scattering effects have to be included in the model. Details on
the computations can be found in Quémerais (2000), Quémerais and Izmodenov
(2002), and Quémerais et al. (2008).
Our method combines a standard iterative computation of the first- and second-
order scattering terms using Monte Carlo simulations to compute the higher or-
ders of scattering (i.e., photons that are scattered more than two times). The
second-order scattering term is computed from integration over the whole sky of
the first-order component. It is also computed independently by the Monte Carlo
simulation. This provides a way to check the validity of both methods (Quémerais
2000).
The iterative scheme, i.e. computing order ns from integration over the whole
sky of terms of scattering order ns + 1, could theoretically be applied to any order
(see Sect. 2.3 of Quémerais 2000). However, starting from ns = 3, it is much more
efficient to use a Monte Carlo approach. The general assumptions of our model are
the following:
• The photon source frequency profile is derived from the Solar Ultraviolet Mea-
surements of Emitted Radiation (SUMER) measurements on SOHO (Lemaire
et al. 1998). The source profile is shown in Quémerais (2000).
• The local distribution of hydrogen atoms is represented by the sum of three
components. The first component is the pristine interstellar component,
the second one is created by charge-exchange with protons in the outer he-
liosheath, and the third one is created by charge exchange with protons in
the inner heliosheath. These three components are described in Izmodenov
et al. (2013, this volume).
• The density distribution of the three populations is symmetrical around the
wind axis going through sun center.
• At each point in the heliosphere, the local distribution of each component is
described by six values: a number density, a mean velocity (two components
in the plane containing the wind axis), and three pseudo-temperatures that
are the widths of the velocity distribution projected on the three directions
of the local frame. These three pseudo-temperatures can be different, which
means that the local distribution is not necessarily a Maxwell–Boltzmann
distribution.
• The scattering process is computed following the Angle Dependent Partial
Frequency Redistribution (Mihalas 1970). This means that the frequency of
the outgoing photons depends on the frequency of the incoming photons, the
scattering angle, and the velocity distribution at the point of scattering. The
numerical expressions of the redistribution function is given in Quémerais
(2000).
144 4. Thirty Years of Interplanetary Background Data
observation was analysed by Quémerais et al. (2009) who showed that the feature
was absent from the data for a different line of sight which was less contaminated
by the Earth exosphere.
Figure 4.1 shows the variation of the interplanetary background data as a func-
tion of time for UVS-Voyager 1 and UVS-Voyager 2, respectively. We should note
that before 1993, observations were randomly done. Starting in 1993, observ-
ing campaigns scanning the outer heliosphere were conducted more systematically
(Quémerais et al. 1995). For both datasets, we see a very similar behaviour. The
general trend of the measurements show a decrease of intensity as a function of
solar distance. For a constant intensity in an optically thin medium, we would
expect a decrease as 1/r, however variations of hydrogen number density with ion-
ization effects and multiple scattering effects modify the dependence with solar
distance. The high frequency variations are due to the variations of the direction
of the line of sight, with higher count rates when line of sight is closer to the
inner heliosphere and the Maximum Emissivity Region. The data also show the
variations of the solar illuminating flux that varies with solar activity. There are
local maxima in 1991 and 2001 at times of solar activity maximum. The Voyager
2 UVS measurements stopped in 1998. On the other hand, Voyager 1 UVS is still
active as of July 2012. Starting in 2003, the outer heliosphere scans were stopped.
It appears clearly in Fig. 4.1 where high frequency variations disappear after 2003.
It should also be noted that the Voyager 1 data seem to flatten after 1997. Through
mid-2012, this flattening is stronger that the predictions of our model computation
including heliospheric and time-dependent effects.
Figure 4.2: Determination of the linear relation between the Voyager IPH data
and the model. The top panel shows Voyager 1 data, the bottom panel shows
Voyager 2. The correlation is very good. The slopes with uncertainties are added.
The agreement between the data and the model is much better for Voyager 1 than
Voyager 2. The uncertainties for the Voyager 2 measurements are about twice as
large as for Voyager 1. This plot includes all data obtained between 10 and 50 AU
from the Sun. If we take into account the difference in solar flux values, our result
is in good agreement with the result of Hall (1992)
150 4. Thirty Years of Interplanetary Background Data
Figure 4.3: Comparison of the UVS Voyager data and model computations in
absolute values. Top panel shows Voyager 1 and bottom panel shows Voyager 2.
The abscissae are in AU and show the distance between the spacecraft and the Sun.
Closer to the Sun (within 10 AU) it is necessary to take into account accurately
the variations of the solar parameters
ALICE New Horizons Roll Data 151
the upwind direction is not far from the direction with minimum intensity. The
three New Horizons rolls have a similar geometry. The roll axis was chosen so that
the plane of the roll passed close to upwind and downwind but avoided most of the
stars along the plane of the milky way. The region close to the Sun is the most
difficult to model because it is more sensitive to the solar parameters of ionization
and radiation pressure. The emission from the rest of the sky is little affected by
variations of solar parameters.
The comparison with our model computed for the correct positions of the New
Horizons spacecraft at the times of the observations is shown in Fig. 4.5. The data
from the three rolls (ACO-1 is incomplete) are shown by the green, black and red
curves with high frequency noise. The spikes correspond to stars in the field of
view of the instrument. In the figure, the data have been scaled to a constant solar
flux by dividing by the SORCE SOLSTICE solar Lyman-α flux of the day and
multiplying by a reference flux of 0.9 × 3.32 × 1011 phot cm−2 s−2 Å−1 . The factor
0.9 Å−1 has been discussed before. It corresponds to the change from the total
line flux to the line center flux at Lyman-α. This is the exact same scheme that
we have applied in the previous section to determine the Voyager UVS calibration
factors.
The agreement between the data and the model in Fig. 4.5 is rather good except
in the region closest to the Sun. This is the region that is most affected by solar
parameters and is more difficult to model accurately. The IPH emission from rest
of the sky shows little dependence on solar parameters. The excellent agreement
between the data and the model suggests that the model correctly represents the
interplanetary background and the interplanetary hydrogen distribution from 10
to 20 AU in the region where these data were obtained.
We performed a linear regression between the data and a model where the solar
flux was not corrected by the 0.9 Å−1 factor (i.e. the SOLSTICE flux was used).
Using the three rolls, it was found that the best ratio between total solar Lyman-
α line and the solar line core value was not 0.9 but 0.86. Given the potential
residual bias in the model, we decided to keep the 0.9 factor with an uncertainty
of about 5 %.
In conclusion, the comparison between the ALICE New Horizon roll data and
our model of the interplanetary background in the range 7–17 AU, validates our
model within about 5 % accuracy. Based on the extrapolation of the model to larger
values of solar distance (up to 50 AU), we have linked the calibration factor of the
two UVS instrument to the calibration of the ALICE spectrometer at Lyman-α.
This means that the Voyager UVS calibration is not based on the model alone but
Inner Heliosphere Data 153
also on the comparison with the ALICE New Horizons roll datasets. This is an
important step in the creating a global view of the interplanetary background from
the inner to the outer heliosphere.
SWAN Calibration
The SWAN instrument is dedicated to the study of the solar wind flux aniso-
tropies (Bertaux et al. 1995). The SOHO mission is a cooperation between ESA
154 4. Thirty Years of Interplanetary Background Data
Figure 4.5: Plot of the ALICE IPH measurements for the three rolls. The data are
shown by the green curve (ACO-1), black curve (ACO-2), and red curve (ACO-4).
The spikes are due to stars crossing the line of sight. The models (see text) are
shown by the blue, green and yellow curves and are plotted over the data. The
data-model agreement is excellent except close to the direction of the Sun
and NASA and was launched in December 1995. The SWAN instrument started its
operation in early 1996 and has been operating since almost continuously, except
during the summer of 1998. It is a UV photometer with a bandwidth from 110
to 190 nm. It consists of two sensor units with identical designs, but sitting on
opposite sides of the SOHO spacecraft. The combination of the data from both
sensors allows to make daily observations of the whole sky.
An initial estimate of the SWAN calibration was published by Quémerais and
Bertaux (2002). The two sensor units of SWAN are compared on a regular basis
because they can both view the same area of the sky at the same time. Therefore,
It is necessary to calibrate only one sensor and the calibration of the second sensor
is derived from that knowledge. As shown by Quémerais and Bertaux (2002),
the two sensors have shown very distinct temporal variations in sensitivity, even
though they have been built following the same design. The sensor on the +Z side
of SOHO, facing the north ecliptic sky most of the time, has been fairly stable over
its 15 years of activity. On the other hand, the −Z sensor (facing south ecliptic
most of the time) has been degrading since the early days of the mission and has
lost more than half of its responsivity since 1995. Therefore, very early in the
mission, it was decided to use the more stable unit as a reference and to use the
almost daily sensor cross-calibration observations to scale the data to the values
of that sensor unit. In the spring of 1996, GHRS on the Hubble Space Telescope
Inner Heliosphere Data 155
measured the IPH background close to the downwind direction (Clarke et al. 1998).
This observation was made at the same time as a regular SWAN observation and
served as a reference point for the uncalibrated SWAN measurements. However,
for unclear reasons, the value used by Quémerais and Bertaux (2002) was not the
final GHRS result and this led to an inexact estimate of the absolute calibration of
SWAN. It should be noted also that because the downwind direction has a lower
brightness by nearly a factor of three compared to the upwind direction as seen
from 1 AU, the derived calibration factor had a large uncertainty.
Following the replacement of GHRS by STIS on HST, new observations were
made (Quémerais et al. (2009), private communication). A comparison between the
four HST (one by GHRS and three by STIS) measurements and the corresponding
SWAN values is shown in Table 4.2. As shown in the table, the SWAN calibration
factor has been corrected to match the HST measurements. The current SWAN
calibration is equal to 4.1 Rayleigh per count per second for the +Z sensor unit.
Figure 4.6 shows the IPH intensity recorded by SOHO SWAN between 1996 and
2011 in the direction of the North ecliptic pole as a function of time. The values are
corrected for solar flux variations based on the SORCE SOLSTICE measurements.
The values are scaled to the reference solar flux of 0.9 × 3.32 × 1011 phot cm−2
s−2 Å−1 as was the case for the ALICE Lyman-α line center values described in
the previous section. The data show two main modulations, one with a period of 1
year corresponding to the rotation of SOHO around the Sun and one with a period
of about 10–11 years corresponding to the solar activity cycle. The 1-year period is
easy to understand. The distribution of hydrogen is not symmetric around the Sun.
In the upwind direction, hydrogen atoms get closer to the Sun and the downwind
region is almost void of hydrogen atoms. During the orbit of the spacecraft around
the Sun it crosses the denser upwind region and then downwind cavity which is
the almost void of hydrogen. The contrast in density is the same in emissivity
and therefore also appears in the intensity. The 1-year modulation disappears if
the hydrogen cavity surrounding the Sun gets much larger than the radius of the
spacecraft orbit.
The 11-year period is caused by the change in solar activity and distribution
between fast and slow solar wind. As mentioned by Quémerais et al. (2006), during
solar minimum, ionization rates at the solar poles are much lower than at the solar
equator. This is because the fast solar wind at the solar poles has a lower mass
flux than the slow solar wind at the equator (Quémerais et al. 2006). But this
difference disappears during solar maximum when the fast and slow solar wind
can be found at different latitudes. This means that the mean ionization rate is an
average between different conditions. As appears in Fig. 4.6, after correction for the
solar illumination, the scaled intensity at the poles is minimum at solar maximum
because this is when the ionizing fluxes are stronger. The ionizing fluxes are also
more or less isotropic. In Fig. 4.6, we have added the isotropic ionization models
computed for two solar conditions detailed in Quémerais et al. (2008). Because the
models are isotropic and independent of time, they give the same intensities each
year. The only variation is due to the position of the observer with respect to the
interstellar wind axis. The polar values at solar minimum (1996 and 2007) are not
well reproduced by the model shown in the figure. To reproduce the polar values,
we would need to use a full three-dimensional model.
Figure 4.7 shows the same curves as Fig. 4.6 but concentrates on the period
around the previous solar maximum in 2001. First, it is interesting to note that the
minimum of intensity, i.e. the lowest column density of hydrogen atoms is obtained
in 2003, almost 2 years after the solar maximum. This is the time necessary for
the changes in ionization rates to modify the hydrogen distribution. Second, the
two models are isotropic models with different solar wind fluxes. The parameters
for the orange model are given in Fig. 1 of Quémerais et al. (2008). The values
correspond to solar minimum conditions in-ecliptic (Quémerais et al. 1996). For the
blue curve, the parameter value correspond to in-ecliptic solar maximum conditions
(i.e. values for 2001 in Fig. 4.1 of the same paper). The two models are not
very different because they are isotropic models with in-ecliptic conditions. Once
anisotropy of the ionizing flux is included, models for solar minimum and solar
maximum conditions become very different at the poles. However, it is interesting
to note that the SWAN data for the year 2003 fall between the two models. The two
Conclusion 157
Figure 4.6: Plot of the IPH intensity recorded by SOHO SWAN between 1996 and
2011 in the direction of the North ecliptic pole as a function of time (diamonds).
The values are corrected for solar flux variations based on the SORCE SOLSTICE
measurements. The data show two main modulations, one with a period of 1 year
corresponding to the rotation of SOHO around the Sun and one with a period of
about 10–11 years corresponding to the solar activity cycle. The bottom curves
(orange and blue) show models for uniform ionization rates for different solar wind
mass flux values (see text)
models give the range of variation for an isotropic model given the uncertainty on
the solar ionizing fluxes, therefore we can conclude that the SWAN data obtained
for solar maximum conditions are in good agreement with the isotropic models
presented here. The blue model is the same one that has been used to compare to
the ALICE-NH rolls and to calibrate the two Voyager UVS.
In conclusion, based on our model and data comparison for solar maximum
conditions and for an observer at 1 AU, the SWAN calibration gives values that
agree with the computed values at 1 AU. Therefore we can conclude that the SWAN
and ALICE calibration are in good agreement assuming that the extrapolation
based on our model is correct.
Conclusion
In this chapter, we have compared four different sets of observations of the In-
terplanetary background covering the inner heliosphere to the outer heliosphere. In
one case, we were able to compare data that were obtained from similar positions
158 4. Thirty Years of Interplanetary Background Data
Figure 4.7: The figure is a close-up of the previous figure for the period 2000–2005,
around the solar maximum of 2001. As shown by Quémerais et al. (2006) the
total solar ionization rate is close to uniform for all latitudes at solar maximum.
The orange model corresponds to medium isotropic ionization and the blue model
corresponds to strong isotropic ionization. The data show that the minimum of
IPH intensity is found almost 2 years after the solar maximum of 2001. This delay
is caused by the latency in the reaction of the ionization processes
(close to 1 AU) and at the same time. In that case the comparison is straight-
forward. The result of this comparison is that the SWAN calibration is directly
derived from the calibration of the STIS and GHRS instruments from the Hubble
Space Telescope.
However, it is not always possible to observe at the same time and position.
In that case, we can use a model combining accurate distributions of hydrogen
atoms in the heliosphere with radiative transfer computations to correct for the
differences in location of the observer. We have also shown that close to the Sun, it
is necessary to use three-dimensional models except for solar maximum conditions
when the ionizing fluxes from the Sun become more or less isotropic. Based on
comparisons with models described in Izmodenov et al. (2013, this volume), we have
cross calibrated ALICE and UVS Voyager 1 and 2. We have also cross calibrated
SOHO SWAN with both instruments.
The calibration factors derived here for Voyager 1 and 2 UVS are different from
the ones published by Hall (1992). However, for Voyager 2, the difference is simply
explained by the change in values of the solar illuminating flux. In 1992, the mean
solar Lyman-α flux was assumed to be smaller than current estimates based on
SORCE SOLSTICE values (Woods et al. 2009).
Bibliography 159
In the models used here, the solar illuminating flux is derived from SORCE
SOLSTICE measurements. However, we use a correction factor of 0.9 Å−1 to com-
pute the line center flux. This was derived from measurements made by SOHO
SUMER (Emerich et al. 2005). This correction is assumed to be independent of
solar activity. This assumption is likely to be inexact and future works will have
to include an updated correction factor.
Future works will extend this analysis. In the last few years, we have made
many mutual observations between various UV instruments measuring the inter-
planetary background. For instance, various cross-calibrations have been performed
between SWAN and the Spectroscopy for Investigation of Characteristics of the At-
mosphere of Venus SPICAV instrument on the Venus-Express spacecraft (Bertaux
et al. 2007). Similarly, we have performed cross-calibrations measurements be-
tween the Mercury Atmospheric and Surface Composition Spectrometer (MASCS)
on the MErcury Surface, Space ENvironment, GEochemistry, and Ranging (MES-
SENGER) spacecraft (McClintock and Lankton 2007) and SPICAV. This will allow
us to link these two instruments to the others that we presented here. And finally,
we will link these results with the results shown in Snow et al. (2013, this volume)
by comparing the Spectroscopy for Investigation of the Characteristics of the At-
mosphere of Mars (SPICAM) and SPICAV observations of the same stars.
Bibliography
V.B. Baranov, Gasdynamics of the solar wind interaction with the interstellar medium.
Space Sci. Rev. 52, 89–120 (1990)
V.B. Baranov, Y.G. Malama, Model of the solar wind interaction with the local interstellar
medium—numerical solution of self-consistent problem. J. Geophys. Res. 98, 15157 (1993)
L. Ben-Jaffel, O. Puyoo, R. Ratkiewicz, Far -ultraviolet echoes from the frontier between
the solar wind and the local interstellar cloud. Astrophys. J. 533, 924–930 (2000)
J-L. Bertaux, J.E. Blamont, Evidence for an extra-terrestrial Lyman-alpha emission, the
interstellar wind. Astron. Astrophys. 11, 200 (1971)
J-L. Bertaux, R. Lallement, V.G. Kurt, E.N. Mironova, Characteristics of the local in-
terstellar hydrogen determined from PROGNOZ 5 and 6 interplanetary Lyman-alpha
line profile measurements with a hydrogen absorption cell. Astron. Astrophys. 150, 1–20
(1985)
J-L. Bertaux et al., SWAN: A study of solar wind anisotropies on SOHO with Lyman alpha
sky mapping. Solar Phys. 162, 403–439 (1995)
J-L. Bertaux et al., SPICAV on Venus Express: three spectrometers to study the global
structure and composition of the Venus atmosphere. Planet. Space Sci. 55, 1673–1700
(2007)
P.W. Blum, H-J. Fahr, The distribution of interplanetary hydrogen. Astrophys. Lett. 6, 127
(1970)
J.C. Brandt, J.W. Chamberlain, Interplanetary gas. I. Hydrogen radiation in the night sky.
Astrophys. J. 130, 670 (1959)
M. Brasken, E. Kyrölä, Resonance scattering of Lyman alpha from interstellar hydrogen.
Astron. Astrophys. 332, 732–738 (1998)
A.L. Broadfoot, B.R. Sandel et al., Ultraviolet spectrometer experiment for the Voyager
mission. Space Sci. Rev. 21, 183 (1977)
A.L. Broadfoot, S.K. Atreya, J-L. Bertaux, J.E. Blamont, A.J. Dessler, T.M. Donahue, W.T.
Forrester, D.T. Hall, F. Herbert, J.B. Holberg, Ultraviolet spectrometer observations of
Neptune and Triton. Science 246, 1459–1466 (1989)
160 4. Thirty Years of Interplanetary Background Data
M. Bzowski et al., in Cross-Calibration of Far UV Spectra of Solar System Objects and the
Heliosphere, ed. by E. Quémerais, M. Snow, R.M. Bonnet. ISSI Scientific Report Series,
SR-013 (2013)
J.T. Clarke, R. Lallement, J-L. Bertaux, E. Quémerais, HST/GHRS observations of the
interplanetary medium downwind and in the inner solar system. Astrophys. J. 448, 893
(1995)
J.T. Clarke, R. Lallement, J-L. Bertaux, H-J. Fahr, E. Quémerais, H. Scherer, HST/GHRS
observations of the velocity structure of interplanetary hydrogen. Astrophys. J. 499, 482
(1998)
G. de Toma, E. Quémerais, B.R. Sandel, Long-term variation of the interplanetary H Ly
alpha glow: Voyager measurements and implications for the solar H Ly alpha irradiance.
Astrophys. J. 491, 980 (1997)
C. Emerich, P. Lemaire, J-C. Vial, W. Curdt, U. Schühle, K. Wilhelm, A new relation be-
tween the central spectral solar H I Lyman α irradiance and the line irradiance measured
by SUMER/SOHO during the cycle 23. Icarus 178, 429–433 (2005)
P. Gangopadhyay, V.V. Izmodenov, D.E. Shemansky, M.A. Gruntman, D.L. Judge, Reap-
praisal of the Pioneer 10 and Voyager 2 Lyα intensity measurements. Astrophys. J. 628,
514–519 (2005)
G.R. Gladstone et al., New Horizons cruise observations of Lyα emissions from the inter-
planetary medium. In: Cross-Calibration of Far UV Spectra of Solar System Objects and
the Heliosphere, ed. by E. Quémerais, M. Snow, R.M. Bonnet. ISSI Scientific Report
Series, SR-013 (2013)
G. Gloeckler, E. Möbius, J. Geiss, M. Bzowski, S. Chalov, H. Fahr, D.R. McMullin, H.
Noda, M. Oka, D. Ruciński, Observations of the helium focusing cone with pickup ions.
Astron. Astrophys. 426, 845–854 (2004)
D.T. Hall, Ultraviolet resonance radiation and the structure of the heliosphere. Dissertation,
University of Arizona, 1992
D.T. Hall, D.E. Shemansky, D.L. Judge, P. Gangopadhyay, M.A. Gruntman, Heliospheric
hydrogen beyond 15 AU - evidence for a termination shock. J. Geophys. Res. 98, 15185–
15192 (1993)
V.V. Izmodenov, M.A. Gruntman, Y.G. Malama, Interstellar hydrogen atom distribution
function in the outer heliosphere. J. Geophys. Res. 106, 10681–10690 (2001)
V.V. Izmodenov, Y.G. Malama, M.S. Ruderman, Solar cycle influence on the interaction of
the solar wind with local interstellar cloud. Astron. Astrophys. 429, 1069–1080 (2005)
V.V. Izmodenov, Y.G. Malama, M.S. Ruderman, Modeling of the outer heliosphere with
the realistic solar cycle. Adv. Space Res. 41, 318–324 (2008)
V.V. Izmodenov et al., in Cross-Calibration of Far UV Spectra of Solar System Objects
and the Heliosphere, ed. by E. Quémerais, M. Snow, R.M. Bonnet. ISSI Scientific Report
Series, SR-013 (2013)
H.U. Keller, K. Richter, G.E. Thomas, Multiple scattering of solar resonance radiation in
the nearby interstellar medium. ii. Astron. Astrophys. 102, 415–423 (1981)
R. Lallement, J.C. Reymond, J. Vallerga, M. Lemoine, F. Delaudier, J.-L. Bertaux, Model-
ing the interstellar-interplanetary helium 58.4 nm resonance glow: towards a reconcilia-
tion with particle measurements. Astron. Astrophys. 426, 875–884 (2004)
R. Lallement, E. Quémerais, J-L. Bertaux, S. Ferron, D. Koutroumpa, R. Pellinen, Deflec-
tion of the interstellar neutral hydrogen flow across the heliospheric interface. Science
307, 1447–1449 (2005)
R. Lallement, E. Quémerais, J-L. Bertaux, B.R. Sandel, V. Izmodenov, Voyager measure-
ments of hydrogen Lyman-α diffuse emission from the milky way. Science 334, 1665
(2011)
P. Lemaire, C. Emerich, W. Curdt, U. Schuehle, K. Wilhelm, Solar HI Lyman alpha full
disk profile obtained with the SUMER/SOHO spectrometer. Astron. Astrophys. 334,
1095–1098 (1998)
W.E. McClintock, M.R. Lankton, The mercury atmospheric and surface composition spec-
trometer for the MESSENGER mission. Space Sci. Rev. 131, 481–521 (2007)
D. Mihalas, Stellar Atmospheres (Freeman, San Francisco, 1970)
Bibliography 161
M. Witte, Kinetic parameters of interstellar neutral helium. review of results obtained dur-
ing one solar cycle with the Ulysses/GAS-instrument. Astron. Astrophys. 426, 835–844
(2004)
T.N. Woods, W.K. Tobiska, G.J. Rottman, J.R. Worden, Improved solar Lyman α irradiance
modeling from 1947 through 1999 based on UARS observations. J. Geophys. Res. 105,
27195–27216 (2000)
T.N. Woods et al., Solar irradiance reference spectra (SIRS) for the 2008 whole heliosphere
interval (WHI). Geophys. Res. Lett. 36, L01101 (2009). doi:10.1029/2008GL036373
—5—
G. Randall Gladstone
Southwest Research Institute, San Antonio, TX, USA
S. Alan Stern
Southwest Research Institute, Boulder, CO, USA
Abstract
From models of the interplanetary Lyman-α glow derived from observations by
the Mercury Atmospheric and Surface Composition Spectrometer (MASCS) inter-
planetary Lyman-α data obtained in 2009–2011 on the MErcury Surface, Space
ENvironment, GEochemistry, and Ranging (MESSENGER) spacecraft mission,
daily all-sky Lyman-α maps were generated for use by the Lunar Reconnaissance
Introduction
The Lunar Reconnaissance Orbiter (LRO) LAMP Lyman-Alpha Mapping
Project (LAMP, Gladstone et al. 2010) experiment is designed to view lunar per-
manently shadowed regions (PSRs) that may contain ice and other volatiles using
the surface reflectance of faint ultraviolet (UV) light from interplanetary hydrogen
emissions and stars (Gladstone et al. 2010). We developed a time-dependent model
for the hydrogen Lyman-α illumination of the Moon that was merged with a stel-
lar illumination model for use in determining far-UV lunar albedos in LAMP data.
Preliminary LAMP results indicate that PSRs are spectrally distinct from neigh-
boring regions, perhaps due to ice deposits and porosity enhancement (Gladstone
et al. 2011). This paper describes the heliospheric Lyman-α model and how it
was tuned with MErcury Surface, Space ENvironment, GEochemistry, and Rang-
ing (MESSENGER) Mercury Atmospheric and Surface Composition Spectrometer
(MASCS) data (McClintock and Lankton 2007), then tested and linearly scaled
in brightness by comparisons with Solar and Heliospheric Observatory (SOHO)
SWAN (Solar Wind ANisotropy) instrument data (Bertaux et al. 1995).
Lyman-α Models
Hydrogen Lyman-α emissions in the inner solar system result from resonance
scattering of solar Lyman-α photons by slow neutral hydrogen. Hydrogen in the
inner heliosphere is dominated by interstellar wind hydrogen: neutral gas that
has penetrated the solar wind termination shock from the “upwind” direction. Hot
models consider processes affecting this hydrogen near the Sun but do not explicitly
consider outer heliospheric processes (Thomas 1978). We apply a hot model which
was previously used to model Cassini Ultraviolet Imaging Spectrograph (UVIS)
cruise Lyman-α data to describe the hydrogen distribution in the inner heliosphere
(Pryor et al. 2008). The following neutral hydrogen parameters are assumed for
large distances from the Sun (but inside the termination shock) in the model:
density n = 0.12 cm−3 to match Cassini UVIS brightness levels obtained with its lab
calibration soon after launch, effective velocity v = 20 km s−1 , and temperature T =
12000 K on the basis of SWAN hydrogen cell studies (Costa et al. 1999). For this
study the downwind flow direction is assumed to be toward 74.7◦ ecliptic longitude
Lyman-α Models 165
and −5.2◦ ecliptic latitude based on Ulysses Interstellar Gas Experiment (GAS)
measurements of neutral helium (Witte 2004). We have neglected the small offset
between hydrogen and helium flow directions (Lallement et al. 2005). The primary
loss processes affecting the slow neutral hydrogen distribution are charge-exchange
with solar wind protons and solar Extreme UltraViolet (EUV) photoionization.
Solar wind proton fluxes have been extracted from the NASA Goddard Space Flight
Center’s OMNI dataset (a multi-spacecraft compilation of near-Earth solar wind
magnetic field and plasma measurements) through their online website OMNIWeb
(King and Papitashvili 2005), and have been time-averaged to reflect the slow
(4.22 AU/year) motion of hydrogen atoms through the solar system. The solar
EUV photoionization rates for hydrogen are from Space Environment Technologies’
Solar Irradiance Platform (SIP) version 2.37, a solar database at http://www.
spacewx.com (Tobiska et al. 2000). They have been time-averaged as follows: solar
wind charge-exchange data, photoionization data, and radiation pressure data are
averaged over 1–8 year intervals, with 1 year used far upwind, 1–2 years used
4.22–8.44 AU upwind of the Sun, 2–3 years used 0–4.22 AU upwind, 3–4 years used
0–4.22 AU downwind of the Sun, 4–5 years used 4.22–8.44 AU downwind, 5–6 years
8.44–12.66 AU downwind, 6–7 years used 12.66–16.88 AU downwind, 7–8 years used
16.88–21.10 AU downwind, and 8 years used beyond 21.10 AU downwind. The idea
is that downwind volume elements contain hydrogen that has had a longer period
of interaction with the solar wind, solar EUV, and the solar Lyman-α radiation
pressure force (Pryor et al. 1998a).
Slowly moving interstellar hydrogen atoms scatter solar Lyman-α photons, with
solar flux values taken here from the solar Lyman-α composite available at the
Laboratory for Atmospheric and Space Physics (LASP) website for the Thermo-
sphere Ionosphere Mesosphere Energetics and Dynamics (TIMED) mission Solar
EUV Experiment (SEE) (Woods et al. 2000). The relationship between integrated
Lyman-α line flux and the line-center Lyman-α line flux is taken from Emerich
et al. (2005). Lyman-α radiative transfer is assumed to be dominated by single
scattering near 1 AU, with a significant multiple scattering correction (Ajello et al.
2004) indicated by full exponential integral multiple scattering calculations (Hall
1992). The brightness of a particular sky direction on a particular day is deter-
mined from an integral over volume elements along a line of sight. A solar Lyman-α
value is assigned to each volume element on the basis of its solar longitude and the
brightness value of that solar longitude when measured at the Earth. The solar
measurements used are those closest in time to the observation date, sometimes up
to 13 days earlier or later on the basis of a 27-day solar rotation as seen from the
Earth. The assumption is that solar active regions and their Lyman-α fluxes change
slowly compared with 13 days. The model uses a simple formalism (Cook et al.
1981) for the modest solar Lyman-α falloff with solar latitude (Pryor et al. 1992).
The quiet sun polar Lyman-α resonance scattering coefficient (g-factor value) at
1 AU is taken as gquiet = 1.7 × 10−3 s−1 .
From the equatorial g-factor value, Lyman-α g-values at other latitudes (glat )
are estimated as:
gpole = gquiet + 0.35 ∗ (gave − gquiet ) (5.1)
166 5. LAMP Lyman-α Models
where gave is the 27-day equatorial average and gpole is the estimated polar g-value.
is the fraction of the equatorially-viewed disk covered by plage and gfact is the
equatorial g-value on a given day, and
is the fraction of the polar-viewed disk covered by plage. The fraction of the disk
covered by plage as viewed at a particular latitude is given by
and finally,
glat = gquiet (1 + 3.6 ∗ flat ) (5.5)
gives the estimated g-value at that latitude (Cook et al. 1981; Pryor et al. 1992).
MESSENGER MASCS
The MASCS instrument on MESSENGER consists of an Ultraviolet-Visible
Spectrometer (UVVS) and a Visible-Infrared Spectrograph (VIRS). UVVS cov-
ers the wavelength ranges of the far ultraviolet (115–180 nm), middle ultraviolet
(160–320 nm), and visible (250–600 nm), with an average spectral resolution of
0.6 nm (McClintock and Lankton 2007). UVVS uses an entrance slit with a field of
view of 1◦ by 0.04◦ for its airglow studies. MASCS UVVS has been studying Mer-
cury’s exosphere and surface reflectance properties initially from flyby observations
(McClintock et al. 2009) and more recently from orbit.
The radiometric sensitivity of the far-ultraviolet (FUV) channel of MASCS was
determined prior to the launch of the MESSENGER spacecraft. Measurements
were conducted in vacuum by observing the output from a monochromator with
both MASCS and a photomultiplier detector, which itself was calibrated against a
National Institute of Science and Technology (NIST) photodiode. Flight observa-
tions of stellar sources provide an opportunity to validate the MASCS radiometric
calibration. An adjustment of approximately 20 % was applied to the MASCS
spectral sensitivity in order to provide agreement with the Solar Radiation and
Climate Experiment (SORCE) satellite SOLar STellar Irradiance Comparison Ex-
periment (SOLSTICE)-measured irradiance of the star Alpha Virginis (Spica) in
the wavelength range 130–190 nm (Snow et al. 2013, this volume) and near Lyman-
α [M. A. Snow, private communication, 2012].
After launch on August 3, 2004, MESSENGER had an extended cruise phase
of its mission in the inner heliosphere until its orbital insertion at Mercury on
March 18, 2011. During part of the cruise phase, in 2009–2011, Lyman-α observa-
tions were obtained by UVVS over great circles as the spacecraft rolled at roughly
right angles to the spacecraft-Sun line at distances from the Sun ranging from 0.30
to 0.57 Astronomical Units (AU). The locations of MESSENGER during these
observations are indicated in Fig. 5.1. Because UVVS is a spectrometer, it is pos-
sible to separate the spectrum into heliospheric Lyman-α emissions (excess counts
MESSENGER MASCS 167
0.6 0.6 AU
2011-045
0.5 AU
0.4 0.4 AU
0.3 AU
0.2
Y_EC (AU)
–0.0 Sun
LISM flow
–0.2
–0.4
2009-124
–0.6
–0.6 –0.4 –0.2 –0.0 0.2 0.4 0.6
X_EC (AU)
Figure 5.2: MESSENGER interplanetary cruise Lyman-α data (top panel) and a
corresponding groove model (lower panel). The model was scaled by a factor of
0.66 to best agree with the data
conditions. Later, Pryor et al. (2003) assessed solar wind latitudinal changes in
Ulysses data from 1990 to 2001 and Quémerais et al. (2006) analyzed the changes
in SOHO SWAN data from 1996 to 2005. The SWAN data showed that the groove
was a narrow well-defined structure in the 1996 solar minimum observations, van-
ished in the 2002 solar maximum data, and reformed as a broader structure by the
next solar minimum in 2005.
We initially tried to fit MASCS data with a Ulysses-derived solar wind groove
model appropriate to the previous solar minimum and detailed by Pryor et al.
(1998b) and McComas et al. (1999). Figure 5.2 shows that this model has too
narrow a groove to describe the current epoch. The width of the groove may be
related to the varying tilt of the heliospheric neutral current sheet.
This Lyman-α model also has a built-in option (Witt et al. 1979) to vary the
solar wind with latitude. When this option is used, the solar latitude distribution
of the lifetime, t, of H atoms against charge-exchange with solar wind protons is
characterized by an asymmetry parameter, A, as follows:
linearly scaled model with the MASCS data is found for AN = 0.8 and AS = 0.5
(Fig. 5.3). This solution indicates that there is a north-south asymmetry with a
longer hydrogen lifetime in the north, explaining the brighter northern Lyman-α
measurements seen in the MASCS data of Fig. 5.3. A scaling factor was needed to
match our model to the MASCS data: the best-fitting model values for the IPH
brightness must be multiplied by 0.62 to match MESSENGER MASCS Lyman-α
data.
The latitudinal behavior of the solar wind lifetime derived from the MESSEN-
GER MASCS data, assuming AN = 0.8 and AS = 0.5, is shown in Fig. 5.4.
SOHO SWAN
SWAN, mounted on the SOHO spacecraft, provides regular Lyman-α maps of
the sky on a 1◦ by 1◦ grid (Bertaux et al. 1997). The SWAN instrument, described
by Bertaux et al. (1995) uses twin two-mirror periscopes mounted on opposite sides
of the spacecraft (+Z and −Z) to construct these maps. The detectors measure 5◦
by 5◦ on the sky at a time, using a Hamamatsu 5×5 multi-anode microchannel plate
array detector tube with a MgF2 window. The CsI photocathode of the detector
renders the instrument solar-blind at wavelengths above 200 nm, but leaves the
instrument with sensitivity to UV stars. Each unit has four unilluminated side
pixels suitable for instrumental dark current subtractions.
The initial published sensitivity of SWAN to Lyman-α radiation in photometer
mode was 0.75 counts per sec per Rayleigh per 1◦ by 1◦ pixel (Bertaux et al. 1995).
This ground calibration was based on comparison with an aluminum standard
photodiode from the National Institute of Standards and Technology (NIST) at a
synchrotron light source at Orsay. In-flight cross-calibration with Hubble Space
Telescope Goddard High Resolution Spectrometer (GHRS) data from March 9,
1996 led to slightly revised values that depended on which sensor was used and
its high-voltage setting (Bertaux et al. 1997). SWAN Lyman-α values cited in this
paper were the current values provided by the SWAN team in 2011.
Figure 5.3: MESSENGER interplanetary cruise Lyman-α data (top panel) and a
corresponding scaled model with AN = 0.8 and AS = 0.5 (lower panel). The result
of the model was scaled by a factor of 0.62 to best agree with the data
Figure 5.5: Sample SWAN Lyman-α all-sky map from March 6, 2011
Before assessing the RMS fit, the sum of the model values is linearly scaled to match
the sum of the data values. Figure 5.7 shows the RMS fits versus time from 2008
to 2011. Although the RMS fits do not change markedly over the orbit in 2010
or 2011, there is some seasonal variation in the fit in 2008 and 2009, suggesting
the solar wind fitting parameters derived from MASCS (mostly from 2010 to 2011
data) may no longer be appropriate. The agreement is adequate for our objective
of providing illumination values for the lunar poles for the LRO LAMP work, since
LRO was launched on June 18, 2009 and mapping operations began on September
15, 2009. Of greater interest is the trend in the derived scale factors shown in
Fig. 5.7. While the scale factor values seem steady and near 0.9 in 2008 and 2009,
a systematic decline is seen in 2010 and 2011. If the model were perfect, this
decline would suggest some degradation in the SWAN sensitivity. Alternatively,
this could be an indicator of model problems or some combination of data and
model problems.
Conclusions
We have identified a substantial offset between the MESSENGER MASCS val-
ues and the model: the best model used here must be scaled by a factor of 0.62 to
match the MASCS data. This same model is a better absolute fit to SWAN values
for 2008 and 2009 than to SWAN values in 2010 and 2011. Other techniques may
be valuable for assessing the hydrogen densities and deciding which brightnesses
are best calibrated. Some examples of such techniques are as follows. Damping at
increasing heliospheric distances of solar Lyman-α 27-day waves due to multiple
172 5. LAMP Lyman-α Models
Figure 5.6: Scaled model output with AN = 0.8 and AS = 0.5 (top panel) and
the corresponding SWAN map for January 1, 2010 (bottom panel). The model
was scaled by a factor of 0.90 to best agree with the data. The diagram shows
the location of the Earth (E) and Sun and indicates the interstellar wind upwind
(u) and downwind (d) directions. The small diagram below the panels shows the
relative locations of the Earth, spacecraft, and the interstellar wind directions
Figure 5.7: Top panel shows the derived RMS fits of the scaled model, with AN =
0.8 and AS = 0.5, to the filtered SWAN maps. The bottom panel shows the scale
factors applied to the model to agree with the data
Acknowledgements
Solar Irradiance Platform historical irradiances were provided courtesy of W.
Kent Tobiska and Space Environment Technologies. The historical irradiances have
been developed with partial funding from the NASA Upper Atmosphere Research
Satellite (UARS), TIMED, and ESA-NASA SOHO missions. We also acknowl-
edge use of the NASA Goddard Space Flight Center Space Physics Data Facility’s
OMNIWeb service and OMNI data. Calibrated SWAN data were provided by
Eric Quémerais (Laboratoire Atmosphères, Milieux, Observations Spatiales (LAT-
MOS)/Centre National de la Recherche Scientifique (CNRS)) and Stephane Ferron
(ACRI-ST). The MESSENGER contributions to this work are supported by the
NASA Discovery Program under contracts NAS5-97271 to The Johns Hopkins Uni-
versity Applied Physics Laboratory and NASW-00002 to the Carnegie Institution
174 5. LAMP Lyman-α Models
Bibliography
J.M. Ajello et al., Observations of interplanetary Lyman-alpha with the Galileo ultraviolet
spectrometer: multiple scattering effects at solar maximum. Astron. Astrophys. 289,
283–303 (1994)
J-L. Bertaux et al., SWAN: a study of solar wind anisotropies on SOHO with Lyman alpha
sky mapping. Sol. Phys. 162, 403–439 (1995)
J-L. Bertaux, E. Quémerais, R. Lallement, Observations of a sky Lyman alpha groove
related to enhanced solar wind mass flux in the neutral sheet. Geophys. Res. Lett. 23,
3675–3678 (1996)
J-L. Bertaux et al., First results from SWAN Lyman α solar wind mapper on SOHO. Sol.
Phys. 175, 737–770 (1997)
M. Bzowski, Density of neutral interstellar hydrogen at the termination shock from Ulysses
pickup ion observations. Astron. Astrophys. 491, 7–19 (2008)
M. Bzowski et al., Solar parameters for modeling the interplanetary background. In: Cross-
Calibration of Far UV Spectra of Solar System Objects and the Heliosphere, ed. by E.
Quémerais, M. Snow, R.M. Bonnet. ISSI Scientific Report Series, SR-013 (2013)
J.W. Cook et al., Latitudinal anisotropy of the solar far ultraviolet flux: effect on the
Lyman-alpha sky background. Astron. Astrophys. 97, 394–397 (1981)
J. Costa et al., Heliospheric interstellar H temperature from SOHO/SWAN H cell data.
Astron. Astrophys. 349, 660–672 (1999)
C. Emerich et al., A new relation between the central spectral solar H I Lyman alpha
irradiance and the line irradiance measured by SUMER/SOHO during the cycle 23. Icarus
178, 429–433 (2005)
G.R. Gladstone et al., LAMP: The Lyman alpha mapping project on NASA’s lunar recon-
naissance orbiter mission. Space Sci. Rev. 150, 161–181 (2010)
G.R. Gladstone et al., Far-ultraviolet reflectance properties of the Moon’s permanently
shadowed regions. J. Geophys. Res. 117, E00H04 (2011). doi:10.1029/2011JE003913
D.T. Hall, Ultraviolet resonance radiation and the structure of the heliosphere Dissertation,
University of Arizona, 1992
D. Hoffleit, W.H. Warren Jr., The Bright Star Catalogue, 5th Revised edn. (National Space
Science Data Center/Astronomical Data Center, 1991)
V.V. Izmodenov et al., Distribution of interstellar hydrogen atoms in the heliosphere and
backscattered solar Lyman-α. In: Cross-Calibration of Far UV Spectra of Solar System
Objects and the Heliosphere, ed. by E. Quémerais, M. Snow, R.M. Bonnet. ISSI Scientific
Report Series, SR-013 (2013)
J.H. King, N.E. Papitashvili, Solar wind spatial scales in and comparisons of hourly
Wind and ACE plasma and magnetic field data. J. Geophys. Res. 110, A02104 (2005).
doi:10.1029/2004JA010649
R. Lallement et al., Deflection of the interstellar neutral hydrogen flow across the heliospheric
interface. Science 307, 1447–1449 (2005)
W.E. McClintock, M.R. Lankton, The mercury atmospheric and surface composition spec-
trometer for the MESSENGER mission. Space Sci. Rev. 131, 481–521 (2007)
W.E. McClintock et al., MESSENGER observations of Mercury’s exosphere: detection of
magnesium and distribution of constituents. Science 324, 610–613 (2009)
D.J. McComas et al., Measurements of variations in the solar wind-interstellar hydrogen
charge exchange rate. Geophys. Res. Lett. 26, 2701–2704 (1999)
Bibliography 175
W.R. Pryor et al., The Galileo and Pioneer Venus ultraviolet spectrometer experiments:
solar Lyman-alpha latitude variation at solar maximum from interplanetary Lyman-alpha
observations. Astrophys. J. 394, 363–377 (1992)
W.R. Pryor, S.J. Lasica, A.I.F. Stewart, D.T. Hall, S. Lineaweaver, W.B. Colwell, J.M.
Ajello, O.R. White, W.K. Tobiska, Interplanetary Lyman alpha observations from Pioneer
Venus over a solar cycle from 1978 to 1992. J. Geophys. Res. 103, 26833–26849 (1998a)
W.R. Pryor, M. Witte, J.M. Ajello, Interplanetary Lyman alpha remote sensing with the
Ulysses interstellar neutral gas experiment. J. Geophys. Res. 103, 26813–26831 (1998b)
W.R. Pryor, J.M. Ajello, D.J. McComas, M. Witte, W.K. Tobiska, Hydrogen atom lifetimes
in the three-dimensional heliosphere over the solar cycle. J. Geophys. Res. 108, A108034
(2003). doi:10.1029/2003JA009878
W.R. Pryor et al., Radiation transport of heliospheric Lyman-alpha from combined Cassini
and Voyager data sets. Astron. Astrophys. 491, 21–28 (2008)
E. Quémerais et al., Interplanetary hydrogen absolute ionization rates: retrieving the solar
wind mass flux latitude and cycle dependence with SWAN/SOHO maps. J. Geophys.
Res. 111, A09114 (2006). doi:10.1029/2006JA011711
M. Snow et al., A new catalog of ultraviolet stellar spectra for calibration. In: Cross-
Calibration of Far UV Spectra of Solar System Objects and the Heliosphere, ed. by E.
Quémerais, M. Snow, R.M. Bonnet. ISSI Scientific Report Series, SR-013 (2013)
T.R. Summanen, R. Lallement, E. Quémerais, Solar wind proton flux latitudinal variations:
comparisons between Ulysses in situ data and indirect measurements from interstellar
Lyman alpha mapping. J. Geophys. Res. 102, 7051–7062 (1997)
T.R. Summanen et al., Interplanetary Lyman alpha observations of SWAN during the rising
phase of the 23rd solar cycle. Adv. Space Res. 29, 457–462 (2002)
G.E. Thomas, The interstellar wind and its influence on the interplanetary environment.
Ann. Rev. Earth Planet. Sci. 6, 173–204 (1978)
W.K. Tobiska et al., The SOLAR2000 empirical solar irradiance model and forecast tool. J.
Atmos. Sol. Terr. Phys. 62, 1233–1250 (2000)
N. Witt, J.M. Ajello, P.W. Blum, Solar wind latitudinal variations deduced from Mariner
10 interplanetary H (1216 A) observations. Astron. Astrophys. 73, 272–281 (1979)
M. Witte, Kinetic parameters of interstellar neutral helium. review of results obtained dur-
ing one solar cycle with the Ulysses/GAS-instrument. Astron. Astrophys. 426, 835–844
(2004)
T.N. Woods, W.K. Tobiska, G.J. Rottman, J.R. Worden, Improved solar Lyman alpha
irradiance modeling from 1947 through 1999 based on UARS observations. J. Geophys.
Res. 105, 27195–27215 (2000)
—6—
S. Alan Stern
Southwest Research Institute,
Boulder, CO, USA
Wayne R. Pryor
Central Arizona College,
Coolidge, AZ, USA
Abstract
Initial results are presented for observations of interplanetary Lyman-α and
Lyman-β emissions in the outer solar system obtained by the Alice ultraviolet
spectrograph on the New Horizons spacecraft (the first new such data from outside
the orbit of Saturn since the Voyager spacecraft). The observations consist of 6◦ ×
360◦ great-circle swaths on the sky, centered on the ecliptic direction λ = 51.3◦ , β =
44.8◦ , which passes within ∼ 33◦ of the upstream and downstream directions of the
interstellar wind. To date, three such scans have been acquired: on October 7, 2007,
October 18, 2008, and June 19, 2010 (at which times the New Horizons spacecraft
was 7.6, 11.3, and 17.0 AU from the Sun, respectively). The data compare fairly
well with model simulations, although the brightness of interplanetary Lyman-α
emissions falls off more slowly than expected with radial distance from the Sun.
The ratio of Lyman-α/Lyman-β brightnesses in the interplanetary medium agrees
well with previous measurements by the Voyager ultraviolet spectrometers.
Introduction
The interplanetary medium (IPM) and nearby interstellar medium (ISM) may be
fruitfully studied by measuring the sky background Lyman-α (121.6 nm) emission,
at nearly any location in the solar system (e.g. Bertaux and Blamont 1971; Thomas
and Krassa 1971; Fahr 1974; Adams and Frisch 1977; Holzer 1977; Thomas 1978;
Ajello et al. 1994; Murthy et al. 1999; Quémerais et al. 2009; Izmodenov 2009).
These emissions result from the scattering of solar Lyman-α photons by neutral
hydrogen atoms of the ISM as they pass by the Sun, and present an interesting
problem in resonance line radiative transfer (e.g. Brandt and Chamberlain 1959;
Meier 1977; Keller et al. 1981; Braskén and Kyrölä 1998; Quémerais and Bertaux
1993; Quémerais 2000).
The characteristic brightness of IPM Lyman-α in the vicinity of the Earth is
about 600 Rayleighs—i.e., about 6 × 108 photons cm−2 s−1 over the entire sky
(Ajello et al. 1987). For comparison, the direct Lyman-α flux from the Sun is
currently about 3–5 ×1011 photons cm−2 s−1 at the Earth. Although the direct
flux is a much more important source for, e.g., photochemistry in the inner solar
system, the interplanetary source falls off much more slowly than the direct flux
(Ajello et al. 1987), so that the two sources are of comparable strength at the orbit
of Neptune (Broadfoot et al. 1989), and the interplanetary source dominates from
there on out into the Kuiper Belt.
The general behavior of the scattered resonance line radiation may be approx-
imated by combining simple expressions for the optically thin and optically thick
limits. The backscattered IPM brightness BIPM may be approximated under opti-
cally thick and optically thin conditions as
1 F
BIPM ≈ (optically thick)
2 r2
1 F ΔλIPM
BIPM ≈ ksca (r) (optically thin),
2 r Δλ
where r is the distance from the Sun, F is the solar line flux, ΔλIPM and Δλ
are the line-widths of the appropriate resonance line in the interplanetary medium
and emitted by the Sun, respectively, and ksca is the line-center extinction due to
scattering. By simply combining these two limits, we obtain an expression useful
for approximating the backscattered mean intensity in both optically thick and
optically thin regimes:
1 F r ksca (r) ΔλIPM
Δλ
BIPM ≈ .
2 r 2 1 + r ksca (r) Δλ IPM
Δλ
The solar Lyman-α line is quite broad (FWHM ∼ 0.1 nm ≈ 250 km/s) com-
pared to the width (T ∼ 12, 000 K ≈ 14 km/s) and offsets (from ∼ −22 km/s to
∼ 22 km/s) of the scattered emissions from the IPM hydrogen, so that the illumi-
nation can be taken to be the same anywhere in the solar system (after accounting
for extinction on the way out from the Sun (Wu and Judge 1979). At a typ-
ical interplanetary H-atom number density of nH ∼ 0.1 cm−3 and temperature
of TH ∼ 12, 000 K (e.g. Izmodenov 2009), the line-center scattering cross section
σ0 is
πe2 f
σ0 = √
me c πΔνD
New Horizons observations 179
πe2 f λ0
= √
me c π 2kTH /mH
√
= (0.02654) (0.4162/ π) (1.216 × 10−5 /1.41 × 106 )
= 5.4 × 10−14 cm2 ,
so that the path length for unit optical depth at line center is
1
L =
nH σ0
1
= cm ∼ 12 AU.
(0.1) (5.4 × 10−14 )
Figure 6.1: During its annual checkouts (ACOs) the NH-Alice ultraviolet spectro-
graph is used to observe the IPM Lyman-α signal along a fixed great circle on the
sky. Such observations can be used to provide a measurement of the distribution
of interstellar hydrogen atoms flowing through the solar system. The trajectory
of New Horizons takes the spacecraft approximately radially upwind at a rate of
∼3 AU/year. Currently, we have IPM scans for ACOs 1, 2, and 4 on October 7,
2007, October 18, 2008, and June 19, 2010, at solar ranges of rNH = 7.6, 11.3, and
17.0 AU, respectively. These observations will continue at approximately yearly
intervals for as long as possible (at least out to 32 AU)
The filled-slit IPM Lyman-α emissions fall on bare, KBr-coated, and CsI-coated
sections of the MCP. The effective area for IPM Lyman-α in the Box is 0.056 cm2 ,
while in the Slot it is 0.007 cm2 . The overall sensitivity SIPM of NH-Alice to IPM
Lyman-α is given by
106
SIPM = [AEFF (Box) ω(Box) + AEFF (Slot) ω(Slot)]
4π
106
= [0.056 (2.0π/180)2 + 0.007 (π/180)2 (0.1)(4.0)]
4π
= 5.43 + 0.07 = 5.5 counts/s/R,
where AEFF is the NH-Alice effective area and ω represents solid angle.
Figures 6.3–6.5 show the great circle swath used during ACO-1, ACO-2, and
ACO-4, respectively, along with the locations of UV-bright stars and the Sun, in
ecliptic coordinates. The IPM swaths are identically positioned at 90◦ from the
ecliptic direction λ = 51.3◦ , β = 44.8◦ (in order to avoid overly bright UV stars and
the Sun). This great circle passes to within about 33◦ of the IPM upstream and
downstream directions. The circumstances during the three ACOs presented here
are given in Table 6.1. The observations made during each ACO were somewhat
different. During ACO-1, although it was planned that time-tagged data be would
New Horizons observations 181
Figure 6.2: NH-Alice nominal effective area, with the average effective areas shown
for IPM Lyman-α filling the 2◦ × 2◦ box and the 0.1◦ × 4◦ slot. Using these av-
erages, the expected sensitivity of NH-Alice to IPM Lyman-α is 5.5 counts/s/R.
The reduced sensitivity around Lyman-α (known as a “Lyman-α gap”) is accom-
plished by leaving the microchannel plate detector bare of any photocathode at
near-Lyman-α wavelengths (done to make the NH-Alice count rates due to bright
Lyman-α emissions more comparable to the count rates from other emissions)
Figure 6.3: Model IPM brightness contours appropriate for ACO-1 (rNH = 7.6 AU,
on October 7, 2007) are shown in blue at intervals of 50 Rayleighs. The great circle
swath of the IPM observed through the 6◦ -long NH-Alice slit is shown in green.
Bright UV stars are shown in orange (NH-Alice counts rates > 1, 000 counts/s) and
red (NH-Alice count rates > 5, 000 counts/s), and the Sun and the NH-Alice 20◦
solar keep-out zone are shown in orange. The upstream and downstream directions
for interstellar H atoms (Lallement et al. 2010) are indicated
Figure 6.4: Model IPM brightness contours appropriate for ACO-2 (rNH = 11.3 AU,
on October 18, 2008) are shown in blue at intervals of 50 Rayleighs. The great circle
swath of the IPM observed through the 6◦ -long NH-Alice slit is shown in green.
Bright UV stars are shown in orange (NH-Alice counts rates > 1, 000 counts/s) and
red (NH-Alice count rates > 5, 000 counts/s), and the Sun and the NH-Alice 20◦
solar keep-out zone are shown in orange. The upstream and downstream directions
for interstellar H atoms (Lallement et al. 2010) are indicated
New Horizons observations 183
Figure 6.5: Model IPM brightness contours appropriate for ACO-4 (rNH = 17.0 AU,
on June 19, 2010) are shown in blue at intervals of 50 Rayleighs. The great circle
swath of the IPM observed through the 6◦ -long NH-Alice slit is shown in green.
Bright UV stars are shown in orange (NH-Alice counts rates > 1, 000 counts/s) and
red (NH-Alice count rates > 5, 000 counts/s), and the Sun and the NH-Alice 20◦
solar keep-out zone are shown in orange. The upstream and downstream directions
for interstellar H atoms (Lallement et al. 2010) are indicated
The brightness data presented in Fig. 6.6 is entirely based on count rate in-
formation from housekeeping data, with a minor correction for nonlinearity, the
subtraction of ∼ 120 counts/s background (∼ 100 counts/s due to particles from
the radioisotope thermoelectric generator (RTG) plus ∼ 20 counts/s due to fidu-
cial “stim” counts), and conversion to brightness units assuming the filled-slit IPM
sensitivity of 5.5 counts/s/R estimated above. The data for the three ACOs are
shown in comparison with model predictions based on Pryor et al. (2008). The IPM
Lyman-α brightness is largest in the direction closest to the Sun, and the ratio of
the brightest/dimmest signal is about 2.2. The model fit is generally reasonable in
the upstream direction but tends to overestimate brightnesses near the Sun.
As mentioned above, one of the interesting differences between direct solar
Lyman-α and IPM Lyman-α is that the latter falls off much more slowly than
1/r 2 , so that it becomes relatively more important in the outer solar system (e.g.,
for methane photochemistry in the atmospheres of Neptune, Triton, and Pluto). As
noted by Gladstone (1993) and more recently by Throop (2011), photochemistry
by IPM Lyman-α was likely very significant in the early solar nebula. While more
NH-Alice data are needed, it appears that the first three ACOs show that the falloff
with rNH is even less than the ∼ 1/r dependence seen by Voyager (Hall 1992; Hall
et al. 1993). Figure 6.7 shows that, in the rNH ∼ 8–17 AU range at least, that
the current dependence is closer to ∼ 1/r 1/2 . It is interesting that the observed
dependence on r is much shallower than predicted by the simple expression for
184 6. New Horizons IPM Observations
Figure 6.6: IPM Lyman-α brightnesses observed by NH-Alice compared with the
model described by Pryor et al. (2008). Excursions due to UV-bright stars have not
been removed. The shifting longitude of the peak emissions is due to the changing
solar ecliptic longitude as seen from the NH spacecraft
BIPM presented in the introduction, even though the measurements encompass the
region where the turnover from an optically thin to an optically thick falloff should
occur. At this point we have no explanation for this behavior, and eagerly wait for
additional data with increased coverage in r to further investigate the problem.
Using the actual science data acquired during the three ACOs (i.e., spectral
images), we have searched for additional IPM emissions besides the dominant
Lyman-α emissions. Figure 6.8 shows the spectra integrated over the swaths ob-
tained during ACO-1, ACO-2, and ACO-4, with contamination from the brightest
stars removed. Although there is considerable structure in these spectra, most
of the behavior is explained by the broad wings of the Lyman-α line and edge
effects of the detector (at both short and long wavelength limits). An expected
few-Rayleigh feature due to resonant scattering of the solar 58.4 nm line by He
atoms in the interstellar wind is not clearly seen. This non-detection is likely due
to a combination of (1) detector edge effects, (2) the relatively large background
(∼ 100 counts/s over the entire detector, ∼ 2 count/s over the 58.4 nm line) from
the New Horizons RTG power source, and (3) the low instrumental sensitivity at
this wavelength (∼ 0.03 counts/s/R). However, a feature at 102.6 nm is identified
as due to resonant scattering of the solar Lyman-β line by IPM hydrogen atoms
(the Lyman-β line is far from the detector edge and the estimated instrumental
sensitivity is ∼ 0.10 counts/s/R). This line is very weak, at about the 0.5–1 R
Conclusions 185
level, as seen in Table 6.1 (the IPM brightnesses and brightness ratios provided
in Table 6.1 are derived from the swath-averaged spectra of Fig. 6.8). The IPM
Lyman-α/Lyman-β ratios are 754 ± 90 (ACO-1), 602 ± 52 (ACO-2), and 811 ± 233
(ACO-4), and are in good agreement with the Voyager result (using data from
inside 15 AU) of 700 ± 200 Murthy et al. 1999). The reason that the IPM Lyman-
α/Lyman-β ratio is so much larger than the solar Lyman-α/Lyman-β ratio (the
solar line center brightness ratio would be even smaller than the integrated line
ratio, since the Lyman-α line is broader than the Lyman-β line) is likely due to
the lack of substantial multiple scattering of Lyman-β in the IPM, due mostly to
its smaller resonance scattering cross section and to the substantial likelihood of a
branching to the Hα transition during each scattering event.
Conclusions
The IPM measurements made by the NH-Alice instrument on the New Hori-
zons mission to the Pluto system presented here represent the first new data from
outside of Saturn’s orbit since Voyager. It is clear that even with total count rate
information only that useful results are obtained with regard to the distribution
and brightness of IPM Lyman-α emissions in the outer solar system. We plan on
continuing such observations on a more or less yearly basis, until they are no longer
possible. Current extrapolations of the RTG performance suggest that NH-Alice
186 6. New Horizons IPM Observations
Figure 6.8: IPM spectra (±1σ) observed by NH-Alice on ACO-1, ACO-2 and ACO-
4, averaged over the great-circle swaths on the sky shown in Figs. 6.3–6.5, using
only detector rows corresponding to the 0.1◦ part of the slit (for improved spectral
resolution) and with star contributions removed. Besides Lyman-α, a feature due
to IPM resonant scattering of the solar line H 102.5 nm (Lyman-β) is detected,
although an expected signal at He 58.4 nm is not clearly seen
could be operable for ∼ 40 or so more years, although other factors are also im-
portant (e.g., how much fuel remains for IPM scans after the planned Kuiper Belt
Object flyby and whether an extended mission is approved). In addition, more
extensive IPM data will be available as a side effect of various airglow observations
that are planned for NH-Alice during the Pluto flyby (and its rehearsal). As New
Horizons moves further from the Sun the effects of multiple scattering in the ter-
mination shock should become more apparent, and the NH-Alice data should allow
for useful remote sensing studies of the termination shock region.
Acknowledgements
We thank the New Horizons mission team and the New Horizons science team. New
Horizons is funded by NASA, whose financial support we gratefully acknowledge. We also
thank ISSI for support of the working group. The authors would also like to thank the
reviewers and Priscilla Frisch for their useful comments.
Bibliography 187
Bibliography
T.F. Adams, P.C. Frisch, High-resolution observations of the Lyman alpha sky background.
Astrophys. J. 212, 300–308 (1977)
J.M. Ajello, A.I.F. Stewart, G.E. Thomas, A. Graps, Solar cycle study of interplanetary
Lyman-alpha variations: pioneer venus orbiter sky background results. Astrophys. J.
317, 964–986 (1987)
J.M. Ajello et al., Observations of interplanetary Lyman-α with the Galileo ultraviolet
spectrometer: multiple scattering effects at solar maximum. Astron. Astrophys. 289,
283–303 (1994)
J-L. Bertaux, J.E. Blamont, Evidence for a source of extraterrestrial hydrogen Lyman-alpha
emission: the interstellar wind. Astron. Astrophys. 11, 200–217 (1971)
J.C. Brandt, J.W. Chamberlain, Interplanetary gas. i. hydrogen radiation in the night sky.
Astrophys. J. 130, 670–682 (1959)
M. Braskén, E. Kyrölä, Resonance scattering of Lyman alpha from interstellar hydrogen.
Astron. Astrophys. 332, 732–738 (1998)
A.L. Broadfoot et al., Ultraviolet spectrometer observations of Neptune and Triton. Science
246, 1459–1466 (1989)
H-J. Fahr, The extraterrestrial uv-background and the nearby interstellar medium. Space
Sci. Rev. 15, 483–540 (1974)
G.R. Gladstone, Photochemistry in the primitive solar nebula. Science 261, 1058 (1993)
D.T. Hall, Ultraviolet resonance line radiation and the structure of the heliosphere. Disser-
tation, University of Arizona, Tucson, 1992
D.T. Hall et al., Heliospheric hydrogen beyond 15 au: evidence for a termination shock. J.
Geophys. Res. 98, 15185–15192 (1993)
T.E. Holzer, Neutral hydrogen in interplanetary space. Rev. Geophys. Space Phys. 15,
467–490 (1977)
V.V. Izmodenov, Local interstellar parameters as they are inferred from analysis of obser-
vations inside the heliosphere. Space Sci. Rev. 143, 139–150 (2009)
H.U. Keller, K. Richter, G.E. Thomas, Multiple scattering of solar resonance radiation in
the nearby interstellar medium. Astron. Astrophys. 102, 415–423 (1981)
R. Lallement, E. Quémerais, D. Koutroumpa, J-L. Bertaux, S. Ferron, W. Schmidt, P. Lamy.
The interstellar H flow: updated analysis of SOHO/SWAN data. In: 12th International
Solar Wind Conference, vol 1216. AIP Conference Proceedings, pp. 555–558 (2010)
R.R. Meier, Some optical and kinetic properties of the nearby interstellar gas. Astron.
Astrophys. 55, 211–219 (1977)
J. Murthy, D. Hall, M. Earl, R.C. Henry, J.B. Holberg, An analysis of 17 years of Voy-
ager observations of the diffuse far-ultraviolet radiation field. Astrophys. J. 522, 904–914
(1999)
W.R. Pryor et al., Radiation transport of heliospheric Lyman-α from combined Cassini and
Voyager data sets. Astron. Astrophys. 491, 21–28 (2008)
E. Quémerais, Angle dependent partial frequency redistribution in the interplanetary
medium at Lyman α. Astron. Astrophys. 358, 353–367 (2000)
E. Quémerais, J-L. Bertaux, Radiative transfer in the interplanetary medium at Lyman
alpha. Astron. Astrophys. 277, 283–301 (1993)
E. Quémerais, R. Lallement, B.R. Sandel, J.T. Clarke, Interplanetary Lyman α observations:
intensities from Voyagers and line profiles from HST/STIS. Space Sci. Rev. 143, 151–162
(2009)
M. Snow et al., A new catalog of ultraviolet stellar spectra for calibration. In: Cross-
Calibration of Far UV Spectra of Solar System Objects and the Heliosphere, ed. by E.
Quémerais, M. Snow, R.M. Bonnet. ISSI Scientific Report Series, SR-013 (2013)
S.A. Stern et al., Alice: the ultraviolet imaging spectrograph aboard the new horizons
Pluto-Kuiper belt mission. Space Sci. Rev. 140, 155–187 (2008)
G.E. Thomas, The interstellar wind and its influence on the interplanetary environment.
Ann. Rev. Earth Planet. Sci. 6, 173–204 (1978)
188 6. New Horizons IPM Observations
G.E. Thomas, R.F. Krassa, OGO 5 measurements of the Lyman alpha sky background.
Astron. Astrophys. 11, 218–233 (1971)
H.B. Throop, UV photolysis, organic molecules in young disks, and the origin of meteoritic
amino acids. Icarus 212, 885–895 (2011)
H.A. Weaver, S.A. Stern, New Horizons: NASA’s Pluto-Kuiper belt mission, in: The so-
lar system beyond neptune, ed. by M.A. Barrucci, H. Boenhardt, D.P. Cruikshank, A.
Morbidelli (University of Arizona Press, Tucson, 2008), p. 592
F.M. Wu, D.L. Judge, Modification of solar lines propagating through the interplanetary
medium. J. Geophys. Res. 84, 979–982 (1979)
Part III
Instrument
Cross-Calibration
—7—
John Clarke
Boston University,
Boston, MA, USA
Abstract
The SOLar-STellar Irradiance Comparison Experiment (SOLSTICE) on the
SOlar Radiation and Climate Experiment (SORCE) observes both the Sun and
stars in the ultraviolet (115–300 nm). Prior to launch, it was calibrated at the
SURF-III synchrotron. Spectra from the International Ultraviolet Explorer (IUE)
corrected to the white dwarf flux scale are in good agreement with SOLSTICE
observations, validating the two completely independent methods of calibration.
Measurements of stars in the SOLSTICE catalog are then used to transfer this
calibration to the SPectroscopy for the Investigation of the Characteristics of the
Atmosphere of Mars (SPICAM) instrument. We describe the steps used to calculate
the effective area for SPICAM to calibrate its stellar observations. Since only a
handful of stars in the IUE archive have been converted to the white dwarf scale
and many of them are relatively faint, the SOLSTICE catalog of bright stars can
be an extremely useful resource for inflight calibration of ultraviolet spectrographs.
Introduction
Early-type stars shine brightly in the ultraviolet (UV), and many can be easily
detected by astronomical space-based instruments. Additionally, the UV spectrum
of these main sequence stars is very stable over time (Mihalas and Binney 1981).
These two properties make them extremely attractive as calibration targets. Ob-
servations of a set of stars by one instrument can be compared to measurements
taken by a different instrument even if the two datasets were collected many years
apart.
A and B main sequence stars such as those included in the SOLar-STellar
Irradiance Comparison Experiment (SOLSTICE; McClintock et al. 2005a) and
SPectroscopy for the Investigation of the Characteristics of the Atmosphere of Mars
(SPICAM; Bertaux et al. 2006) target lists have higher effective temperatures than
a star like the Sun. Consequently, the peak wavelength of the blackbody emission
from their photospheres is in the ultraviolet rather than in the visible range. On
the Sun, the ultraviolet spectrum is formed in the chromosphere and is far more
variable than the visible spectrum which is formed at the photosphere. For A and B
spectral type stars, the stable photosphere produces a bright ultraviolet spectrum
over long timescales (cf Mihalas and Binney 1981). These stars can therefore be
an important transfer standard in cross-calibrating a large number of instruments.
All that is required is a set of absolutely calibrated spectra, ideally at a wide range
of apparent magnitude and widely distributed across the celestial sphere.
Instruments in orbit around the Earth or other celestial body often have sea-
sonal observing constraints much like ground-based observatories. A star in one
part of the sky may only be within viewing constraints for part of the year. An
ensemble of stars distributed around the sky will increase the probability that on
any given day, at least one of them will be available for observation. A wide range
of stellar brightnesses is also desired to make the catalog useful to the greatest
number of instruments. Depending on the design of the spectrometer, a given star
may be either too dim to produce a reasonable signal or be too bright and saturate
the detector system. Additionally, well-calibrated spectra of different levels may
provide diagnostic information about an instrument’s linearity.
In this paper, we will focus on two main topics. The first is a description of
the catalog of stars observed by the absolutely-calibrated SOLSTICE, including
comparisons to widely used archived spectra. These SOLSTICE stellar ultraviolet
spectra cover the wavelength range of 120–300 nm and are only observable from
space. The Earth’s atmosphere completely absorbs radiation at these shorter ul-
traviolet wavelengths. The second goal of this paper is to show how the catalog of
stellar spectra can be used to calibrate an instrument inflight that was launched
without a highly accurate ground calibration. In particular, we will apply this
technique to the UV channel of the SPICAM instrument onboard Mars Express
(Bertaux et al. 2006). After describing the observations by SOLSTICE and
SPICAM, we will compare their measurements to spectra of IUE standard stars
corrected to the white dwarf scale, and then finally discuss the results.
SOLSTICE 193
Grating Drive
Entrance Slit Assembly Sphere-Sphere
Assembly Kinematic Mount
Purge/
Fill Assy
Sunshade Fold Mirror
Assembly Assembly
Sphere-Translate
Kinematic Mount
SOLSTICE
SOLSTICE (McClintock et al. 2005a) onboard the SOlar Radiation and Climate
Experiment (SORCE; Rottman 2005) is a grating spectrometer whose primary
mission is to measure the solar spectral irradiance on a daily basis. SORCE was
launched in January 2003, and began taking measurements of both solar and stellar
spectral irradiance soon thereafter. The SOLSTICE instrument has two channels:
one observing the far ultraviolet (FUV) from 115 nm to 180 nm, and the other
observing the middle ultraviolet (MUV) from 180 nm to 300 nm. Figure 7.1 shows
the optical layout of the SOLSTICE instrument.
The instrument is designed to use the spectral irradiance from a set of early-
type stars as a calibration reference in order to track long-term changes in the
instrument. It accomplishes this goal by observing both the Sun and the stars
with the same optics and detector, changing only the size of the entrance aperture
and exit slit, in a highly controlled manner (Snow et al. 2005). Since the instru-
ment observes the solar spectral irradiance every orbit, it is expected to degrade
significantly over its lifetime. Repeated observations of the stellar calibration stars
allow this degradation to be corrected in the published data.
As described in McClintock et al. (2005b), the instrument was calibrated
before launch at the NIST SURF III facility in Gaithersburg, MD (Arp et al. 2000)
with an accuracy of about 3 %. During an early on-orbit observing campaign, this
calibration was transfered to the stars. All the stars in the SOLSTICE catalog were
observed in the first few days of the mission before the instrument was exposed to
the Sun. Therefore, SOLSTICE stellar spectra are absolutely calibrated and can
be useful for cross-calibration of other instruments that observe the same stars in
the ultraviolet.
194 7. Catalog of UV Stellar Spectra
Star name Bright star RA (2000) Dec (2000) V mag Spectral type
C(λ, τ, Dc , Sl , St )
E(λ) = (7.2)
RC (λ, T )F OV (λ, Ω, θ, φ)AEntrance ΔλBP DEG(t, λ, Ω, θ, φ)
S(λ)N (τ ) − Dc − Sl (λ) − St
C(λ, τ, Dc , Sl , St ) = (7.3)
Δt
where C is the instrument count rate computed by correcting the observed signal,
S(λ), for nonlinearity, N (τ ), where τ is the electronics dead time. Dc is the dark
current, Sl is the contribution from scattered light and St is the amount of stray
light. Δt is simply the integration time of the observation, and R is the instru-
ment responsivity at the beginning of life. The responsivity of the instrument was
measured before flight at the Synchrotron Ultraviolet Radiation Facility (SURF;
Arp et al. 2000) and is shown in Fig. 7.2. R is a function of wavelength and de-
tector temperature. In the case of stellar observations, the field-of-view is fully
SOLSTICE 197
Figure 7.2: Responsivities of SOLSTICE A and B in both the far ultraviolet (FUV)
and middle ultraviolet (MUV) channels. These responsivities were determined
based on calibration at the SURF-III synchrotron source before the spacecraft was
launched (From McClintock et al. 2005b)
filled and the integration over the angular size of the target described in McClin-
tock et al. (2005b). Equation 7.3 can be simplified because the contribution from
scattered light, stray light, and nonlinearity are all negligible in the SOLSTICE
stellar observations. It then becomes:
S(λ) − Dc (t)
C(λ) = . (7.4)
Δt
Dc is measured by pointing the instrument at dark space on a daily basis.
Figure 7.3 shows the contributions to the uncertainty in the SURF calibration.
The SURF beam has a well-known absolute irradiance (0.75 %), but since it is
a featureless continuum, it cannot be used to confirm the instrument wavelength
scale the way a line source would. The uncertainty in the wavelength scale and the
counting statistics of the observation represent about equal contributions to the
final uncertainty. The largest source of uncertainty in the preflight calibration is
the geometrical correction shown as “Gamma” in Fig. 7.3. This correction factor
enters into the calculation of the uncertainty in the responsivity because the SURF
beam does not fill the field-of-view of the instrument. Instead, it is more like a point
source. The calibration at SURF included a map of the instrument response to
the synchrotron point source over the field-of-view. The uncertainty in converting
this grid of map positions to the smooth field-of-view, Γ, is shown in the plot.
198 7. Catalog of UV Stellar Spectra
Figure 7.3: (top) Major sources of uncertainty in the SOLSTICE absolute cal-
ibration. The largest component, Gamma, is the uncertainty in the geometric
correction between the point source SURF beam and the full instrument field-of-
view. (bottom) The combined standard uncertainty of the SOLSTICE absolute
calibration from the components is shown in the top panel. For most wavelengths,
the uncertainty in responsivity is around 2 % (From McClintock et al. 2005b)
The lower panel of this figure shows the rms combination of these effects. For most
wavelengths, the combined uncertainty in the responsivity is around 2 %.
Unlike most instruments which measure stellar fluxes, SOLSTICE spends the
vast majority of its life staring directly at the Sun. This exposure to the solar
flux causes the instrument to degrade over time, so correction for changes in the
instrument responsivity are a necessary and important step in the data reduction.
The full description of the SOLSTICE technique for tracking degradation is avail-
able in Snow et al. (2005), but the following summary should provide the reader
with a sufficient understanding of the process to give confidence in the uncertainty
estimates that follow.
During the eclipse portion of each orbit, SOLSTICE observes stars that are
available based on celestial geometry. Rather than measuring the entire spectrum,
the typical observation is at a single wavelength in order to increase the signal-
to-noise ratio. Observations are planned to yield at least 104 counts in order to
SOLSTICE 199
reduce the fractional uncertainty due to counting statistics to less than 1 %. This
may take anywhere from 200 to 500 integrations of 1 s each, depending on the
star’s brightness. The wavelengths used to track changes in the responsivity curve
are evenly spaced throughout the wavelength range of each channel. There are 18
wavelengths in the FUV and 22 in the MUV. Five of the FUV wavelengths are
contaminated by geocoronal Lyman-α emission and a special technique is required
to isolate the stellar signal in this range. As described in Snow et al. (2005), the
Lyman-α airglow is removed by observing a dark region of the sky adjacent to the
target star before and after the stellar observation. The dark region measurement
includes the geocoronal emission, so the pair of observations can be used to estimate
the contribution of the airglow to the stellar signal taken between the two dark
region measurements. The analysis in this paper will be limited to the range 130–
300 nm to avoid the contribution of the airglow.
Figure 7.4 shows a typical fixed-wavelength observation of α Cru. A series
of 1 s integrations are analyzed statistically to determine the mean and standard
deviation as well as the likelihood that the data represent a normal distribution.
Observations passing these statistical tests are then used to determine the long-
term degradation of the instrument.
200 7. Catalog of UV Stellar Spectra
The top left panel of Fig. 7.5 shows the individual stellar irradiances from the
fixed-wavelength observations at 167 nm. The gradual decrease of irradiance with
time is at the same rate in all the stars, and therefore must be due to the degra-
dation of the instrument’s responsivity. The bottom left panel shows the same
irradiances normalized and fit with an exponential decay model. The relative
brightness of each star (i.e. the normalization factor) and the decay model are
determined simultaneously through a least-squares fit. Statistical outliers are re-
moved and the fit is repeated to produce the final degradation function for a given
wavelength. The whole procedure is repeated for each of the 40 wavelengths that
span the SOLSTICE sensitivity range.
The panel on the upper right shows the histogram of normalized stellar observa-
tions after correction for degradation. It is well fit by a Gaussian and is statistically
consistent with a single mean and an uncertainty due to random noise (χ2ν ∼ 1).
SOLSTICE 201
The bottom right panel identifies the stars used in the other plots and also shows
the average count rate. Since the detector dark rate is only a few counts per second
for the FUV channel, all the stellar irradiances used in the degradation analysis
are well above the background. The dark rate in the MUV channel is significantly
higher, and the stellar count rates at the long wavelength end of the spectrum
(λ > 270 nm) are only a few times larger than the dark rate. Stellar observations
above 270 nm are therefore less certain than those at shorter wavelength due to
statistical uncertainty in the dark rate correction.
SOLSTICE Uncertainty
The combined standard uncertainty in stellar irradiance for a SOLSTICE stellar
measurement is the combination of a variety of sources of error, some random
and some systematic. The main sources of systematic error are uncertainty in
the preflight calibration, uncertainty in the dark current subtraction, and inflight
correction for instrument degradation. These three sources of uncertainty are much
larger than the random uncertainty due to statistical noise in each observation,
although an error in the calculated degradation correction will appear to be a
larger spread in observed count rates from one observation to another.
Figure 7.6: Comparison of SORCE SOLSTICE and IUE spectra for three stars
calibrated to the white dwarf scale. The top panel shows the SOLSTICE spectra
for the three stars, while the lower panel shows the ratio of each observation to
the corresponding IUE spectrum from the CALOBS archive. The SOLSTICE
calibration based on a NIST standard is in good agreement with the IUE calibration
based on the white dwarf scale. The IUE spectra themselves are shown in Figs. 7.17,
7.18, and 7.27 in the appendix section
SPICAM-UV 203
take care to use only the spectra which have been corrected to the white dwarf
scale. The CALSPEC archive on the STScI web page also contains well-calibrated
spectra, but the only star which overlaps the SOLSTICE catalog is α Lyr. The
CALSPEC and CALOBS data files for α Lyr are identical in the 130–300 nm range.
SPICAM-UV
SPICAM-UV (referred to as simply SPICAM hereafter) is an instrument onboard
the Mars Express mission which was launched in 2003 (Bertaux et al. 2006) and
is currently in orbit around Mars. The primary science objective of the SPICAM
instrument is to make measurements of stars occulted by the Martian atmosphere.
These data are used to retrieve the density profiles of various components of the
Martian atmosphere. Because this only requires a relative measurement of changes
in the star’s spectrum as the line-of-sight passes through the atmosphere, it was not
necessary to have an accurate absolute calibration of the instrument before flight.
Linearity is important for an occultation measurement, but that can be tested
on the ground by varying the integration time while observing a stable source.
Absolute calibration can be a challenging task that may require taking the entire
instrument to a specialized facility. However, instead of using a ground-based
source tied to international standards, one can use a calibrated stellar reference
spectrum and transfer the calibration of that instrument to SPICAM. This section
will describe the process of using stellar reference spectra to determine SPICAM’s
calibration.
Stellar spectral measurements are performed by SPICAM during occultations of
stars by the atmosphere of Mars. These observations are fully described in Bertaux
et al. (2006). During the observation, the instrument boresight is pointed towards
a chosen star. The stellar spectrum is measured for about 10 min with the line-
of-sight well above the atmosphere of Mars. These data are used to compute the
reference spectrum. As the spacecraft travels along its orbit around Mars, the star
moves behind the planet and the line of sight approaches the surface of Mars. The
part of the observation where the star is occulted by the Martian atmosphere lasts
only a few minutes. The variation of the transmission is used to characterize the
composition and density profile of the atmosphere (Bertaux et al. 2006).
Initially, the SPICAM team had selected 39 stars to provide sufficient longi-
tudinal and latitudinal coverage for the study of the atmosphere of Mars. More
recently, the catalog has been extended by more than 200 stars and dedicated ob-
servations have been performed outside of the atmosphere. The stars observed by
SPICAM are listed in Table 7.4 in the Appendix section. At the time of publication
of this paper, SPICAM has observed 147 individual stars. It should be noted that
similar observations have been performed by the SPICAV-UV instrument of the
Venus-Express mission (Bertaux et al. 2007).
Figure 7.7: Optical layout of the UV channel of the SPICAM instrument on Mars
express. (1) aperture blend of the UV channel, (2) off-axis parabolic mirror, (3)
slit, (4) concave UV grating, (5) intensifier, (6) CCD, (7) flat mirror, (8) solar
opening (closed by a shutter when not viewing the Sun) (Adapted from Fig. 1 of
Bertaux et al. 2006)
device (CCD), i.e. an image intensifier coupled to a CCD through stacks of fiber
optics. The Hamamatsu image intensifier has a CsTe photocathode deposited on
a MgF2 window. Photoelectrons created at the photocathode by UV photons are
accelerated by the high voltage applied to a microchannel plate (MCP), creating
a cloud of electrons that is again accelerated to a phosphor screen. The resulting
pulse of green light is transmitted through the fiber optics to the CCD pixels. The
charge on each pixel is read and digitized to 12 bits, giving a maximum of 4,095
ADU (Analog to Digital units). Therefore the detector is used in a pseudo-counting
mode (Sandel and Broadfoot 1986). The gain of the MCP may be adjusted so
that a single photoelectron created at the cathode will produce 1–40 ADUs in the
CCD, spread over ≈ 6 pixels on the CCD matrix.
The CCD is an array of 408 by 288 pixels. The 288 lines are perpendicular to
the slit axis and parallel to the dispersion plane of the grating. The spectra are
spread along the lines of the CCD. Of the 408 pixels in each line, 384 are used to
measure the spectra. The remaining pixels, some of which are masked, are used to
determine the dark current of the CCD.
Each measurement has an integration time of a few hundred milliseconds. A
typical value is 640 min. The gain of the MCP is adjusted according to the bright-
ness of the source. Due to limited telemetry bandwidth, only five bands of 408
pixels are transmitted to Earth. The five bands are formed by summing individual
lines by 1, 2, 4, 8, 16 or 32 pixels. Typically, for stellar occultations, each band
is the sum of 16 physical lines of the CCD. Most of the stellar signal is included
in the central band. There has been very little degradation of the MCP over the
course of the Mars Express mission, likely because the voltage applied to it is fairly
low (Bertaux et al. 2006).
SPICAM-UV 205
The wavelength assignment is the relation between wavelength and pixel num-
ber on the spectral axis. Because the instrument behaves as an objective grating
spectrograph, the position of the source in the field-of-view influences the posi-
tion of the spectrum on the detector. Therefore the wavelength-pixel relation is
determined for a reference position, i.e. the center of the slit, but must be cor-
rected for the position of the source. This relation was initially measured on the
ground. However, this assignment had to be verified in space due to the possibility
of a change in alignment during or after launch. The wavelength assignment after
launch was measured by comparing SPICAM observations with an IUE spectrum
of γ Vel (HR 3207, see Fig. 7.8). The spectrum of this star shows various narrow
emission or absorption lines that allow for a good calibration of the spectrum posi-
tion on the CCD. The star spectrum can also be used to check whether the relation
between the photon wavelength and the pixel number is linear. For SPICAM-UV,
the wavelength assignment after launch is given by
where λ is the wavelength in nm and pn is the pixel number. The change between
ground and inflight calibration was very small. Figure 7.8 shows the spectrum
of γ Vel as determined by IUE. The spectrum obtained by SPICAM with the
correct wavelength assignment is shown by the red curve. The blue curve shows
the spectrum derived with the wavelength assignment measured on the ground.
There is a small shift between the two curves. This reference star can be used on a
206 7. Catalog of UV Stellar Spectra
Figure 7.8: The wavelength assignment was performed using the spectrum of γ
Vel (HR 3207) that shows narrow emission and absorption lines between 100 nm
and 300 nm. This allows us to check and correct the ground measurements of line
positions. The black curve shows the IUE spectrum. The spectrum obtained by
SPICAM with the correct wavelength assignment is shown by the red curve. The
blue curve shows the spectrum derived with the wavelength assignment measured
on the ground. The red and blue curves diverge at shorter wavelengths, but are
in close agreement at longer wavelengths. These spectra have been normalized to
arbitrary units for ease of comparison
SPICAM data are expressed in units of ADU (Analog to Digital Units), the
output of the CCD readout. When a photoelectron is created in the photocathode
(a photoevent), it eventually results in a pulse of light on the phosphor screen at
the back of the MCP. This pulse of light is distributed over a few pixels of the CCD.
It is detected by the CCD readout electronics as a total charge which is converted to
ADUs (NADU ). The high voltage (noted as HT hereafter) of the image intensifier
may be adjusted to give more or fewer electrons (hence ADUs) per photoevent, a
parameter that we note G, the gain of the UV detector. Therefore, the number of
counts, or NADU , depends on the gain of the detector, on the spectral resolution
Δλ and on the integration time Ti . We have the relation:
NADU = G × Np × Δλ × Ti (7.6)
SPICAM-UV 207
Figure 7.9: Linear relation between the signal SADU and its variance when observ-
ing a star at a given high voltage setting. The absolute gain is obtained from the
slope of the linear relation between the signal SADU and its variance
where Np is the number of photoevents detected by SPICAM per second and per
nanometer.
In other words, the gain G of a CCD camera is the conversion between the
number of photoevents (Np × Δλ × Ti ) collected at the photocathode during an
integration time Ti , and the number of digital units (NADU ) contained in the CCD
image.
There are two steps used to determine the absolute gain for all high voltage
settings. The first step is the determination of the gain Gk for a given high volt-
age. The reference gain Gk can be obtained inflight by an independent statistical
method. This is done by determining the mean number of photoevents contained
in the whole band devoted to the star signal outside of the atmosphere. The total
ADU signal fluctuations are analyzed, and the gain Gk is determined as:
a.b V ar(SADU ) V ar(SADU )
Gk = = 0.65 , (7.7)
c SADU SADU
where SADU is the mean value of a series of measurements and V ar(SADU ) is the
variance of the series. The scaling factor depends on different effects that modify
the statistical distribution of measurements. A detailed analysis by Dimarellis
(private communication, 1998) has shown that the scaling factor is affected by
• Factor c : The number of pixels used to determine the signal (binning of lines
of the CCD).
Figure 7.10: Variation of the number of counts (NADU ), for a given source, as a
function of the digital level of the high voltage setting. This curve (from Eq. 7.9)
shows the relative gain of the intensifier. This curve was obtained by comparing
stellar counts measured with different high voltage settings. This curve is very
stable over the whole dataset
G V
ln = ln
Gk Vk
and
Vi = a + b × HTi
PSF(x, y, xi , yi ) = Wi × H( ai , ki .r ) (7.10)
r= (x − xi )2 + (y − yi )2 (7.11)
2
a +∞
e−u
H(a, x) = du. (7.12)
π −∞ a + (u − x)2
2
Figure 7.11: Point spread function measured in the spatial axis for star ζ Pup
centered on line 144 of the CCD (black line). This figure was obtained by binning
all columns between pixels 100 and 200 on the CCD image. The blue curve shows
the result of the fit of the function given in Eq. 7.10. The same binning between
pixels 100 and 200 has been applied. The fit is the convolution of a Gaussian (core)
and a Lorentzian (wings). Outside 20 pixels from the star position the signal is
dominated by the background noise
spectrum observed by SPICAM were in a good agreement thus validating the as-
sumption.
Table 7.3: SPICAM-UV calibration stars. The columns show the SPICAM catalog
number, the bright star catalog number, the star’s Bayer designation, spectral type,
the level of the high voltage (HT) used for the observation, the number of repeat
observations for that star, and the source for the reference spectrum. The column
listing figure numbers refers to the spectra shown in the appendix
SPICAM HR Star Spectral HT Nobs Source Figure
number number name type
The main characteristics of the SPICAM stellar observations are also given in
Table 7.3. The High Voltage setting (HT) is low when the star is bright in the UV.
The number of observations shown in the table lists the number of separate obser-
vations with unique dates. For each distinct observation, the number of individual
spectra can go from a few hundred to a few thousand. Each spectrum has an inte-
gration time of 450 min or 640 min and the sampling rate during one observation
is 1 s. The star ζ Pup has been observed 132 times. This corresponds to more than
35,000 individual spectra. This star has been used to show the excellent stability
of the SPICAM-UV channel between October 2004 and April 2010. During that
period of time all spectra of that star show a variability lower then 5 % with no
systematic trend.
A visual inspection of the ratio between the SPICAM count rate and the SOL-
STICE irradiance shows that three stars, α Vir (O55), β Cen (O60) and α Cen
(O48), are outliers and will therefore not be used in the analysis. The other 14 stars
observed by both SOLSTICE and SPICAM will be used to derive the SPICAM
212 7. Catalog of UV Stellar Spectra
effective area. Six of the stars in the CALOBS archive—η UMa, ζ Oph, α Lyr,
α Leo, ζ Cas, and η Aur—have IUE measurements converted to the white dwarf
scale using the FLXCOR procedure from Bohlin (1996). Two other CALOBS
stars, ζ Pup and β Cen, are based on OAO-2 measurements that have been cor-
rected to the IUE scale but not the white dwarf scale (Bohlin and Holm 1981).
Since these two stellar spectra have not been converted to the white dwarf scale,
we have decided not to use them in our analysis.
The spectra in the CALOBS archive combine measurements from both the short
and long wavelength IUE cameras. The short wavelength channel covers the range
115–197.5 nm, while the long wavelength channel includes 191–330 nm. We note
that the long wavelength section of the spectrum for ζ Oph does not appear to be
consistent with the short wavelength portion.
As shown in Fig. 7.6, the calibration between SOLSTICE and IUE is consis-
tent to within their combined uncertainties. For wavelengths above 150 nm, the
uncertainty quoted for SOLSTICE is less than the nominal 3 % uncertainty quoted
for IUE spectra (Bohlin 1996). There are also 14 well-calibrated stars observed
by both SOLSTICE and SPICAM, while there are only six well-calibrated IUE
stars observed by SPICAM. Therefore we have chosen to use SOLSTICE irradi-
ance spectra to determine the SPICAM effective area for wavelengths greater than
150 nm and IUE irradiance spectra for wavelengths shorter than 150 nm.
Np = Aeff Φ. (7.13)
In this relation, Np is the number of photoevents per second per nm reported by
the instrument electronics and Φ is the incident stellar flux in photons/s/cm2 /nm
through the aperture.
If we assume that an input spectrum for a given star measured by another
instrument is well calibrated, then we can use it to derive the effective area. Aeff is
therefore only a function of the instrument and the ratio Np /Φ should be the same
for all stars. Figure 7.12 shows the derived Aeff using IUE spectra below 150 nm
and SOLSTICE spectra above 150 nm. There is excellent agreement between the
average of the six IUE spectra and the average of the 14 SOLSTICE spectra at
150 nm where the crossover occurs. This implies that there is internal consistency
in all three instruments for all the measured stars. The red and blue curves in
Fig. 7.12 represent the mean effective area computed for each wavelength range
respectively.
Figure 7.13: Irradiance ratio of SOLSTICE to SPICAM (top panel) and SPICAM
to IUE (lower panel) for stars observed by both instruments. The SPICAM to IUE
ratios are only for stars in the CALOBS archive that have been converted to the
white dwarf scale. The large ratio near Lyman-α is likely due to a mismatch in
spectral resolution rather than a true divergence in calibration
large variation with wavelength. This is due to the sharp features near Lyman-α
in all the spectra. Slight errors in either the wavelength scale or in matching the
spectral resolution before taking the ratio will lead to the large variance in the
ratio. There is no reason to believe that the true calibration changes so rapidly
with wavelength.
The other interesting feature in the SPICAM/IUE ratio is the broad feature
near 190 nm. In the 185–205 nm range, the SPICAM to IUE ratio, with a value of
1.06, is significantly larger than the mean ratio of SOLSTICE to SPICAM, which
is equal to 1.00 in the same range. This is the wavelength range where both IUE
channels have greater uncertainty in their responsivity and correspondingly lower
signal-to-noise ratio in all their spectra. Note that the transition between the two
SOLSTICE channels occurs at 180 nm. The two SOLSTICE channels appear to
be consistent since there is no discontinuity when switching from one to the other
at that wavelength. So we are confident that the region around 190 nm is less well
calibrated, and that explains the “bump” in the SPICAM/IUE ratio.
Discussion 215
Discussion
It has been more than 30 years since IUE was launched, yet it still forms the basis
of absolute calibration for ultraviolet astronomy. IUE observed so many targets
during its mission that nearly all current missions can find an IUE measurement
of any target that it observes. SPICAM and SOLSTICE observe many of the
same stars, and the IUE observations of those stellar spectra allow them to be
intercompared.
SPICAM uses SOLSTICE and IUE reference spectra from a limited number of
stars to determine its inflight calibration. The effective area calculated from the
average of those few spectra is then used to calibrate the observations of the rest of
its stellar catalog. This system produces spectra that are consistent to about 5 %.
That is presumably the uncertainty of the reduction method.
SOLSTICE, on the other hand, uses a preflight calibration at the SURF-III syn-
chrotron source (McClintock et al. 2005b) and then tracks changes to the instru-
ment responsivity by making regular observations of the ensemble of catalog stars
(Snow et al. 2005). The absolute uncertainty in the preflight calibration shown in
Fig. 7.3 is estimated to be about 3 % or better. The measurement uncertainty and
degradation correction are no larger than 2 %. These two sources of uncertainty
are independent and can therefore be added in quadrature.
The uncertainty of the IUE absolute calibration was initially estimated to be
10 % when based on only η UMa (Bohlin et al. 1990). Subsequent work using
model atmospheres of white dwarfs has reduced the estimated uncertainty of the
IUE absolute fluxes to just a few percent (Bohlin 1996). The repeatability of an
IUE observation is similarly at the few percent level (Bohlin et al. 1990). The
two panels of Fig. 7.13 confirm that the consistency from star to star for the three
instruments is on the order of a few percent. The average ratio between IUE and
SOLSTICE is close to 1 with a 1-σ spread of less than 5 %. The IUE spectra in
the CALOBS archive which have been converted to the white dwarf scale have
irradiances that match SOLSTICE’s, but the user should be alert to the fact that
some spectra in that archive have not been converted. Careful inspection of the
headers of the files in the CALOBS archive is strongly advised.
SORCE SOLSTICE’s absolute stellar irradiance is derived from a measured
calibration (McClintock et al. 2005b), while IUE’s ultimate calibration is from
a theoretical understanding of white dwarf atmospheres (Bohlin 1996). These
two methods are in good agreement over the full wavelength range considered here
(150–300 nm), which provides independent validation of the white dwarf model.
The ground calibration of SPICAM-UV was not very accurate. This study has
provided a new absolute calibration based on SOLSTICE’s absolute calibration
and which is therefore in agreement with the white dwarf absolute scale. This
calibration can then be used to produce more spectra in absolute units based on
SPICAM’s stellar observations. The SOLSTICE and SPICAM spectra available
on the FONDUE web service, http://bdap.ipsl.fr/fondue/, are consistent with the
best calibrated IUE spectra.
216 7. Catalog of UV Stellar Spectra
Acknowledgements
The work was supported by NASA contract NAS5-97045 (SORCE) at the Uni-
versity of Colorado. We would like to thank the referees for their assistance in
understanding the IUE white dwarf scale and their helpful comments on the initial
draft of the paper. The authors would also like to thank the International Space
Science Institute, Bern, Switzerland for their support of the working group.
Appendix
For each of the 19 stars listed in Table 7.3, we show spectra from each of
the three instruments—IUE, SOLSTICE, and SPICAM—if good spectra exist for
that star. The ratio between the spectra is also shown in the lower panel of each
figure. The corresponding data are available on-line from the FONDUE web service,
http://bdap.ipsl.fr/fondue/. The spectra are in FITS files including statistical
uncertainties.
224 7. Catalog of UV Stellar Spectra
Table 7.4: List of all stars observed by SPICAM. Columns give the SPICAM ob-
serving number, the bright star catalog number, star’s Bayer designation, spectral
type, celestial coordinates, and visual magnitude
SPI HR Name and constellation Spectral type RA2000 DEC2000 Mag
Bibliography
U. Arp, R. Friedman, M.L. Furst, S. Makar, P.S. Shaw, SURF III—an improved storage
ring for radiometry. Metrologia 37, 357–360 (2000). doi:10.1088/0026-1394/37/5/2
J-L. Bertaux et al., SPICAM on Mars express: observing modes and overview of
uv spectrometer data and scientific results. J. Geophys. Res. 111, E10S90 (2006).
doi:10.1029/2006JE002690
J-L. Bertaux et al., SPICAV on Venus express: three spectrometers to study the global
structure and composition of the Venus atmosphere. Planet. Space Sci. 55, 1673–1700
(2007)
R.C. Bohlin, Spectrophotometric standards from the far-uv to the near-ir on the white dwarf
flux scale. Astrophys. J. 111, 1743–1747 (1996). doi:10.1086/117914
R.C. Bohlin, A.V. Holm, New software — absolute calibration — iue. IUE ESA Newsletter
11, 18 (1981)
R.C. Bohlin, W.M. Sparks, A.V. Holm, B.D. Savage, M.A.J. Snijders, Photometric calibra-
tion of the international ultraviolet explorer (IUE): low dispersion. Astron. Astrophys.
85, 1–13 (1980)
R.C. Bohlin, A.W. Harris, A.V. Holm, C. Gry, The ultraviolet calibration of the hubble space
telescope iv. absolute IUE fluxes of hubble space telescope standard stars. Astrophys. J.
Suppl. 73, 413–439 (1990). doi:10.1086/191474
R.C. Bohlin, M.E. Dickinson, D. Calzetti, Spectrophotometric standards from the far-
ultraviolet to the near-infrared STIS and NICMOS fluxes. Astrophys. J. 122, 2118–2128
(2001). doi:10.1086/323137
R.C. Bohlin et al., Absolute flux calibration of the IRAC instrument on the Spitzer space
telescope using hubble space telescope flux standards. Astrophys. J. 141, 173–185 (2011).
doi:10.1088/004-6256/141/5/173
E.H. Eberhardt, Gain model for microchannel plates. Appl. Opt. 18, 1418–1423 (1979).
doi:10.1364/AO.18.001418
E. Marcq, D. Belyaev, F. Montmessin, A. Fedorova, J-L. Bertaux, A.C. Vandaele, E. Neefs,
An investigation of the SO2 content of the venusian mesosphere using SPICAV-UV in
nadir mode. Icarus 211, 58–69 (2011). doi:10.1016/j.icarus.2010.08.021
W.E. McClintock, G. Rottman, T.N. Woods, Solar stellar irradiance comparison ex-
periement II (SOLSTICE II): instrument concept and design. Sol. Phys. 230, 225–258
(2005a). doi:10.1007/s11207-005-7432-x
W.E. McClintock, M. Snow, T.N. Woods, Solar stellar irradiance comparison experiment II
(SOLSTICE II): pre-launch and on-orbit calibrations. Sol. Phys. 230, 259–294 (2005b).
doi:10.1007/s11207-005-1585-5
D. Mihalas, J. Binney, Galactic astronomy structure and kinematics (Freeman, New York,
1981), p. 135
G. Rottman, The SORCE mission. Sol. Phys. 230, 7–25 (2005). doi:10.1007/s11207-005-
8112-6
G. Rottman, T.N. Woods, T. Sparn, Solar stellar irradiance comparison experiment
1. I — instrument design and operations. J. Geophys. Res. 98, 10667 (1993).
doi:10.1029/93JD00462
B.R. Sandel, A.L. Broadfoot, Statistical performance of the intensified ccd. Appl. Opt. 25,
4135–4140 (1986). doi:10.1364/AO.25.004135
M. Snow, W.E. McClintock, G. Rottman, T.N. Woods, Solar stellar irradiance comparison
experiment II (SOLSTICE II): examination of the solar-stellar comparison technique. Sol.
Phys. 230, 295–324 (2005). doi:10.1007/s11207-005-8763-3
—8—
Abstract
The Moon has been shown to be an extremely stable radiometric reference for
calibration and long-term stability measurements of on-orbit sensors. The majority
of the previous work on characterizing the lunar reflectance has been in the visible
part of the spectrum using ground-based lunar images. The SOLar-STellar Irradi-
ance Comparison Experiment (SOLSTICE) on the SOlar Radiation and Climate
Experiment (SORCE) can be used to extend the lunar spectral irradiance dataset
to include the 115–300 nm range. SOLSTICE can directly measure both the solar
and lunar spectra from orbit, using the same optics and detectors. An observing
campaign to map out the reflectance as a function of phase angle began in mid 2006
and continued through 2010. The geometry of SORCE’s orbit is very favorable for
lunar observations, and we have measurements spanning a range 0–170◦ in phase
angle. In addition to Earth Observing Systems using the Moon for calibration, re-
cent planetary missions have also made ultraviolet observations of the Moon during
Earth flyby, and these SOLSTICE measurements can be useful in calibrating the
absolute responsivity of those instruments as well.
Introduction
One of the many challenges to quantitative, space-based, absolute remote-
sensing measurements is tracking long-term instrumental changes. Comparison
to a standard source is one way to monitor such changes. Some missions may carry
a comparison source onboard, while others may rely on cross-calibration with other
observations such as a rocket underflight. But if a suitable celestial target whose
radiometric properties are well known is available, then it can be a valuable weapon
in the calibration arsenal. The Moon is such a radiometric standard.
Over 25 years ago, it was recognized that the Moon could be used for on-orbit
calibration (Kieffer and Wildey 1985, 1996; Pugacheva et al. 1993), and a ground-
based observatory program to determine the radiometric properties of the Moon
in the visible and near-infrared part of the spectrum began in 1996 (Kieffer and
Anderson 1998; Stone and Kieffer 2002). The goal of the LASP Lunar Albedo
Measurement and Analysis from SOLSTICE (LLAMAS) project is to extend the
measurements of the lunar irradiance to the ultraviolet using data from the SOLar-
STellar Irradiance Comparison Experiment (SOLSTICE) (McClintock et al. 2005a)
on the SOlar Radiation and Climate Experiment (SORCE) spacecraft (Rottman
2005).
The first comprehensive campaign to characterize the disk-integrated
wavelength-dependent phase curve of the Moon was conducted by Lane and Irvine
(1973). The geometric albedo was determined for each bandpass filter across the
visible region and an empirical phase curve was derived over the phase range of
6–120◦ . The results from that study effectively demonstrated the phase-reddening
effect. This is the phenomenon where the color of the Moon tends to become
increasingly red at larger phase angles. In this context, color is the ratio of the
brightness of the Moon at a longer wavelength to a shorter wavelength.
The RObotic Lunar Observatory (ROLO) was a program within the United
States Geological Survey (USGS) to accurately determine the exoatmospheric spec-
tral irradiance and phase curve of the Moon from 350 to 2450 nm (Stone and Kieffer
2002). The Moon was observed every clear night throughout a six-year period cov-
ering phase angles of −90 to +90◦ (negative phase angles refer to lunar phases prior
to full Moon) through multiple bandpass filters in the visible and near-infrared. An
empirical model with 18 fitting coefficients in each of 32 bands for the lunar disk-
integrated reflectance was developed. Eight of these coefficients are common to all
bands for a total of 328 coefficients in the ROLO photometric model. The residual
from this fit was about 1% (Kieffer and Stone 2005).
The Moon has been observed at ultraviolet wavelengths by many other space-
based missions, and a careful comparison of their results with the SOLSTICE
measurements is planned for future publication. Figure 8.1 shows the primary
datasets for the ultraviolet wavelength range as a function of wavelength and phase
angle coverage.
Starting at the shortest wavelengths, the Hopkins Ultraviolet Telescope (HUT)
(Henry et al. 1995) and Apollo 17 (Lucke et al. 1976) determined the albedo from
radiance measurements of relatively small regions of the lunar disk. The Nozomi
ultraviolet scanner took an image of the Moon at a phase angle of about 130◦ ,
but the only wavelength that had good signal was Lyman-α (Taguchi et al. 2000).
Mariner 10 (Wu and Broadfoot 1977), Galileo (Hendrix 1996), Cassini (Hendrix
et al. 2009), and the MErcury Surface, Space ENvironment, GEochemistry, and
Ranging (MESSENGER) spacecraft (Holsclaw et al. 2010) are all planetary mis-
sions that took data during lunar flybys. The footprint of their fields-of-view (FOV)
on the Moon depend on their distance from the Moon during the observation.
Instrument Description 229
Signal levels also vary widely among datasets, and we have just begun to work
on intercomparing them. The International Ultraviolet Explorer (IUE) took sev-
eral spectra of the Moon, but those results are still unpublished (Michael Combi,
private communication).
In the middle ultraviolet (MUV, 180–300 nm), Galileo and the Student Nitric
Oxide Explorer (SNOE) made partially overlapping measurements at a phase angle
of about 20◦ (Hendrix 2002). These were also radiance observations over varying
FOVs. The Global Ozone Monitoring Experiment (GOME; Dobber 1996, 1997)
and the Shuttle Solar Backscatter Ultraviolet (SSBUV; Janz et al. 1996) made spec-
tral irradiance measurements of the Moon, but since these were specially planned
events, they span a very small range of phase angle. The Advanced Camera for
Surveys (ACS) on the Hubble Space Telescope (HST) obtained measurements of
the lunar radiance at 250 nm and a few visible wavelengths (Robinson et al. 2007).
Finally, in the near-UV, dedicated observational campaigns by Lane and Irvine
(1973) using the Harvard College Observatory and Kieffer and Stone (2005) using
ROLO have filled in large ranges of phase angle over these wavelengths and into
the visible. The ROLO dataset is by far the most complete.
The Lunar Reconnaissance Orbiter (LRO) has been in orbit around the Moon
since June 2009 and includes the Lyman Alpha Mapping Project (LAMP) instru-
ment (Gladstone et al. 2010). In addition to its primary mission of searching for
ice in the permanently shadowed polar regions, it is also taking measurements of
the lunar albedo in the ultraviolet and will ultimately map out a large fraction of
the Moon on the dayside. The Wide Angle Camera on the Lunar Reconnaissance
Orbiter Camera (LROC) has a number of filter bandpasses in the near ultravi-
olet, but the shortest wavelength they observe is 320 nm (Robinson et al. 2010).
Although there is no overlap between LROC/WAC and SOLSTICE, trends in spec-
tral and photometric properties vary continuously with wavelength, so trends seen
in the two instruments will provide useful insight into the surface properties of
the Moon.
As Fig. 8.1 clearly shows, there have been many ultraviolet lunar observational
datasets, but none have the wavelength and phase angle coverage that the SOL-
STICE measurements provide. Prior to SOLSTICE, there was very little overlap in
phase and wavelength between the various sets of observations, so intercomparisons
were generally quite difficult and had fairly large uncertainties. Our new dataset
can be a bridge between them all and will greatly improve our knowledge of the
lunar albedo throughout the UV for all observing geometries.
Instrument Description
SOLSTICE is a grating spectrometer that can measure ultraviolet irradiance
over an extremely large dynamic range (McClintock et al. 2005a). Its primary
mission is to measure the absolute solar irradiance on a daily basis. It is on board
the SOlar Radiation and Climate Experiment (SORCE) (Rottman 2005) which is
a small satellite in low earth orbit. During orbit day, SOLSTICE measures the
solar spectral irradiance, but during the eclipse portion of the orbit, the spacecraft
turns and SOLSTICE measures the irradiance from a variety of celestial targets
230 8. Ultraviolet Irradiance of the Moon
Figure 8.1: Catalog of lunar ultraviolet albedo measurements. The shaded areas
represent the phase and wavelength coverage for each dataset. An underlined
dataset name indicates that the measurement comes from images. Dataset names
in parentheses are not full-disk measurements. (From Snow et al. 2007)
including the Moon. The altitude of the orbit is 580 km with an inclination of 40◦ .
This 96-minute orbit allows lunar viewing on a routine basis, including nearly the
full range of phase angles (0–170◦ ).
Before launch, SOLSTICE was calibrated using Beam Line 2 at the Synchrotron
Ultraviolet Radiation Facility III (SURF III) at the National Institute of Standards
and Technology (NIST) in Gaithersburg, MD (Arp et al. 2000). This synchrotron
beam is a standard source in the ultraviolet and the uncertainty in its irradiance
is only 0.75% (Arp et al. 2000). The final preflight accuracy of SOLSTICE’s re-
sponsivity is on the order of 5% (McClintock et al. 2005b). To monitor long-term
degradation of the instrument response, SOLSTICE uses an ensemble of early-type
stars as its calibration reference source, thus the need for a dynamic range of 108 .
The wavelength range measured by SOLSTICE is from 115 to 300 nm using two
channels. In solar mode, the entrance aperture is a 0.1 mm square and the exit slit
for each channel is 0.375 mm wide. The spectral resolution in this mode is 0.1 nm.
The instrument can exchange these small entrance and exit slits with much larger
apertures to convert to stellar mode (i.e. lunar mode).
Replacing the small solar entrance aperture with a large 16 mm circular aper-
ture and increasing the exit slit width to 0.75 mm decreases the resolution dur-
ing a stellar measurement to 1.1 nm in the far ultraviolet (FUV) channel (115–
180 nm). The exit slit in the middle ultraviolet (MUV) channel (180–300 nm) is
1.5 mm for a spectral resolution of approximately 2.2 nm in the 180–300 nm range.
Determination of the spectral resolution of a lunar measurement is somewhat
Instrument Description 231
Grating Drive
Entrance Slit Assembly Sphere-Sphere
Assembly Kinematic Mount
Purge/
Fill Assy
Sunshade Fold Mirror
Assembly Assembly
Sphere-Translate
Kinematic Mount
more complicated and will be explained below. The rest of the optical system is
unchanged in the switch from solar to stellar mode. For part of the MUV solar
measurement, a neutral density filter is placed into the optical path just before
the exit slit to reduce the detector count rate. In-flight tracking of this filter has
shown no measureable changes to its transmission since launch. Figure 8.2 shows
the opto-mechanical layout of the SORCE SOLSTICE instrument.
In stellar mode, the instrument operates as an objective grating spectrometer.
While this mode is well-suited to lunar observations, it does introduce a few com-
plications to the data processing. The two major issues are wavelength shifts as
the target moves within the FOV and vignetting of the incoming rays. Both of
these effects will be discussed in detail below.
SOLSTICE has the unique ability to measure both the solar and lunar irradi-
ances with the same optics and detectors. The entrance and exit slits are different
between the two modes, but the elements that disperse, reflect, or detect light, i.e.
the optical elements that might change properties over time, are the same for both.
The ratio of the aperture areas for the two modes was measured by NIST before
flight and has an uncertainty of 0.5% (McClintock et al. 2005b). A detailed un-
certainty analysis will be included in the section “Uncertainty” below. Most other
albedo measurements rely on either model solar spectra or must include the un-
certainties from relative calibration of two separate instruments. The ratio of solar
and lunar irradiances can be calculated directly from the SOLSTICE observations
and a large number of calibration factors in this ratio cancel out.
In the discussions that follow, we will need to make reference to the x, y, and
z coordinate frame of the instrument. The z-axis is the direction of the optic axis,
often referred to as the boresight. The x-axis is in the dispersion direction for
SOLSTICE B, and the y-axis is the cross-dispersion direction. The plane defined
by the instrument, the Moon, and the Sun is known as the scattering plane. The
232 8. Ultraviolet Irradiance of the Moon
roll angle of the instrument will be defined as the angle between the x-axis and the
surface normal to the scattering plane.
Libration Coverage
Although only one side of the Moon generally faces the Earth, there is a small
oscillation of the selenographic longitude and latitude of the sub-observer point
due to the eccentricity of the lunar orbit and the inclination of the Moon’s axis
of rotation relative to the plane of its orbit around the Earth. Similarly, the
sub-spacecraft point on the Moon varies as a function of time as the celestial
geometry evolves. With regular observations from mid-2006 until the end of 2010,
SOLSTICE has achieved fairly good coverage of the full range of possible libration
angles. Unlike a ground-based observatory, the selenographic coordinates of the
observation can sometimes change by a significant amount on short timescales
due to the orbital motion of the spacecraft. Figure 8.4 shows the position of the
sub-spacecraft point on the Moon at the beginning and end of each of the FUV
observations (diamonds). The start and stop points for each scan are connected
by a line. Occassionally, the sub-spacecraft point will have moved by as much as
a degree during the observation, but in general the change is much smaller. The
MUV observations have a very similar distribution of selenographic positions, so
we only show the FUV observations in Fig. 8.4.
Data Reduction
In order to create a photometrically accurate model of the lunar reflectance
over the entire wavelength and phase angle region of interest, the raw data must
be carefully corrected and reduced. Some of the reduction steps are the same
as for a standard SOLSTICE stellar observation, but many of them are specific
234 8. Ultraviolet Irradiance of the Moon
Figure 8.4: Selenographic coordinates (in degrees) of the sub-spacecraft point. Due
to the orbital motion of the spacecraft, the sub-spacecraft point changes throughout
the observation. Pairs of diamonds (connected by lines) show the start and end
selenographic position for each scan
to the lunar observing conditions. We will briefly describe the additional correc-
tions needed to fully reduce the lunar observations in this section. The discussion
in the following sections extends the basic data reduction process described in
Snow et al. (2007).
Pointing
The SORCE spacecraft has only two tracking modes: inertial targets and the
Sun. Unfortunately, the Moon is neither of those. In order to observe the Moon,
we schedule and execute the observation as if it were an inertial target. Parallax
due to spacecraft orbital motion causes the Moon to drift away from the center
of the FOV during an observation. To minimize this problem, we have developed
an observing strategy where we re-point at the current position of the Moon every
2 min, so it never gets more than 10 arc minutes from the center. Figure 8.5 shows
the position of the Moon in the FOV for typical observations. The two axes are the
pitch and yaw directions in the instrument frame of reference. When the spacecraft
re-points at the current position of the Moon, it also makes adjustments to the roll
angle. The roll angle refers to rotations about the spacecraft z-axis. Therefore the
drift direction of the Moon can be slightly different after each pointing correction.
Because SOLSTICE operates as an objective grating spectrometer, offsets in the
position of the target from the instrument boresight have the effect of displacing
Data Reduction 235
Figure 8.5: Position of the Moon in the field-of-view of the instrument for a few
typical observations. The axes are the pitch and yaw directions for the instrument.
The crosses show the position of the center of the lunar disk as it drifts away
from the center of the FOV during the observation. The spacecraft is periodically
re-pointed to the new position of the Moon to re-center it, but during each such
maneuver, the roll angle of the spacecraft is allowed to change, so the new direction
of the Moon’s drift can be different after each re-pointing
the effective wavelength sampled by the detector. This shift can be positive or
negative and changes continuously throughout an observation depending on the
attitude of the spacecraft. Uncorrected wavelength shifts would lead to errors in
the reflectance that would be particularly large when there are sharp features in the
spectrum. This effect can be characterized analytically from the grating equation:
where m is an integer representing the diffraction order, and θS is the grating angle,
given by
β+α
θS = , (8.2)
2
and the deviation angle is defined as
β−α
φG = . (8.3)
2
236 8. Ultraviolet Irradiance of the Moon
α and β in these relationships are the incidence and diffraction angles, respectively.
The stellar-mode wavelength equation is given by:
where is the offset angle of the star (or in our case, the Moon) from the boresight
in the grating dispersion plane. The nominal value (i.e. when the target is located
on the optic axis of the instrument) of φG is fixed by the instrument geometry
and has a value of 2.32◦ (McClintock et al. 2005a). For all wavelengths such that
|β| > |α|, both incident and diffracted beams lie on the same side of the grating
normal, and the adopted sign convention is such that m = +1 with both α and β
positive. The angle is effectively a change in the grating incidence angle, α, but
does not change the diffraction angle, β.
McClintock et al. (2005a,b) provide the stellar-mode wavelength equation as a
function of pointing offset. It should be noted that Eqs. 2 and 3 of McClintock
et al. (2005a) are meant to be equivalent to Eqs. 27 and 28 of McClintock et al.
(2005b); however, the definition of the constant deviation angle (φG ) is different in
these two papers which leads the stellar-mode wavelength equation to be incorrect
in both. The wavelegth relationships presented here use a consistent definition and
should be used instead of the equations in the two McClintock papers.
Using the reconstructed spacecraft attitude, the angular offset of the Moon
center from the instrument boresight is known. However, the attitude derived from
the spacecraft star trackers is sampled at a different cadence and less frequently
than the SOLSTICE detector readout. The quaternions representing the spacecraft
attitude need to be interpolated to allow a determination of the effective wavelength
of each detector sample.
There is a secondary wavelength shift that must also be established. The grating
position reported by the instrument may be in error by an amount that is fixed
for each observation but will vary from one observation to the next. This grating
offset is described in Snow et al. (2005) and can produce a shift in the wavelength
scale of up to a few nanometers, although it is typically less than 1 nm. In data
processing, the magnitude of this offset is determined by an iterative least-squares
procedure to align the spectrum to the solar spectrum.
Vignetting
SOLSTICE was designed to accommodate either the diverging beam from the
solar entrance aperture (a pinhole) or the collimated beam from the stellar aperture
(16 mm diameter) and all of the optical surfaces were sized appropriately. Table 8.1
lists the sizes of each optic in the SOLSTICE optical path along with the size of the
lunar beam as determined from raytracing. For lunar observations, we must use
the large stellar entrance aperture, and the diverging beam from the Moon nearly
fills the first mirror. If the Moon is not properly centered in the FOV, some of the
light will be vignetted. As is shown in Table 8.1, the lunar image very nearly fills
the first optic. Therefore, the pointing errors during lunar observations caused by
parallax discussed in the section “Pointing” become important for even relatively
small angles.
Data Reduction 237
Figure 8.6 shows the loss of signal as a function of pointing offset angle as
determined by a raytrace model developed with ZEMAX optical design software.
The upper panel shows the loss of signal in the cross-dispersion direction and the
lower panel shows the loss in the dispersion direction. The loss of signal due to
vignetting in the dispersion direction is a function of wavelength. The vignetting
correction becomes more significant at longer wavelengths because the projected
area of the grating becomes smaller as the grating is rotated to a larger angle
relative to the incident beam. If this effect were not taken into account during
data processing, it would introduce a wavelength dependent artifact into the lunar
reflectance.
Since the vast majority of the measurements are made within 5 arc minutes of
the spacecraft boresight, this is a small correction for the MUV channel (less than
5 %). But in the case of the FUV channel, there is a 20 arc minute offset between
the spacecraft and instrument boresights. Therefore, we must make a significant
correction for vignetting to the FUV data.
Spectral Resolution
The spectral resolutions for the various observing modes of SOLSTICE men-
tioned in the section “Instrument Description” and discussed in McClintock et al.
(2005a) are applicable for the Sun or for a stellar point source. The situation for
lunar observations is somewhat more complicated. Given the optical layout of SOL-
STICE, the image of the full Moon at the exit slit has a diameter of approximately
1.57 mm (wavelength dependent). The width of the exit slit in MUV stellar mode
is 1.5 mm, and 0.75 mm in the FUV channel (McClintock et al. 2005a). Convolving
this circular image with the rectangular exit slits for each channel produces the
two instrument profiles shown in Fig. 8.7. These profiles assume ideal imaging con-
ditions and will be assumed to be lower limits to the true spectral resolution while
observing the Moon. We discuss an estimate of the uncertainty of these profiles in
the section “Uncertainty”.
We have calculated the spectral resolution for lunar observations as a func-
tion of phase angle and the angle between the projection of the Moon’s rotational
axis on the sky and the spacecraft y-axis based upon geometrical analysis of the
instrument. The lunar axis angle relative to the instrument dispersion direction
becomes particularly important at large phase angles because a crescent Moon ori-
ented parallel to the exit slit will yield a very different spectral resolution than
if the lunar axis is perpendicular to the slit. A non-zero angle will also lead to
an asymmetric profile. The roll angle of the spacecraft is part of the telemetered
cross-dispersion plane
1.00
0.95
0.90
fraction
100nm
0.85 150nm
200nm
250nm
0.80 300nm
350nm
0.75
0.70
-30 -20 -10 0 10 20 30
angle (arcmin)
dispersion plane
1.00
0.95
0.90
fraction
100nm
0.85 150nm
200nm
250nm
0.80 300nm
350nm
0.75
0.70
-30 -20 -10 0 10 20 30
angle (arcmin)
Figure 8.6: Lunar vignetting correction. As the Moon moves away from the center
of the field-of-view, the image on the optics is vignetted. This curve shows the
fraction of rays which reach the exit slit as a function of angle based upon a
raytrace model of the instrument. The left panel shows the vignetting for offsets in
the cross-dispersion direction, the right panel is for the dispersion direction. The
wavelength dependence is only significant in the dispersion direction
spacecraft quaternion. Information about the lunar axis can be determined from
the ephemeris, and the mean geometry is used for each observation. Although
it is possible to convolve the lunar image with the exit slit for each sample, it is
much more computationally intensive. The celestial geometry generally changes
very slowly over the orbit, so using the mean geometry does not introduce a large
systematic uncertainty. However, we recognize that this effect should be quantified
to determine if it is worth the extra computation time to implement in the next
Data Reduction 239
Figure 8.7: Instrumental spectral profiles for FUV and MUV channels. These
curves come from the convolution of a circular lunar image (full Moon) with the
rectangular exit slit for each instrument configuration. The exit slit in FUV mode
is about half the diameter of the full Moon image, while the MUV exit slit is about
equal to the size of the image. (From Snow et al. 2007)
version of the data processing system. These spectral profiles are used as the convo-
lution kernel for the high-resolution solar spectrum in order to match the resolution
of the lunar observation.
240 8. Ultraviolet Irradiance of the Moon
Figure 8.8: (Top) FUV lunar spectral scan at full Moon, acquired on 7 October
2006 at a phase angle of 2.0◦ . Data points marked with X contain contribution
from airglow. Error bars shown are due to counting statistics and do not include
systematic uncertainties. (Bottom) MUV lunar irradiance spectrum at full Moon,
acquired on 2 April 2007 at a phase angle of 3.4◦ . Uncertainty due to counting
statistics is smaller than plot symbols. (From Snow et al. 2007)
feature in the spectral scan data. Correction for airglow contamination is fairly
straightforward, but the quantitative analysis has not been completed in time for
this publication.
242 8. Ultraviolet Irradiance of the Moon
Albedo
Since the spectrum of the Sun changes with solar activity, the irradiance
spectrum of the Moon will also change. For purposes of using the Moon as a
calibration target, a more useful quantity than the irradiance may be the disk-
equivalent albedo (Kieffer and Stone 2005), which is derived from the ratio of lunar
to solar irradiance. This quantity is a property of the Moon alone, and it is there-
fore stable over long time periods (Kieffer 1997). The relationship between albedo
and irradiance is given by:
EM /ΩM
AM = . (8.5)
E /π
In this equation, AM is the disk-equivalent albedo, ΩM is the solid angle subtended
by the Moon, EM is the irradiance of the Moon, and E is the irradiance of the
Sun. In our notation, ΩM varies with the Moon-spacecraft distance. Since the
solar and lunar observations are taken close in time, there is no need to correct the
irradiances to 1 AU; such a correction would cancel out in the ratio. The ratio of
lunar to solar irradiance for the SOLSTICE instrument is simply:
EM CM a Δλ
= Γ. (8.6)
E C aM ΔλM
In this expression, C and CM are the solar and lunar count rates, a and Δλ are
the entrance aperture area and radiometric bandpass of the two instrument modes.
The function Γ is a geometrical correction factor which accounts for differences
in responsivity for targets of different angular size. The correction factor is a few
percent at most wavelengths [Figure 4 of McClintock et al. (2005b)]. Both C
and CM include correction factors for dark current, scattered light, temperature,
filter transmission. We believe that these instrumental corrections have all been
accounted for, but it is possible that there could still be subtle systematic differences
between the solar and lunar observing conditions that have not yet been identified.
Figure 8.9 shows the measured disk-equivalent albedo for the FUV and MUV
channels on SOLSTICE. For simplicity, we show only the albedo for observations
in the range of 9–11◦ phase in the figure. The feature at about 215 nm in the lower
panel of Fig. 8.9 is likely an artifact. There is a neutral density filter that is inserted
at that wavelength during the solar scan, so it is suspicious that there is a feature
in the reflectance at that same wavelength. Further investigation of the properties
of this filter will resolve this uncertainty.
Since SOLSTICE directly measures both solar and lunar irradiance with the
same optics and detectors, this ratio is independent of solar variability and long-
term instrument degradation. Random and systematic errors in this ratio will
be discussed in the section “Uncertainty”. Most importantly, this ratio is also
independent of the absolute calibration of SOLSTICE since it is a direct ratio.
The ratios shown in Fig. 8.9 can be measured at all phase angles. Figure 8.10
shows the typical behavior of the albedo as a function of phase angle for one
wavelength (282 nm). Near full Moon an effect known as the opposition surge causes
a rapid increase in brightness as the phase angle decreases. The behavior of the
lunar integral photometric function at large phase angles is less well studied. The
only other dataset of a similar surface with comparable phase coverage is the visible
Lunar Irradiance and Albedo 243
Far Ultraviolet
0.10
0.08
Disk-Equivalent Albedo
0.06
0.04
0.02
0.00
130 140 150 160 170 180 190
Wavelength (nm)
Middle Ultraviolet
0.10
0.08
Disk-Equivalent Albedo
0.06
0.04
0.02
0.00
180 200 220 240 260 280 300
Wavelength (nm)
Figure 8.9: (Top) FUV disk-equivalent albedo for phase angles between 9 and 11◦ .
Measurements have been averaged over 2 nm (approximately two resolution ele-
ments). Data from eleven SOLSTICE scans were used in this plot. (Bottom) MUV
disk-equivalent albedo for phase angles between 9 and 11◦ . Measurements from
sixteen SOLSTICE scans were averaged over 2 nm intervals (about one resolution
element)
244 8. Ultraviolet Irradiance of the Moon
Figure 8.10: Disk equivalent albedo as a function of phase angle for 282 nm.
Negative phase angles indicate a waxing Moon
phase curve of the planet Mercury as measured by the Large Angle Spectroscopic
COronograph (LASCO) onboard the Solar and Heliophysical Observatory (SOHO).
The phase curve of Mercury exhibits a similar sharp decline in albedo at large phase
angles (Mallama et al. 2002). This SOLSTICE dataset will further understanding
of the lunar surface in this regime.
Uncertainty
In order for the SOLSTICE lunar observations to be a useful calibration dataset,
a thorough understanding of the uncertainties is critical. We present here a pre-
liminary analysis of the major contributors to uncertainty. From McClintock et al.
(2005b), the irradiance measurement equation for SOLSTICE is given by:
C
E= , (8.7)
R · FOV · AEntrance · ΔλBP · DEG
where C is the count rate (corrected for dead time, dark current, and scattered
light), R is the responsivity, FOV is the variation in responsivity over the field-of-
view of the instrument, and AEntrance · ΔλBP is the aperture area and radiometric
bandpass. DEG is a function of time and wavelength that accounts for changes
in the instrument responsivity. From Eq. (23) of McClintock et al. (2005b), the
uncertainty of the ratio between a solar and stellar irradiance measurement is:
2 2 2 2 2
σE/E = σC
+ σC
+ σG/G + σR A−BP
+ σf2AU + σΓ2 . (8.8)
In this equation, σC and σC are the uncertainties in solar and stellar (or lunar)
count rates. These are sources of random error. σG/G is the uncertainty in ratio
Uncertainty 245
of the gain factors for the two modes. Γ is a correction factor that accounts for
the slightly different illumination between solar and stellar configurations. When
comparing solar to stellar irradiances, a correction to 1 AU is necessary to account
for the eccentricity of the Earth’s orbit, and σfAU is the uncertainty of that correc-
tion. However, there is no 1 AU correction needed in the lunar case so σfAU can be
neglected. Finally, σRA−BP is the uncertainty in the responsivity due to errors in
the aperture and radiometric bandpass. These four terms are systematic sources
of error and will be discussed first.
Systematic Error
The gain and FOV (i.e. Γ) terms contribute 0.1% and 1.0% respectively
[Table III of McClintock et al. (2005b)]. These two terms are wavelength-
independent sources of systematic error. The illumination of the optics during
lunar observations is slightly different than for stellar mode. The stellar illumina-
tion is basically a collimated beam. The lunar illumination is a diverging beam
that illuminates a somewhat larger portion of each optic than the stellar beam,
but the difference in responsivity for the two cases is very small. Therefore we are
using the stellar value for Γ in the lunar processing. We do not expect that the
contribution to systematic error from an error in Γ will ever be larger than the 1%
that we have already added to the error budget. The correction to 1 AU can be
neglected in the albedo measurement since both solar and lunar count rates are
affected by the same factor if the observations are made close in time. For the
case of lunar observations, we also must account for systematic uncertainties in our
vignetting correction and convolution kernel.
The vignetting correction is derived from a raytrace model of the instrument,
using the measured sizes of the optical elements. We can test the model for the
stellar case with observations. Once a week, the SORCE spacecraft performs a
cruciform alignment maneuver while SOLSTICE observes a star. The maneuver
covers the range ±5.5◦ in both dispersion and non-dispersion planes. Our raytrace
model agrees quite well with the stellar alignment data, which gives us confidence
that it is also accurate in the lunar case. We will assume for now that we know this
correction factor to the 5% level. A quantitative estimate of the accuracy in the
stellar case is planned for the next version of the analysis. The vignetting correction
for the MUV is typically less than 0.5%; therefore uncertainty in this correction to
the MUV data makes a negligible contribution to the overall error budget. In the
case of the FUV, where the magnitude of the correction is aproximately 7%, we
estimate a 0.35% contribution to the systematic uncertainty that is independent of
wavelength.
For the uncertainty in the convolution kernel, we can empirically estimate the
effect of an error in its assumed value. The lunar albedo is presumed to be slowly
varying with wavelength at this spectral resolution, so an error in the convolution
kernel will cause the features in the solar spectrum to show up as features in the
albedo. The uncertainty in the ratio due to counting statistics then sets an upper
limit on our knowledge of the convolution kernel. The size of an artifact due to
error in the kernel must be larger than the random errors in the measurement. We
estimate that the lunar convolution kernel is accurate to 5%. This estimate will
246 8. Ultraviolet Irradiance of the Moon
likely decrease as we analyze the data further. The profile that we are currently
using is from the simple convolution which assumes ideal imaging conditions. A
more detailed model may result in changes to our assumed instrument profile.
Since these sources of systematic error are independent, we can add them in
quadrature. By far, the uncertainty in the convolution kernel is the dominant
source of systematic error. The total contribution from systematic errors in the
lunar albedo measurement is likely no larger than 5.1%.
Instrumental artifacts that could influence the value of EM /E include sys-
tematic differences in temperature, libration angle, etc. since the solar and lunar
observations are not done simultaneously. However, we think that we have made
conservative estimates for the contribution to the total combined uncertainty for
these corrections and included them. We will continue to make improvements to
our data processing system and update the results.
Random Error
Random error sources also contribute to the uncertainty of the lunar albedo
measurement. These errors can be reduced in magnitude by repeated observations
at each phase and libration angle. This category of error includes counting statistics
and uncertainties in the lunar wavelength scale. The former is a straightforward
calculation of the square root of the number of counts. The latter is somewhat
more complex. In the current version of data processing, we are neglecting the
variation due to libration angle and averaging all observations from a given phase
angle together. The random variation due to statistics is likely much larger than
the systematic variation due to libration angle. Based upon the ROLO photometric
model, variations due to libration are at most a few percent for phase angles smaller
than 90. We have not yet quantified the magnitude of the change in reflectance
with libration angle in the ultraviolet. A future version of the data processing will
take libration angle into account.
As discussed in the section “Pointing”, the SOLSTICE wavelength scale is very
repeatable, although there can be an offset introduced when the instrument is re-
set, i.e. in Eq. 8.4 (Snow et al. 2005). But more importantly, since SOLSTICE
acts as an objective grating spectrometer in stellar mode, the wavelength scale will
be continually changing during an observation due to the parallax motion of the
Moon in the FOV. A change of 5 arc minutes in position is about 16 of the lunar
diameter. This translates to about a 0.4 nm shift in the wavelength scale. The cur-
rent processing system calculates the appropriate wavelength shift for each sample
based on spacecraft telemetry and applies it to the lunar spectrum. Our estimate
of the residual error in this method is about 0.1 nm. The uncertainty shown in
Fig. 8.11 is the change in irradiance for a wavelength scale offset of that size.
Polarization
Understanding the polarization of the lunar irradiance is important for both
calibration and scientific investigation of the lunar surface. In this section we will
briefly discuss both aspects. The SURF beam used during SOLSTICE’s preflight
Polarization 247
Figure 8.11: Sources of random error in a single SOLSTICE lunar albedo measure-
ment. The top panel shows the FUV channel while the bottom panel shows the
MUV. The contribution from these errors to the final albedo measurement can be
reduced by averaging multiple observations. A systematic error of about 5% needs
to be added in quadrature to the random errors to get the final uncertainty in each
measurement
248 8. Ultraviolet Irradiance of the Moon
0.025
sensitivity 0.020 H
0.015
0.010 V
0.005
0.000
180 200 220 240 260 280
wavelength (nm)
0.8
0.6 (H-V)/(H+V)
polarization
0.4
0.2
0.0
180 200 220 240 260 280
wavelength (nm)
Figure 8.12: Response of SOLSTICE MUV channel to polarized light. H and V are
the signal levels detected by the instrument when the dispersion direction is either
Horizontal or Vertical relative to the SURF polarization plane. The difference of
these two orientations divided by the sum gives the fractional polarization detected
by the instrument
0.25
WUPPE (not full disk) [8] waning
waxing
0.20
polarization SOLSTICE
0.15
0.05
0.00
200 300 400 500
wavelength (nm)
Figure 8.13: Percent polarization of the lunar irradiance 55◦ phase angle. Box and
X symbols show the results from Coyne and Pellicori (1970). Diamonds mark the
SOLSTICE observations. The SOLSTICE measurement is derived from two spec-
tra taken at orthogonal spacecraft roll angles. The middle ultraviolet wavelength
band appears to be very consistent with measurements from the visible band
Summary
We have presented preliminary results of measurements of the lunar irradiance,
and more importantly, the lunar albedo at a large range of phase angles throughout
the UV (115–300 nm). The current processing system accounts for the dynamic
observing geometry from a satellite in low earth orbit and accurately produces lunar
reflectances. Improvements to this system are continuing as part of the LASP Lunar
Albedo Measurement and Analysis from SOLSTICE (LLAMAS) project funded by
NASA and final data products are scheduled for delivery in 2013.
This dataset will be used to create a photometric model of the lunar irradiance
that can then be used for both calibration of any Moon-viewing instrument and
for scientific investigations of the surface properties of the Moon. A photometric
model of the lunar observations for UV wavelengths over the full range of phase and
libration angles will allow Hapke photometric parameters (Hapke 1981, 1984, 1986,
1993) to be derived and compared with previously published results in the visible
(e.g. Helfenstein and Veverka 1987). This comparison will help us to understand
the physical mechanisms for scattering from the lunar surface at ultraviolet wave-
lengths. Discriminating mineralogic features are known to exist in the ultraviolet
(Wagner et al. 1987) and their identification in this data would demonstrate an
additional method of detection with remote sensing instruments.
Initial analysis of the polarization properties of the lunar irradiance indicates
that trends observed in the visible continue through at least the middle ultraviolet.
Measuring the polarization of the sunlight reflected off the Moon will help us to
understand the physical state of the lunar surface.
The current status of the data processing system is that the MUV channel data
is nearly ready for photometric modeling, but the FUV dataset processing is not
yet ready for routine production. We have summarized the major sources of both
systematic and random error in the measurements. Further analysis and additional
measurements will likely further reduce these uncertainties in the future.
The data from SOLSTICE overlap the observations from a wide variety of
previous experiments, none of which had the full range of wavelengths and phase
angles contained in the new dataset. This overlap will allow the cross-calibration
of a large number of planetary missions once the SOLSTICE data has been fully
reduced.
The surface reflectance properties of the Moon are known to be stable on very
long time scales (Kieffer 1997). The solar irradiance in the ultraviolet varies much
more than it does in the visible, so accurate solar measurements (even if based on
a proxy) will be necessary to make the Moon a useful UV calibration target in the
future. Beyond the SORCE mission, there will likely be MUV of the Sun observa-
tions from an instrument such as the Spectral Irradiance Monitor (SIM) which is
currently flying on SORCE and a similar instrument is presently manifested as the
Total and Spectral Irradiance Sensor (TSIS) on the Joint Polar Satellite System
(JPSS) scheduled for launch later this decade (http://www.nesdis.noaa.gov/jpss).
Alternatively the MUV solar spectrum can also be fairly well-modeled through
correlations with the Mg II index (Chandra et al. 1995). This index is currently
measured on a daily basis by the National Oceanic and Atmospheric Adminis-
tration (NOAA) and will certainly be continued by future missions. Uncertainties
Bibliography 251
inherent in solar models based on proxies are larger than those from measurements,
but since the amount of solar variability in the MUV is only a few percent, the
model irradiance may be adequate. The FUV solar irradiance measurement after
SORCE will also have to come from models and proxy measurements. The current
NASA plan does not include a FUV solar measurement. The SORCE primary
mission ended in January, 2008, but it is continuing to operate on an extended
mission. Spacecraft operations on SORCE have curtailed the lunar observations
during eclipse, but during the lunar observing campaign, we acquired a very exten-
sive dataset of lunar irradiance that will produce valuable scientific and calibration
results.
Acknowledgements
This research is supported by NASA grant NNX09AQ60G (LLAMAS) and contract
NAS5-97045 (SORCE) at the University of Colorado. The authors would like to thank
the International Space Science Institute, Bern, Switzerland, for their generous support
of the working group.
Bibliography
U. Arp, R. Friedman, M.L. Furst, S. Mkar, P.-S. Shaw, SURF III—an improved storage
ring for radiometry. Metrologia 37, 357–360 (2000). doi:10.1088/0026-1394/37/5/2
S. Chandra, J.L. Lean, O.R. White, D.K. Prinz, G.J. Rottman, G.E. Brueckner, Solar uv
irradiance variability during the declining phase of the solar cycle 22, Geophys. Res. Lett.
22, 2481–2484 (1995). doi:10.1029/95GL02476
G.V. Coyne, S.F. Pellicori, Wavelength dependence of polarization. xx. the integrated disk
of the moon. Astron. J. 75, 54–60 (1970). doi:10.1086/110940
M.R. Dobber, Moon observations with gome. Proc. SPIE 2831 154–168 (1996)
M.R. Dobber, GOME moon measurements, including instrument characterization and moon
albedo. In: Guyenne T-D, Danesy D (eds) Third ERS Symposium on Space at the service
of our Environment, ESA SP-414, p. 743, Florence, Italy, March 1997
A. Dollfus, E. Bowell, Polarimetric properties of the lunar surface and its interpretation.
part 1. telescopic observations. Astron. Astrophys. 10, 29–53 (1971)
A. Dollfus, Lunar surface imaging polarimetry: i. roughness and grain size. Icarus 136,
69–103 (1998)
G.K. Fox, A.D. Code, C.M. Anderson, B.L. Babler, K.S. Bjorkman, J.J. Johnson, M.R.
Meade, K.H. Nordsieck, A.J. Weitenbeck, N.E.B. Zellner, Solar system observations
by the wisconsin utraviolet photopolarimeter experiment—iii. the first ultraviolet lin-
ear spectropolarimetry of the moon. Mon. Not. R. Astron. Soc. 298, 303–309 (1998).
doi:10.1046/j.1365–8711.1998.01633.x
G.R. Gladstone et al., LAMP: the lyman alpha mapping project on nasa’s lunar reconnais-
sance orbiter mission. Space Sci. Rev. 150, 161–181 (2010). doi:10.1007/s11214-009-9578-
6
B. Hapke, Bidirectional reflectance spectroscopy. 1. theory. J. Geophys. Rev. 86, 3039–3054
(1981)
B. Hapke, Bidirectional reflectance spectroscopy. 3. correction for macroscopic roughness.
Icarus 59, 41–59 (1984)
B. Hapke, Bidrectional reflectance spectroscopy. 4. the extinction coefficient and the
opposition effect. Icarus 67, 264–280 (1986)
B. Hapke, Theory of reflectance and emitance spectroscopy. (Cambridge University Press,
Cambridge, 1993)
252 8. Ultraviolet Irradiance of the Moon
A.S. McEwen, E. Turtle, B.L. Jolliff, H. Heisinger, Lunar reconnaissance orbiter camera
(LROC) instrument overview, Space Sci. Rev. 150, 81–124 (2010). doi:10.1007/s11214-
010-9634-2
G. Rottman, Solar ultraviolet irradiance and its temporal variation. J. Atmos. Sol. Terr.
Phys. 61, 37–44 (1999). doi:10.1016/S1364-6826(98)00114-X
G. Rottman, The sorce mission. Sol. Phys. 230, 7–25 (2005). doi:10.1007/s11207-005-8112-6
M. Snow, W.E. McClintock, G. Rottman, T.N. Woods, Solar stellar irradiance comparison
experiment ii (solstice ii): examination of the solar-stellar comparison technique. Sol.
Phys. 230, 295–324 (2005). doi:10.1007/s11207-005-8763-3
M. Snow, G. Holsclaw, W.E. McClintock, T.N. Woods, Absolute ultraviolet irradiance of
the moon from sorce solstice. Proc. SPIE 6677, 66770D (2007). doi:10.1117/12.732498
T.C. Stone, H.H. Kieffer, Absolute irradiance of the moon for on-orbit calibration. Proc.
SPIE 4814, 211–221 (2002)
M. Taguchi, G. Funabashi, S. Watanabe, Y. Takahashi, H. Fukunishi, Lunar albedo at
hydrogen lyman α by the nozomi/uvs. Earth Planets & Space 52, 645–647 (2000)
J.K. Wagner, B.W. Hapke, E.N. Wells, Atlas of reflectance spectra of terrestrial, lu-
nar, and meteoric powders and frosts from 92 to 1800 nm. Icarus 69, 14–28 (1987).
doi:10.1016/0019-1035(87)90003-0
H.H. Wu, A.L. Broadfoot, The extreme ultraviolet albedos of the planet mercury and of the
moon. Geophys. Res. Lett. 82, 759–761 (1977). doi:10.1029/JB082i005p00759
—9—
Martin Snow
Laboratory for Atmospheric and Space Physics,
University of Colorado,
Boulder, CO, USA
Eric Quémerais
LATMOS-IPSL
Université Versailles-SaintQuentin, Guyancourt, France
Stéphane Ferron
ACRI-ST,
Guyancourt, France
Abstract
In November 2007, the Solar Wind Anisotropy (SWAN) instrument on the Solar
and Heliospheric Observatory (SOHO) and the SOLar-STellar Irradiance Compar-
ison Experiment (SOLSTICE) on the SOlar Radiation and Climate Experiment
(SORCE) both made observations of comet 17P/Holmes at Lyman-α wavelengths.
The intent of these observations was to compare the calibration of SWAN and
SOLSTICE, rather than to study the properties of comets. The two datasets re-
quire significant processing to remove interplanetary and geocoronal backgrounds
which increased the uncertainty of the result. However, both instruments measured
similar Lyman-α irradiances, confirming their calibration agreement.
Introduction
Comet 17P/Holmes brightened in a spectacular outburst on 23–24 October 2007
and remained unusually bright for several months. Comets can provide a localized
source of Lyman-α photons which are potentially easy to distinguish from geo-
coronal and interplanetary hydrogen sources. This paper uses nearly simultaneous
Lyman-α observations of this bright comet from two independent satellite instru-
ments to compare their calibrations. The Solar Wind Anisotropy (SWAN; Bertaux
et al. 1995) instrument on the Solar and Heliospheric Observatory (SOHO) maps
the ultraviolet (UV) emission from the entire sky on a daily basis. It has observed
Lyman-α emission from over 20 comets since launch on 2 December 1995 from
its location near the L1 Sun-Earth Lagrange point (Combi et al. 2011, 2012). The
SOLar-STellar Irradiance Comparison Experiment (SOLSTICE) instrument on the
Upper Atmosphere Research Satellite (UARS; Rottman et al. 1993) observed the
spectrum of comet Hale-Bopp (Woods et al. 2000), but did not measure its Lyman-
α irradiance due to the complications of the airglow. In this paper, we will ana-
lyze observations of comet Holmes from the SOLSTICE (McClintock et al. 2005a)
instrument on the SOlar Radiation and Climate Experiment (SORCE; Rottman
2005). The absolute calibration of SORCE SOLSTICE is also discussed in Snow
et al. (2012, this volume). The SORCE observations of comet Holmes used a differ-
ential technique to remove the geocoronal contribution to the comet signal. This
technique will be described in the section “SOLSTICE Comet Observations”.
In mid November 2007, there was a region about 0.6◦ in diameter around the
comet (larger than the Moon) which was bright in visible light. However, the
Lyman-α emitting region was likely much smaller than 0.6◦ since it had a fairly
modest gas production rate in November 2007 (Schleicher 2009).
The brightening of comet Holmes provided an opportunity to compare the cur-
rent SWAN Lyman-α calibration to the Lyman-α calibration of the SOLSTICE
on SORCE. SORCE was launched on 25 January 2003 into low Earth orbit. In
addition to observing the Sun, SOLSTICE is designed to use bright stars to track
changes in the instrument over time (Snow et al. 2005, 2012). Comet Holmes was
reported to be as bright as some of SOLSTICE’s calibration stars in November
2007, so observations were specially planned and executed.
Both SOLSTICE and SWAN must make significant corrections to the data
to account for the contribution from local or interplanetary hydrogen in order
to determine the emission from the comet alone. These corrections are highly
variable in time and depend on the detailed observation geometry. The uncertainty
introduced by these corrections is significant for the case of 17P/Holmes, but both
instruments did get a reliable measurement of the flux of Lyman-α photons coming
from the comet.
SWAN
SWAN provides regular Lyman-α maps of the sky on a 1◦ × 1◦ grid (Bertaux
et al. 1997). It uses twin 2-mirror periscopes mounted on opposite sides of the
spacecraft (+Z and −Z) to construct these maps. The detectors measure 5◦ by 5◦
SWAN 257
Figure 9.1: SWAN Lyman-α all-sky map from 1 November 2007. Comet Holmes
is visible as a bright patch near ecliptic latitude 30◦ and ecliptic longitude 60◦
maneuver on November 20 (marked by the vertical line). After the roll manuever, a
different detector was used for the (62,120) pixel measurement. The two detectors
have somewhat different sensitivity to starlight. On the left side of the vertical line,
the irradiance initially appears almost level until the comet begins drifting into the
1◦ × 1◦ field-of-view (FOV). On the right side of the vertical line, the irradiance
initially declines as the comet outburst fades away, then levels off.
One should note that after the roll maneuver, the irradiance measured by this
pixel remained very stable throughout November and December. The signal during
October had a larger variance, but was also relatively constant before the comet
entered the field of view. Therefore, the SWAN detectors show a stable value for
the irradiance of the background stars, although the level is different at each end
because of the two different detectors used. To estimate the comet brightness on
the cross-calibration days 22 and 24 November, we subtracted a straight-line fit to
the last 19 days of the observation period to obtain estimated comet brightnesses
for 22 November and 24 November of 146.2 R and 99.7 R, or irradiance values of
3,545 and 2,417 photons/s/cm2 /nm.
Figure 9.2: The signal from the SWAN pixel containing the comet as a function
of time (+ symbols) and after subtraction of the estimated Lyman-α background
(lower solid curve). The two diamonds after the spacecraft roll maneuver (vertical
line) are the two points used in the cross-calibration
which the comet was embedded. After the spacecraft roll maneuver on 20 Novem-
ber, over the course of several days, a diagonal group of pixels at one edge of the
band of stars went from low brightness levels to a higher steady level. A diagonal
group of pixels at the other edge of the band of stars went from a high initial level
to a lower level over the same period of time. This behavior can be understood as
a slight temporary shift in pixel positions with respect to the star field. The size
of the brightness shift is of the same order of magnitude as the comet brightness
found for the cross-calibration period.
SOLSTICE
SOLSTICE is a grating spectrometer whose primary mission is to measure
the ultraviolet solar spectral irradiance. It also observes the irradiance from an
ensemble of stable, bright, early-type stars in order to monitor changes in the
instrument response. As detailed in Snow et al. (2005, 2012), SOLSTICE can
make measurements over such a large dynamic range simply by changing apertures
and integration times. All other optical elements are used for both solar and stellar
observations.
Unlike SWAN, SOLSTICE is on a spacecraft in low Earth orbit. SORCE orbits
at an altitude of about 580 km with an orbital period of 97 min. The inclina-
tion of the orbit is 40◦ . This orbital geometry means that comet observations
at wavelengths near Lyman-α are contaminated by geocoronal airglow emissions
(Snow et al. 2005). In order to correct for contributions to the measured irradiance
from the airglow, comet observations are combined with observations of the nearby
260 9. Comet Holmes
sky background before and after each comet radiance measurement. This technique
is more fully described in Snow et al. (2005) and is the same observing strategy
used by SOLSTICE to measure stellar irradiances near Lyman-α on a routine ba-
sis. All the observations described here were conducted while the spacecraft was
in the eclipse portion of its orbit, so only the nightglow need be considered.
Figure 9.3: Typical SOLSTICE comet observation. In order to correct for geo-
coronal airglow emission, a region of the sky with no background stars is observed
before and after the comet observation. The top panel shows both the raw counts
of the background airglow and the comet and the polynomial model fit to the
background alone during the comet observation. The diamonds are observed back-
ground counts and the plus symbols indicate signal from the comet plus background.
The small circles in the top panel are the counts from the comet after subtracting
the background estimated from the model. The middle panel shows the observation
after the polynomial background has been removed. The bottom panel shows the
histogram of the final count rate measurement for the observation
on the zero order position derived from the stellar measurements before and after
the cometary observations to determine the absolute offset. Given the shape of
the airglow spectrum in Fig. 9.5, a 1.7 nm shift in grating angle produces a signal
that is reduced from the peak by a factor of 0.96. Therefore the observed count
rate must be divided by an additional factor of fλ = 0.96. The final count rate is
therefore
Craw − Cairglow
C= . (9.1)
DEG · fλ
The final count rate was then converted to irradiance using the SOLSTICE
measurement equation in McClintock et al. (2005b). Lyman-α irradiances for all
observations of comet Holmes are shown in Table 9.1.
262 9. Comet Holmes
Figure 9.4: Wavelength stability during the time period of the comet observations.
The position of the zero order offset for stellar observations throughout the day
changed very little in November 2007. The large offset between the stellar and
cometary datasets is due to the offset between the instrument boresight and the
peak of the background Lyman-α radiance
the SOLSTICE FUV boresight is offset by 20 arc minutes from the spacecraft z-
axis. It is this z-axis which is commanded to point at the ephemeris position of the
comet nucleus. The position angle of the comet in the instrument FOV will depend
on the roll angle of the spacecraft. For each observation, the spacecraft planning
system chooses a roll angle which minimizes obscuration of the star tracker’s FOV.
Therefore the roll angle may vary from one observation to the next. In the worst
case scenario, i.e. two roll angles separated by 180◦ , the comet’s wavelength scale
264 9. Comet Holmes
Figure 9.6: Time series of SOLSTICE Lyman-α irradiances from comet Holmes.
The error bars shown on the plot only include the statistical error for each obser-
vation. Systematic errors due to pointing are discussed in the text
could shift by up to 0.7 nm. This would change fλ by about 10 %. Based on the
preflight calibration, the absolute uncertainty of SOLSTICE at 121.6 nm is about
four percent (McClintock et al. 2005b), so the uncertainty in the fλ correction
factor is the dominant source of uncertainty in the SOLSTICE measurement.
Since these three sources of systematic error are all independent, they can be
added in quadrature to give a combined standard uncertainty of about 12 %, which
is consistent with the observed spread in the irradiance measurements shown in
Fig. 9.6.
Figure 9.7: SOLSTICE spectra of comet Holmes from 13 November 2007. The
top panel shows the average count rate from 30 co-added spectra after background
subtraction. The lower panel shows those counts converted to irradiance units
(photons/s/cm2 /nm). The signal-to-noise ratio was very low during these obser-
vations, so the uncertainty in the derived irradiance is significant
there is a weak signal from the comet. The top panel shows the residual count rate
after background subtraction. The signal-to-noise ratio is only about 1 for each
spectrum, so even co-adding 30 observations yields an irradiance with about a 30 %
uncertainty. Integrating over the entire FUV band gives a contribution of about
1,300 photons/s/cm2 /nm to the signal measured by SWAN.
If the comet did not change UV brightness over the 9 days from 13 November
to 22 November, then the rest of the FUV band would contribute about a third of
the spectral irradiance measured by SWAN. However, based on Fig. 9.2, the total
irradiance in the SWAN band decreased by about a factor of two over those 9 days.
Therefore the non-Lyman-α contribution to the comet flux on 22 November is
probably also half as much as it was on 13 November when the SOLSTICE spectra
were acquired. Therefore the FUV contribution to the SWAN measurement on 22
November was probably no more than 15 % of the total signal. If so, the corrected
SWAN value on 22 November would be about 2,800 photons/s/cm2 /nm (Table 9.2).
These observations of Lyman-α emission from a comet by both SWAN and
SOLSTICE confirm their similar calibration. As discussed in Snow et al. (2012,
this volume), the SOLSTICE calibration is in good agreement with observations
from the International Ultraviolet Explorer based on the white dwarf flux scale
(Bohlin et al. 1990; Bohlin 1996). This white dwarf scale is also used to calibrate
266 9. Comet Holmes
the instruments on the HST, and the inflight calibration of SWAN used the God-
dard Hight Resolution Spectrograph on HST (Quémerais and Bertaux 2002). So
agreement between SOLSTICE and SWAN is to be expected.
In summary, the comparison technique of using a target of opportunity such as
a comet was moderately successful. However, due to the relatively low signal levels
in these observations, the uncertainty in the Lyman-α measurement is significant.
The agreement between SWAN and SOLSTICE is not tightly constrained by these
observations due to the large uncertainty in the SWAN measurement caused by
the contribution of the background stars. Comet Holmes, a rare naked-eye comet,
apparently was not a particularly strong Lyman-α source. Two other comets were
easier to see in the SWAN data from this period. This may indicate a larger role for
dust than water in the 2007 comet Holmes outburst. Other comets, not necessarily
naked-eye comets, might actually be better targets for cross-calibration.
Unfortunately, no follow-up observations will be possible with SORCE SOL-
STICE. Spacecraft systems on SORCE, particularly the batteries, have forced op-
erations during the eclipse portion of the orbit to be curtailed. Starting in late
2010, only a few of the standard stellar calibration observations can be scheduled,
and only during orbits with eclipse durations less than 30 min.
Bibliography
J-L. Bertaux et al., SWAN: a study of solar wind anisotropies on SOHO with Lyman-alpha
sky mapping. Solar Phys. 162, 403 (1995) doi:10.1007/BF00733435
J-L. Bertaux et al., First results from SWAN Lyman alpha solar wind mapper on SOHO.
Solar Phys. 175, 737–770 (1997) doi:10.1023/A:1004979605559
R.C. Bohlin, Spectrophotometric standards from the far-uv to the near-ir on the white dwarf
flux scale. Astrophys. J. 111, 1743–1747 (1996) doi:10.1086/117914
R.C. Bohlin, A.W. Harris, A.V. Holm, C. Gry, The ultraviolet calibration of the hubble space
telescope iv. absolute IUE fluxes of Hubble Space Telescope standard stars. Astrophys.
J. Suppl. 73, 413–439 (1990) doi:10.1086/191474
M.R. Combi, Y. Lee, T.S. Patel, J.T.T. Mäkinen, J-L. Bertaux, E. Quémerais,
SOHO/SWAN observations of short-period spacecraft target comets. Astrophys. J. 141,
128 (2011) doi:10.1088/0004-6256/141/4/128
M.R. Combi, Z. Boyd, Y. Lee, T.S. Patel, J-L. Bertaux, E. Quémerais, J.T.T. Mäkinen,
SOHO/SWAN observations of comets with small perihelia: C/2002 V1 (NEAT), C/2002
X5 (Kudo-Fujikawa), 2006 P1 (McNaught) and 96P/Macholtz 1. Icarus 216, 449–461
(2012) doi:10.1016/j.icarus.2011.09.019
Bibliography 267
W.E. McClintock, G. Rottman, T.N. Woods, Solar stellar irradiance comparison experiment
ii (SOLSTICE II): instrument concept and design, Solar Phys. 230, 225–258 (2005a)
doi:10.1007/s11207-005-7432-x
W.E. McClintock, M. Snow, T.N Woods, Solar stellar irradiance comparison
experiment ii (SOLSTICE II): pre-launch and on-orbit calibrations. Solar Phys. 230,
259–294 (2005b) doi:10.1007/s11207-005-1585-5
E. Quémerais, J-L. Bertaux, Radiometric calibration of the SWAN instrument. In The
Radiometric Calibration of SOHO, ed. by A. Pauluhn, M.C.E. Huber, R. von Steiger.
(ISSI Scientific Series, SR-002, 2002)
G. Rottman, The SORCE mission. Sol. Phys. 230, 7–25 (2005) doi:10.1007/s11207-005-
8112-6
G.Rottman, T.N. Woods, T. Sparn, Solar stellar irradiance comparison experi-
ment 1. i—instrument design and operations. J. Geophys. Res. 98, 10667 (1993)
doi:10.1029/93JD00462
D.G. Schleicher, The long-term decay in production rates following the extreme out-
burst of comet 17P/Holmes. Astrophys. J. 138, 1062–1071 (2009) doi:10.1088/0004-
6256/138/4/1062
M. Snow, W.E. McClintock, G. Rottman, T.N. Woods, Solar stellar irradiance comparison
experiment ii (SOLSTICE II): examination of the solar-stellar comparison technique. Sol.
Phys. 230, 295–324 (2005) doi:10.1007/s11207-005-8763-3
M. Snow, A. Reberac, E. Quémerais, J. Clarke, W.E. McClintock, T.N. Woods, A new
catalog of ultraviolet spectra. In Cross-Calibration of Past and Present Far UV Spectra of
Solar System Objects and the Heliosphere, ed. by E. Quémerais, M. Snow, R.M. Bonnet.
(ISSI Scientific Series, SR-013, 2012) (this volume)
T.N. Woods, P.D. Feldman, G.J. Rottman, Ultraviolet observations of comet Hale-
Bopp (C/1995 O1) by the UARS SOLSTICE. Icarus 144, 182–186 (2000)
doi:10.1066/icar.1999.6262
Index
Heliopause, 14, 16, 18, 19, 21, 22, 25, 164, 165, 167, 172, 177,
26, 28, 30, 57, 70, 80, 144 179, 180, 182–184
Heliosheath, 9, 10, 14–21, 25–28, 70, Interstellar magnetic field (IsMF), 9,
71, 82, 143 14, 21–24, 57, 58, 142
Heliotail, 58 Interstellar medium (ISM), 15, 16,
High-energy, 9 21, 24, 26, 148, 177
Hopkins Ultraviolet Telescope IP. See Interplanetary (IP)
(HUT), 228 IPH. See Interplanetary hydrogen
Hot-type, 28, 58, 59 (IPH)
Hubble Space Telescope (HST) IPM. See Interplanetary medium
Goddard High Resolution (IPM)
Spectrograph (GHRS), 145, IPS. See Interplanetary scintillation
153–155, 169, 257 (IPS)
Space Telescope Imaging ISM. See Interstellar medium (ISM)
Spectrograph (STIS), 2, IsMF. See Interstellar magnetic field
145, 153, 155 (IsMF)
HUT. See Hopkins Ultraviolet Isothermal, 146
Telescope (HUT) Isotropic, 9, 12, 24, 25, 96, 155–158,
167
IMP-8. See Interplanetary IUE. See International Ultraviolet
Monitoring Platform Explorer (IUE)
(IMP-8)
Inflow, 68, 85, 103, 106, 108 Joint Polar Satellite System (JPSS),
Intercalibration, 71, 98, 109, 250
126 JPSS. See Joint Polar Satellite
International Ultraviolet Explorer System (JPSS)
(IUE), 2, 3, 179, 192, Jupiter, 107, 108
201–203, 205, 206, 210,
212–215, 229, 265 Kinetic, 1, 8, 18, 19, 24, 25, 27,
Interplanetary (IP), 1, 2, 8–10, 12, 29–36, 38, 40–44, 46, 51,
14, 24, 58, 59, 67–127, 52, 55–57, 59, 80, 93, 94
141–159, 167, 168, 170, 178, Kinetic-continuum, 57, 58
257, 260 Kinetic-gasdynamic, 23
Interplanetary hydrogen (IPH), 8, Kinetic-MHD, 23, 27, 28
20, 149, 151, 152, 154–158,
164, 169, 170, 256, 260 LASP Interactive Solar IRradiance
Interplanetary medium (IPM), 1, 75, Datacenter (LISIRD), 127
142, 148, 177–186 LASP Lunar Albedo Measurement
Interplanetary Monitoring Platform and Analysis from
(IMP-8), 98 SOLSTICE (LLAMAS),
Interplanetary scintillation (IPS), 227–251
102, 103, 105, 107, 111–117, LIC. See Local interstellar cloud
123, 125–126 (LIC)
Interstellar, 1, 7–59, 68–73, 77–83, Lightcurves, 106
85, 86, 93, 96, 97, 100, 101, Limb, 116
103, 104, 108, 109, 126, Line-center, 8, 45, 76–78, 148, 152,
142, 143, 145, 148, 156, 156, 159, 165, 178, 179, 185
272 Index
Spectrograph, 145, 164, 166, 179, eta Aur, 211, 212, 220
180, 205, 266 eta UMa, 194, 201, 211–213,
Spectrometer, 1–3, 142, 144, 145, 215, 220
151, 152, 159, 164, 166, kappa Vel, 194, 211, 220, 224
169, 172, 192, 193, 205, sigma Sgr, 194, 211, 221, 225
229, 231, 232, 234, 240, tau Sco, 194, 211, 221, 225
246, 257, 259, 260 zeta Cas, 211, 212, 221, 225
Spectroscopy for Investigation of zeta Oph, 211, 212, 222, 225
Characteristics of the zeta Pup, 209–212, 222, 224
Atmosphere of Mars Strahl, 94
(SPICAM), 2, 3, 159, 192, Streamlines, 22, 23, 97
203–215, 224 Student Nitric Oxide Explorer
Spectroscopy for Investigation of (SNOE), 229
Characteristics of the Subsonic, 9, 16, 70
Atmosphere of Venus Sunspot, 72, 89
(SPICAV), 159 Suprathermal, 27
SPICAM. See Spectroscopy for Synchrotron Ultraviolet Radiation
Investigation of Facility (SURF), 196–198,
Characteristics of the 201, 246, 248
Atmosphere of Mars
(SPICAM)
SPICAV. See Spectroscopy for Temperature, 11, 12, 16–18, 22,
Investigation of 30–32, 35, 36, 38, 40–46,
Characteristics of the 51–57, 70, 71, 93–96, 98,
Atmosphere of Venus 99, 126, 146, 164, 178, 192,
(SPICAV) 196, 242, 246
SSBUV. See Shuttle Solar Thermodynamics, 24, 25
Backscatter Ultraviolet Thermoelectric, 183
(SSBUV) Thermosphere Ionosphere
Stars Mesosphere Energetics and
alpha CMa, 194, 211, 216, Dynamics (TIMED), 92,
224 165
alpha Cru, 194, 199, 211, 216 Solar EUV Explorer (SEE), 73,
alpha Gru, 194, 211, 216, 225 88–90
alpha Leo, 194, 201, 211–213, Three-component, 73, 94, 143
217, 224 Three-dimensional, 21, 31, 36, 86,
alpha Lyr, 194, 201, 203, 155, 156, 158
211–213, 217, 225 Three-Gaussian, 76
alpha Pav, 194, 211, 217, 225 TIMED. See Thermosphere
alpha Vir, 194, 209, 211, 218, Ionosphere Mesosphere
224 Energetics and Dynamics
beta Cen, 194, 201, 211, 212, (TIMED)
218, 225 Tomography, 103, 112–117
beta Lup, 225 Total and Spectral Irradiance Sensor
delta Cen, 194, 211, 219, 224 (TSIS), 250
delta Sco, 194, 211, 219, 225 Two-component, 15, 33, 35, 47,
epsilon Per, 194, 211, 219, 224 52–57, 73, 143
Index 275