This action might not be possible to undo. Are you sure you want to continue?
C:l:«tíal ¸atíçatíou
Copyr¡ght © 19972009 Henn¡ng 1mIand
IermIssIon Is granted to copy, dIstrIbute and/or modIIy thIs document under the terms oI the
GÞ! Iree ÐocumentatIon IIcense, VersIon 1.3 or any Iater versIon pubIIshed by the Iree
SoItware IoundatIon, wIth no ¡nvarIant SectIons, no IrontCover Texts and no ÐackCover
Texts. A copy oI the IIcense Is IncIuded In the sectIon entItIed "GÞ! Iree ÐocumentatIon
IIcense".
Bev¡sed December 1
st
, 2009
1¡rst PubI¡shed May 20
th
, 1997
Index
Preface
Chapter 1 The Basics oI Celestial Navigation
Chapter 2 Altitude Measurement
Chapter 3 Geographic Position and Time
Chapter 4 Finding One's Position (Sight Reduction)
Chapter 5 Finding the Position oI a Moving Vessel
Chapter 6
Determination oI Latitude and Longitude, Direct Calculation oI
Position
Chapter 7 Finding Time Longitude by Lunar Distances
Chapter 8 Rise, Set, Twilight
Chapter 9 Geodetic Aspects oI Celestial Navigation
Chapter 10 Spherical Trigonometry
Chapter 11 The Navigational Triangle
Chapter 12 General Formulas Ior Navigation
Chapter 13 Charts and Plotting Sheets
Chapter 14 Magnetic Declination
Chapter 15 Ephemerides oI the Sun
Chapter 16 Navigational Errors
Chapter 17 The Marine Chronometer
Appendix
GNU Free
Documentation
License
Felix qui potuit boni
fontem visere lucidum,
felix qui potuit gravis
terrae solvere vincula.
Boethius
Why should anybody still practice celestial navigation in the era oI electronics and GPS? One might as well ask why
some photographers still develop blackandwhite photos in their darkroom instead oI using a digital camera. The
answer would be the same: because it is a noble art, and because it is rewarding. No doubt, a GPS navigator is a
powerIul tool, but using it becomes routine very soon. In contrast, celestial navigation is an intellectual challenge.
Finding your geographic position by means oI astronomical observations requires knowledge, judgement, and
skillIulness. In other words, you have to use your brains. Everyone who ever reduced a sight knows the thrill I am
talking about. The way is the goal.
It took centuries and generations oI navigators, astronomers, geographers, mathematicians, and instrument makers to
develop the art and science oI celestial navigation to its present level, and the knowledge thus accumulated is a treasure
that should be preserved. Moreover, celestial navigation gives an impression oI scientiIic thinking and creativeness in
the preelectronic age. Last but not least, celestial navigation may be a highly appreciated alternative iI a GPS receiver
happens to Iail.
When I read my Iirst book on navigation many years ago, the chapter on celestial navigation with its Iascinating
diagrams and Iormulas immediately caught my particular interest although I was a little deterred by its complexity at
Iirst. As I became more advanced, I realized that celestial navigation is not nearly as diIIicult as it seems to be at Iirst
glance. Studying the literature, I Iound that many books, although packed with inIormation, are more conIusing than
enlightening, probably because most oI them have been written by experts and Ior experts. Other publications are
designed like cookbooks, i. e., they contain stepbystep instructions but avoid much oI the theory. In my opinion, one
can not reallv comprehend celestial navigation and enjoy the beauty oI it without knowing the mathematical
background.
Since nothing really complied with my needs, I decided to write a compact manual Ior my personal use which had to
include the most important deIinitions, Iormulas, diagrams, and procedures. The idea to publish it came in 1997 when I
became interested in the internet and Iound that it is the ideal medium to share one's knowledge with others. I took my
manuscript, rewrote it in the HTML Iormat, and published it on my web site. Later, I converted everything to the PDF
Iormat, which is an established standard Ior electronic publishing now.
The style oI my work may diIIer Irom standard books on this subject. This is probably due to my diIIerent perspective.
When I started the project, I was a newcomer to the world oI navigation, but I had a background in natural sciences and
in scientiIic writing. From the very beginning, it has been my goal to provide accurate inIormation in a highly
structured and comprehensible Iorm. The reader may judge whether this attempt has been successIul.
More people than I ever expected are interested in celestial navigation, and I would like to thank readers Irom all over
the world Ior their encouraging comments and suggestions. However, due to the increasing volume oI correspondence,
I am no longer able to answer every individual question or to provide individual support. UnIortunately, I have still a
Iew other things to do, e. g., working Ior my living. Nonetheless, I keep working on this publication at leisure.
This publication is released under the terms and conditions oI the GNU Free Documentation License. A copy oI the
latter is included.
December 1
st
, 2009
Henning Umland
Correspondence address:
Dr. Henning Umland
Rabenhorst 6
21244 Buchholz i. d. N.
Germany
Fax: ¹49 3212 1197261
Chapter 1
The Basics of Celestial Navigation
Celestial navigation, also called astronomical navigation, is the art and science oI Iinding one's geographic position
through astronomical observations, particularly by measuring altitudes oI celestial bodies sun, moon, planets, or
stars.
An observer watching the night sky without knowing anything about geography and astronomy might spontaneously
get the impression oI being on a horizontal plane located at the center oI a huge hollow sphere with the celestial bodies
attached to its inner surIace. This naive concept oI a spherical universe has probably been in use since the beginning oI
mankind. Later, astronomers oI the ancient world (Ptolemy et al) developed it to a high degree oI perIection. Still
today, spherical astronomy is Iundamental to celestial navigation since the navigator, like the astronomers oI old,
measures apparent positions oI bodies in the sky without knowing their absolute positions in space.
Accordingly, the apparent position oI a body in the sky is best characterized by a spherical coordinate system, in this
special case by the horizon system of coordinates. In this system, an imaginary observer is located at the center oI the
celestial sphere, a hollow sphere oI inIinite diameter, which is divided into two hemispheres by the plane oI the
celestial horizon (Fig. 11). The center oI the celestial sphere coincides with the center oI the earth which is also
assumed to be a sphere. The Iirst coordinate oI the observed body is the geocentric altitude, H. H is the vertical angle
between the celestial horizon and a straight line Irom the center oI the celestial sphere to the body. H is measured Irom
0° through ¹90° above the horizon and Irom 0° through 90° below the horizon. The geocentric zenith distance, z, is
the corresponding angular distance between the body and the zenith, an imaginary point vertically overhead. The zenith
distance is measured Irom 0° through 180°. H and z are complementary angles (H ¹ z ÷ 90°). The point opposite to
the zenith, deIined by the direction oI gravity, is called nadir (H ÷ 90°, z ÷ 180°). H and z are also arcs oI the vertical
circle going through zenith, nadir, and the observed body. The second coordinate oI the body, the geocentric true
azimuth, Az
N
, is the horizontal direction oI the body with respect to the geographic north point on the celestial
horizon, measured clockwise Irom 0°(N) through 360°. The third coordinate, the distance oI the body Irom the center
oI the celestial sphere, remains unknown.
In reality, the observer is not located on the plane oI the celestial horizon but on or above the surIace oI the earth. The
horizontal plane passing through the observer's eye is called sensible horizon (Fig. 12).
The latter merges into the geoidal horizon, a plane tangent to the earth at the observer's position, when the observer's
eye is at sea level. The three planes oI horizon are parallel to each other and perpendicular to the direction oI gravity at
the observer's position.
11
Since sensible and geoidal horizon are relatively close to each other (compared with the radius oI the earth), they can be
considered as identical under most practical conditions. None oI the above Iictitious horizons coincides with the visible
horizon, the line where the earth's surIace and the sky appear to meet.
Usually, the trigonometric calculations oI celestial navigation are based on the geocentric altitudes (or geocentric
zenith distances) oI bodies. Since it is not possible to measure the geocentric altitude oI a body directly, it has to be
derived Irom the altitude with respect to the visible or sensible horizon (altitude corrections, chapter 2).
The altitude oI a body with respect to the visible sea horizon is usually measured with a marine sextant. Measuring
the altitude with respect to the (invisible) sensible horizon requires an instrument with an artificial horizon, e. g., a
theodolite (chapter 2). An artiIicial horizon is a device, e. g., a pendulum, that indicates the local direction oI gravity.
Geocentric altitude and zenith distance oI a celestial body depend on the distance between the terrestrial observer and
the geographic position of the body, GP. GP is the point where a straight line Irom the center oI the earth, C, to the
celestial body intersects the earth's surIace (Fig. 13).
A body appears in the zenith (H ÷ 90°, z ÷ 0°) when GP is identical with the observer's position. A terrestrial (earth
bound) observer moving away Irom GP will experience that the geocentric zenith distanze oI the body varies in direct
proportion with his growing distance Irom GP. The geocentric altitude oI the body decreases accordingly. The body is
on the celestial horizon (H ÷ 0°, z ÷ 90°) when the observer is one quarter oI the circumIerence oI the earth away Irom
GP. II the observer moves Iurther away Irom GP, the body becomes invisible.
For any given altitude oI a body, there is an inIinite number oI terrestrial positions having the same distance Irom GP
and thus Iorming a circle on the earth's surIace (Fig 14). The center oI this circle is on the line CGP, below the earth's
surIace. An observer traveling along the circle will measure a constant altitude (and zenith distance) oI the body,
irrespective oI his position on the circle. ThereIore, such a circle is called a circle of equal altitude.
The arc length r, the radial distance oI the observer Irom GP measured along the surIace oI the earth, is obtained
through the Iollowing Iormula:
One nautical mile (1 nm ÷ 1.852 km) is the great circle distance (chapter 3) oI one minute oI arc on the surIace oI the
earth. The mean perimeter oI the earth is 40031.6 km.
12
r nm 60: ° or r km
Perimeter of Earth km
360°
: °
As shown in Fig. 14, light rays originating Irom a distant object (Iixed star) are virtually parallel to each other when
they arrive at the earth. ThereIore, the altitude oI such an object with respect to the geoidal (or sensible) horizon, called
topocentric altitude, equals the geocentric altitude. In contrast, light rays coming Irom a relatively close body (moon,
sun, planets) diverge signiIicantly. This results in a measurable diIIerence between topocentric and geocentric altitude,
called parallax. The eIIect is greatest when observing the moon, the body closest to the earth (chapter 2).
The true azimuth oI a body depends on the observer's position on the circle oI equal altitude and can assume any value
between 0° and 360°. Usually, the navigator is not equipped to measure the azimuth oI a body with the same precision
as the altitude. However, there are methods to calculate the azimuth Irom other quantities.
Whenever we measure the altitude or zenith distance oI a celestial body, we have already gained some inIormation
about our own geographic position because we know we are somewhere on a circle oI equal altitude deIined by the
center, GP (the geographic position oI the body), and the radius, r. OI course, the inIormation available so Iar is still
incomplete because we could be anywhere on the circle oI equal altitude which comprises an inIinite number oI
possible positions and is thereIore also reIerred to as a circle of position (chapter 4).
We extend our thought experiment and observe a second body in addition to the Iirst one. Logically, we are on two
circles oI equal altitude now. Both circles overlap, intersecting each other at two points on the earth's surIace. One oI
these two points oI intersection is our own position (Fig. 15a). Theoretically, both circles could be tangent to each
other. This case, however, is unlikely. Moreover, it is undesirable and has to be avoided (chapter 16).
In principle, it is not possible Ior the observer to know which point oI intersection Pos. 1 or Pos. 2 is identical with
his actual position unless he has additional inIormation, e. g., a Iair estimate oI his position, or the compass bearing
(approximate azimuth) oI at least one oI the bodies. Solving the problem oI ambiguity can also be achieved by
observation oI a third body because there is only one point where all three circles oI equal altitude intersect (Fig. 15b).
Theoretically, the observer could Iind his position by plotting the circles oI equal altitude on a globe. Indeed, this
method has been used in the past but turned out to be impractical because precise measurements require a very big
globe. Plotting circles oI equal altitude on a map is possible iI their radii are small enough. This usually requires
observed altitudes oI almost 90°. The method is rarely used since such altitudes are not easy to measure. Usually,
circles oI equal altitude have diameters oI several thousand nautical miles and do not Iit on nautical charts. Further,
plotting circles oI such dimensions is very diIIicult due to geometric distortions caused by the respective map projection
(chapter 13).
Usually, the navigator has at least a rough idea oI his position. It is thereIore not required to plot a complete circle oI
equal altitude. In most cases only a short arc oI the circle in the vicinity oI the observer's estimated position is oI
interest. II the curvature oI the arc is negligible, depending on the radius oI the circle and the map scale, it is possible to
plot a straight line (a secant or a tangent oI the circle oI equal altitude) instead oI the arc. Such a line is called a line of
position or position line.
In the 19
th
century, navigators developed very convenient mathematical and graphic methods Ior the construction oI
position lines on nautical charts. The point oI intersection oI at least two suitable position lines marks the observer's
position. These methods, which are considered as the beginning oI modern celestial navigation, will be explained in
detail later.
In summary, Iinding one's geographic position by astronomical observations includes three basic steps:
1. Measuring the altitudes or zenith distances of two or more celestial bodies (chapter 2).
2. Finding the geographic position of each body at the instant of its observation (chapter 3).
3. Deriving one's own position from the above data (chapter 4&5).
13
Chapter 2
Altitude Measurement
Although altitudes and zenith distances are equally suitable Ior navigational calculations, most Iormulas are
traditionally based upon altitudes since these are easily accessible using the visible sea horizon as a natural reIerence
line. Direct measurement oI the zenith distance, however, requires an instrument with an artiIicial horizon, e. g., a
pendulum or spirit level indicating the local direction oI gravity (perpendicular to the sensible horizon), since a suitable
reIerence point in the sky does not exist.
Instruments
A marine sextant consists oI a system oI two mirrors and a telescope mounted on a metal Irame (brass or aluminum).
A schematic illustration (side view) is given in Fig. 21. The horizon glass is a halIsilvered mirror whose plane is
perpendicular to the plane oI the Irame. The index mirror, the plane oI which is also perpendicular to the Irame, is
mounted on the socalled index arm rotatable on a pivot perpendicular to the Irame. The optical axis oI the telescope is
parallel to the Irame. When measuring an altitude, the instrument Irame is held in an upright position, and the visible
sea horizon is sighted through the telescope and horizon glass. A light ray coming Irom the observed body is Iirst
reIlected by the index mirror and then by the back surIace oI the horizon glass beIore entering the telescope. By slowly
rotating the index mirror on the pivot the superimposed image oI the body is aligned with the image oI the horizon line.
The corresponding altitude, which is twice the angle Iormed by the planes oI horizon glass and index mirror, can be
read Irom the graduated limb, the lower, arcshaped part oI the triangular sextant Irame (Fig. 22). Detailed inIormation
on design, usage, and maintenance oI sextants is given in 3 (see appendix).
Fíg. 22
On land, where the horizon is too irregular to be used as a reIerence line, altitudes have to be measured by means oI
instruments with an artificial horizon.
21
A bubble attachment is a special sextant telescope containing an internal artificial horizon in the Iorm oI a small
spirit level whose image, replacing the visible horizon, is superimposed with the image oI the body. Bubble attachments
are expensive (almost the price oI a sextant) and not very accurate because they require the sextant to be held absolutely
still during an observation, which is diIIicult to manage. A sextant equipped with a bubble attachment is reIerred to as a
bubble sextant. Special bubble sextants were used Ior air navigation beIore electronic navigation systems became
standard equipment.
A pan Iilled with water or, preIerably, a more viscous liquid, e. g., glycerol, can be utilized as an external artificial
horizon. As a result oI gravity, the surIace oI the liquid Iorms a perIectly horizontal mirror unless distorted by
vibrations or wind. The vertical angular distance between a body and its mirror image, measured with a marine sextant,
is twice the altitude. This very accurate method is the perIect choice Ior exercising celestial navigation in a backyard.
Fig. 23 shows a proIessional Iorm oI an external artiIicial horizon Ior land navigation. It consists oI a horizontal mirror
(polished black glass) attached to a metal Irame with three leg screws. Prior to an observation, the screws have to be
adjusted with the aid oI one or two detachable highprecision spirit levels until the mirror is exactly horizontal in every
direction.
Fíg. 23
Fíg. 24
A theodolite (Fig. 24) is basically a telescopic sight which can be rotated about a vertical and a horizontal axis. The
angle oI elevation (altitude) is read Irom the graduated vertical circle, the horizontal direction is read Irom the
horizontal circle. The specimen shown above has vernier scales and is accurate to approx. 1'.
22
The vertical axis oI the instrument is aligned with the direction oI gravity by means oI a spirit level (artiIicial horizon)
beIore starting the observations. Theodolites are primarily used Ior surveying, but they are excellent navigation
instruments as well. Some models can resolve angles smaller than 0.1' which is not achieved even with the best
sextants. A theodolite is mounted on a tripod and has to stand on solid ground. ThereIore, it is restricted to land
navigation. Mechanical theodolites traditionally measure zenith distances. Electronic models can optionally measure
altitudes. Some theodolites measure angles in the unit gon instead oI degree (400 gon ÷ 360°).
BeIore viewing the sun through an optical instrument, a proper shade glass must be inserted, otherwise the retina might
suIIer permanent damage!
Altitude corrections
Any altitude measured with a sextant or theodolite contains errors. Altitude corrections are necessary to
eliminate systematic altitude errors AND to reduce the topocentric altitude of a body to the geocentric altitude
(chapter 1). Altitude corrections do NOT remove random observation errors.
Index error (IE)
A sextant or theodolite may display a constant error (index error, IE) which has to be subtracted Irom the reading
beIore it can be used Ior Iurther calculations. The error is positive iI the angle displayed by the instrument is greater
than the actual angle and negative iI the displayed angle is smaller. Errors which vary with the displayed angle require
the use oI an individual correction table iI the error can not be eliminted by overhauling the instrument.
The sextant altitude, Hs, is the altitude as indicated by the sextant beIore any corrections have been applied.
When using an external artiIicial horizon, H
1
(not Hs!) has to be divided by two.
A theodolite measuring the zenith distance, z, requires the Iollowing Iormula to obtain H
1
:
Dip of horizon
II the earth's surIace were an inIinite plane, visible and sensible horizon would be identical. In reality, the visible
horizon appears several arcminutes below the sensible horizon which is the result oI two contrary eIIects, the curvature
oI the earth's surIace and atmospheric reIraction. The geometrical horizon, the surIace oI a Ilat cone, is Iormed by an
inIinite number oI straight lines tangent to the earth and radiating Irom the observer's eye. Since atmospheric reIraction
bends light rays passing along the earth's surIace toward the earth, all points on the geometric horizon appear to be
elevated, and thus Iorm the visible horizon. II the earth had no atmosphere, the visible horizon would coincide with the
geometrical horizon (Fig. 25).
The vertical angular distance oI the sensible horizon Irom the visible horizon is called dip and is a Iunction oI the
height of eye, HE, the vertical distance oI the observer's eye Irom the sea surIace (the distance between sensible and
geoidal horizon):
23
1st correction: H
1
HsIE
H
1
90° :IE
Dip´ 1.76HE m 0.97HE ft
The above Iormula is empirical and includes the eIIects oI the curvature oI the earth's surIace and oI atmospheric
reIraction*.
*At sea, the dip oI horizon can be obtained directly by measuring the angular distance between the visible horizon in Iront oI the observer and the
visible horizon behind the observer through the zenith. Subtracting 180° Irom the angle thus measured and dividing the resulting angle by two
yields the dip oI horizon. This very accurate method can not be accomplished with a sextant but requires a special instrument (prismatic reIlecting
circle) which is able to measure angles greater than 180°.
The correction for the dip of horizon has to be omitted (Dip ÷ 0) if any kind of an artificial horizon is used since
the latter is solely controlled by gravity and thus indicates the plane of the sensible horizon (perpendicular to the
vector of gravity).
The altitude obtained aIter applying corrections Ior index error and dip is also reIerred to as apparent altitude, Ha.
Atmospheric refraction
A light ray coming Irom a celestial body is slightly deIlected toward the earth when passing obliquely through the
atmosphere. This phenomenon is called refraction, and occurs always when light enters matter oI diIIerent density at
an angle smaller than 90°. Since the eye is not able to detect the curvature oI the light ray, the body appears to be on a
straight line tangent to the light ray at the observer's eye, and thus appears to be higher in the sky. R is the vertical
angular distance between apparent and true position oI the body measured at the observer's eye (Fig. 26).
Atmospheric reIraction is a Iunction oI Ha (÷ H
2
). Atmospheric standard refraction, R
0
, is 0' at 90° altitude and
increases progressively to approx. 34' as the apparent altitude approaches 0°:
Ha ¡°] 0 1 2 5 10 20 30 40 50 60 70 80 90
R
0
¡']
~34 ~24 ~18 9.9 5.3 2.6 1.7 1.2 0.8 0.6 0.4 0.2 0.0
The navigator has the choice between several Iormulas to calculate R
0
. Smart´s Iormula gives highly accurate results
Irom 15° through 90° apparent altitude 2,9:
For the purpose oI navigation, Smart´s Iormula is still accurate enough at 10° apparent altitude. Below 5°, the error
increases progressively.
For altitudes between 0° and 15°, the Iollowing Iormula is recommended 10. H
2
is measured in degrees:
24
R
0
´
34.1334.197H
2
0.00428H
2
2
10.505H
2
0.0845H
2
2
2nd correction: H
2
H
1
Dip
Ha H
2
R
0
´
0.97127
tan H
2
°
0.00137
tan
3
H
2
°
A lowprecision reIraction Iormula including the whole range oI altitudes Irom 0° through 90° was Iound by Bennett:
The accuracy is suIIicient Ior navigation. The maximum systematic error, occurring at 12° altitude, is approx. 0.07' 2.
II necessary, Bennett´s Iormula can be improved (max. error: 0.015') by the Iollowing correction:
The argument oI the sine is stated in degrees (although R
0
is measured in arcminutes) 2.
Atmospheric reIraction is inIluenced by atmospheric pressure and air temperature. ThereIore the standard reIraction,
R
0
, has to be multiplied with a correction Iactor, I, to obtain the reIraction Ior a given combination oI pressure and
temperature iI higher precision is required.
P is the atmospheric pressure and T the air temperature. Standard conditions (I ÷ 1) are 1010 hPa (29.83 in) and 10°C
(50°F). The eIIects oI air humidity are comparatively small and can be ignored. The correction Ior pressure and
temperature is sometimes omitted (I ÷ 1) since the resulting error is usually small.
The common reIraction Iormulas reIer to a Iictitious standard atmosphere with an average density gradient. The actual
reIraction may diIIer Irom the calculated one iI anomalous atmospheric conditions are present (temperature inversion,
mirage eIIects, etc.). The inIluence oI atmospheric anomalies increases strongly with decreasing altitude. ThereIore,
reIraction at altitudes below ca. 5° may become erratic, and calculated values in this range are not always reliable. It
should be mentioned that the dip oI horizon, too, is inIluenced by atmospheric reIraction and may become
unpredictable under certain meteorological conditions.
H
3
represents the topocentric altitude of the body, the altitude with respect to the sensible horizon.
Parallax
The trigonometric calculations oI celestial navigation are based upon geocentric altitudes. Fig. 27 illustrates that the
geocentric altitude oI an object, H
4
, is always greater than the topocentric altitude, H
3
. The diIIerence H
4
H
3
is called
parallax in altitude, P. P decreases as the distance between object and earth increases. Accordingly, the eIIect is
greatest when observing the moon since the latter is the object nearest to the earth. On the other hand, P is too small to
be measured when observing Iixed stars (see chapter 1, Fig. 14). Theoretically, the observed parallax reIers to the
sensible, not to the geoidal horizon. However, since the height oI eye is by several magnitudes smaller than the radius
oI the earth, the resulting error is usually not signiIicant.
The parallax (in altitude) oI a body being on the geoidal horizon is called horizontal parallax, HP (Fig. 27). The
horizontal parallax oI the sun is approx. 0.15'. Current values Ior the HP oI the moon (approx. 1°!) and the navigational
planets are given in the Nautical Almanac 12 and similar publications, e.g., 13*.
*Tabulated values Ior HP reIer to the equatorial radius oI the earth (equatorial horizontal parallax, see chapter 9).
25
R
0
´
1
tan H
2
°
7.31
H
2
° 4.4
R
0, improved
´ R
0
´ 0.06sin14.7R
0
´ 13
3rd correction: H
3
H
2
f R
0
H
2
R
0
f
p hPa
1010
283
273T ° C
pin.Hg
29.83
510
460T ° F
P is a Iunction oI topocentric altitude and horizontal parallax oI a body.* It has to be added to H
3
.
An additional correction Ior the oblateness oI the earth is recommended (∆P, see p. 28).
H4 represents the geocentric altitude of the body, the altitude with respect to the celestial horizon.
Semidiameter
When observing sun or moon with a marine sextant or theodolite, it is not possible to locate the center oI the body
precisely. It is thereIore common practice to measure the altitude oI the upper or lower limb oI the body and add or
subtract the apparent semidiameter, SD. The latter is the angular distance oI the respective limb Irom the center oI the
body (Fig. 28).
We have to correct Ior the geocentric SD, the SD measured by a Iictitious observer at the center the earth, because H
4
is measured at the center the earth (see Fig. 24)**. The geocentric semidiameters oI sun and moon are given on the
daily pages oI the Nautical Almanac 12. The geocentric SD oI a body can be calculated Irom its tabulated horizontal
parallax. This is oI particular interest when observing the moon.
The Iactor k is the ratio oI the radius oI the respective body to the equatorial radius oI the earth (r
Earth
÷ 6378 km).
**Note that Fig. 28 shows the topocentric semidiameter.
Although the semidiameters oI the navigational planets are not quite negligible (the SD oI Venus can increase to 0.5'),
the apparent centers oI these bodies are usually observed, and no correction Ior SD is applied. With a strong telescope,
however, the limbs oI the brightest planets can be observed. In this case the correction Ior semidiameter should be
applied. Semidiameters oI stars are much too small to be measured (SD ÷ 0).
(lower limb: add SD, upper limb: subtract SD)
When using a bubble sextant, we observe the center oI the body and skip the correction Ior semidiameter.
The altitude obtained after applying the above corrections is called observed altitude, Ho:
Ho H
5
Ho represents the geocentric altitude of the center of the body.
*To be exact, the parallax Iormula shown above is rigorous Ior the observation oI the center oI a body only. When observing the lower or upper
limb, there is a small error caused by the curvature oI the body's surIace which is usually negligible. The rigorous Iormula Ior the observation oI
either oI the limbs is:
26
4th correction: H
4
H
3
P
5th correction: H
5
H
4
SD
geocentric
SD
geocentric
arcsin k sin HP k HP k
Moon
r
Moon
r
Earth
0.2725
P arcsin sin HPcos H
3
HPcos H
3
Lower limb: P arcsin
sin HPcos H
3
k
arcsink sin HP HPcos H
3
Upper limb: P arcsin
sin HPcos H
3
k
arcsin k sin HP HPcos H
3
Combined corrections for semidiameter and parallax
H
3
can be reduced to the observed altitude in one step. The Iollowing Iormula includes the corrections Ior semidiameter
and parallax in altitude:
(lower limb: add k, upper limb: subtract k)
Alternative procedure for semidiameter and parallax
Correcting Ior semidiameter before correcting Ior parallax is also possible. In this case, however, we have to calculate
with the topocentric semidiameter, the semidiameter oI the respective body as seen Irom the observer's position on the
surIace oI the earth (see Fig. 28).
With the exception oI the moon, the body nearest to the earth, there is no signiIicant diIIerence between topocentric and
geocentric semidiameter. The topocentric SD oI the moon is only marginally greater than the geocentric SD when the
moon is on the sensible (geoidal) horizon but increases measurably as the altitude increases because oI the decreasing
distance between observer and moon. The distance is smallest (decreased by about the radius oI the earth) when the
moon is in the zenith. As a result, the topocentric SD oI the moon being in the zenith is approximately 0.3' greater than
the geocentric SD. This phenomenon is called augmentation (Fig. 29).
The rigorous Iormula Ior the topocentric (augmented) semidiameter oI the moon is:
(observation oI lower limb: add k, observation oI upper limb: subtract k)
The approximate topocentric semidiameter oI the moon can be calculated with a simpler Iormula given by Meeus 2. It
reIers to the center oI the moon but is still accurate enough Ior the purpose oI navigation (error · 1'') when applied to
the altitude oI the upper or lower limb, respectively:
A similar Iormula was proposed by Stark 14:
27
SD
topocentric
arctan
k
1
sin
2
HP
cos H
3
k
2
sin H
3
SD
topocentric
k HP1 sin HPsin H
3
SD
topocentric
k HP
1 sin HPsin H
3
Ho H
3
arcsin
sin HPcos H
3
k
H
3
HPcos H
3
k
Thus, the alternative Iourth correction is:
(lower limb: add SD, upper limb: subtract SD)
H4,alt represents the topocentric altitude of the center of the moon.
Using the parallax Iormula explained above, we calculate P
alt
Irom H
4,alt:
Thus, the alternative IiIth correction is:
Since the geocentric SD is easier to calculate than the topocentric SD, it is usually more convenient to correct Ior the
semidiameter in the last place or, better, to use the combined correction Ior parallax and semidiameter unless one has to
know the augmented SD oI the moon Ior special reasons.
The topocentric semidiameter oI the moon can also be calculated Irom the observed altitude (the geocentric altitude oI
the center oI the moon), Ho:
Instead oI Ho, the computed altitude, Hc, can be used (see chapter 4).
Correction for the oblateness of the earth
The above Iormulas Ior parallax and semidiameter are rigorous Ior spherical bodies. In Iact, the earth is not a sphere but
rather resembles an oblate spheroid, a sphere Ilattened at the poles (chapter 9). In most cases, the navigator will not
notice the diIIerence. However, when observing the moon, the Ilattening oI the earth may cause a small but measurable
error (up to ±0.2') in the parallax, depending on the observer's position. ThereIore, a small correction, P, should be
added to P iI higher precision is required 12. When using the combined Iormula Ior semidiameter and parallax, P is
added to Ho.
* Replace H with H3 or H4,alt, respectively.
Lat is the observer's estimated latitude (chapter 4). Az
N
, the true azimuth oI the moon, is either measured with a
compass (compass bearing) or calculated using the azimuth Iormulas given in chapter 4.
Phase correction (Venus and Mars)
Since Venus and Mars show phases similar to the moon, their apparent center may diIIer somewhat Irom the actual
center. The coordinates oI both planets tabulated in the Nautical Almanac 12 include the phase correction, i. e., they
reIer to the apparent center. The phase correction Ior Jupiter and Saturn is too small to be signiIicant.
In contrast, coordinates calculated with Interactive Computer Ephemeris reIer to the actual center. In this case, the
upper or lower limb oI the respective planet should be observed iI the magniIication oI the telescope is suIIicient, and
the correction Ior semidiameter should be applied.
28
5th correction (alternative): H
5, alt
H
4,alt
P
alt
Ho H
5, alt
SD
topocentric
arcsin
k
1
1
sin
2
HP
2
sin Ho
sin HP
P
improved
P P
P f HP sin2Lat cos A:
N
sin H sin
2
Lat cos H * f
1
298.257
P
alt
arcsin sin HPcos H
4, alt
HPcos H
4, alt
4th correction (alt.): H
4, alt.
H
3
SD
topocentric
Altitude correction tables
The Nautical Almanac provides sextant altitude correction tables Ior sun, planets, stars (pages A2 A4), and the moon
(pages xxxiv xxxv), which can be used instead oI the above Iormulas iI small errors (· 1') are tolerable (among other
things, the tables cause additional rounding errors).
Other corrections
Sextants with an artiIicial horizon can exhibit additional errors caused by acceleration Iorces acting on the bubble or
pendulum and preventing it Irom aligning itselI with the direction oI gravity.
Such acceleration Iorces can be accidental (vessel movements) or systematic (coriolis Iorce). The coriolis Iorce is
important to air navigation (high speed!) and requires a special correction Iormula.
In the vicinity oI mountains, ore deposits, and other local irregularities oI the earth's crust, the vector oI gravity may
slightly diIIer Irom the normal to the reIerence ellipsoid, resulting in altitude errors that are diIIicult to predict (local
deflection of the vertical, chapter 9). Thus, the astronomical position oI an observer (resulting Irom astronomical
observations) may be slightly diIIerent Irom his geographic (geodetic) position with respect to a reference ellipsoid
(GPS position). The diIIerence is usually small at sea and may be ignored there. On land, particularly in the vicinity oI
mountain ranges, position errors oI up to 50 arcseconds (Alps) or even 100 arcseconds (Himalaya) have been Iound.
For the purpose oI surveying there are maps containing local corrections Ior latitude and longitude, depending on the
respective reIerence ellipsoid.
29
Chapter 3
Geographic Position and Time
Geographic terms
In celestial navigation, the earth is regarded as a sphere. Although this is an approximation only, the geometry oI
the sphere is applied successIully, and the errors caused by the Ilattening oI the earth are usually negligible (chapter 9).
A circle on the surIace oI the earth whose plane passes through the center oI the earth is called a great circle. Thus, a
great circle has the greatest possible diameter oI all circles on the surIace oI the earth. Any circle on the surIace oI the
earth whose plane does not pass through the earth's center is called a small circle. The equator is the only great circle
whose plane is perpendicular to the polar axis, the axis of rotation. Further, the equator is the only parallel of latitude
being a great circle. Any other parallel oI latitude is a small circle whose plane is parallel to the plane oI the equator. A
meridian is a great circle going through the geographic poles, the points where the polar axis intersects the earth's
surIace. The upper branch oI a meridian is the halI Irom pole to pole passing through a given point, e. g., the
observer's position. The lower branch is the opposite halI. The Greenwich meridian, the meridian passing through the
center oI the transit instrument at the Royal Greenwich Observatory, was adopted as the prime meridian at the
International Meridian Conference in 1884. Its upper branch is the reIerence Ior measuring longitudes (0°...¹180°
east and 0°...180° west), its lower branch (180°) is the basis Ior the International Dateline (Fig. 31).
Each point oI the earth's surIace has an imaginary counterpart on the surIace oI the celestial sphere obtained by central
projection. The projected image oI the observer's position, Ior example, is the zenith. Accordingly, there are two
celestial poles, the celestial equator, celestial meridians, etc.
The equatorial system of coordinates
The geographic position of a celestial body, GP, is deIined by the geocentric equatorial system of coordinates, a
spherical coordinate system the origin oI which is at the center oI the earth (Fig. 32). The Greenwich hour angle,
GHA, is the angular distance oI the upper branch oI the meridian passing through GP Irom the upper branch oI the
Greenwich meridian (Lon ÷ 0°), measured westward Irom 0° through 360°. The meridian going through GP (as well as
its projection on the celestial sphere) is called hour circle. The Declination, Dec, is the angular distance oI GP Irom
the plane oI the equator, measured northward through ¹90° or southward through 90°. GHA and Dec are geocentric
coordinates (measured at the center oI the earth).
31
Although widely used, the term ,geographic position' is misleading when applied to a celestial body since it actually
describes a geocentric position in this case (see chapter 9).
GHA and Dec are equivalent to geocentric longitude and latitude with the exception that longitudes are measured
westward through 180° and eastward through +180°.
Since the Greenwich meridian rotates with the earth from west to east, whereas each hour circle remains linked
with the almost stationary position of the respective body in the sky, the Greenwich hour angles of all celestial
bodies increase by approx. 15° per hour (360° in 24 hours). In contrast to stars (15° 2.46' /h), the GHA's oI sun,
moon, and planets increase at slightly diIIerent (and variable) rates. This is caused by the revolution oI the planets
(including the earth) around the sun and by the revolution oI the moon around the earth, resulting in additional apparent
motions oI these bodies in the sky. For several applications it is useIul to measure the angular distance between the hour
circle oI a celestial body and the hour circle oI a reIerence point in the sky instead oI the Greenwich meridian because
the angle thus obtained is independent oI the earth's rotation. The sidereal hour angle, SHA, oI a given body is the
angular distance oI its hour circle (upper branch) Irom the hour circle (upper branch) oI the first point of Aries (also
called vernal equinox, see below), measured westward Irom 0° through 360°. Thus, the GHA oI a body is the sum oI
its sidereal hour angle and the GHA oI the Iirst point oI Aries, GHA
Aries
:
(II the resulting GHA is greater than 360°, subtract 360°.)
The angular distance oI a body eastward Irom the hour circle oI the vernal equinox, measured in time units (24h ÷
360°), is called right ascension, RA. The latter is mostly used by astronomers whereas navigators preIer the SHA.
Fig. 33 illustrates the diIIerent hour angles on the plane oI the equator (seen Irom the celestial north pole).
Declinations are not aIIected by the rotation oI the earth. The declinations oI sun and planets change primarily due to
the obliquity of the ecliptic, the inclination oI the earth's equator to the ecliptic. The latter is the plane oI the earth's
orbit and Iorms a great circle on the celestial sphere. The declination oI the sun, Ior example, varies periodically
between ca. ¹23.5° at the time oI the summer solstice and ca. 23.5° at the time oI the winter solstice. (Fig.34).
The two points on the celestial sphere where the great circle oI the ecliptic intersects the celestial equator are called
equinoxes. The term equinox is also used Ior the instant at which the apparent sun, moving westward along the ecliptic
during the course oI a year, crosses the celestial equator, approximately on March 21 and on September 23. There is a
vernal equinox (first point of Aries, vernal point) and an autumnal equinox. The Iormer is the reIerence point Ior
measuring sidereal hour angles (Fig. 35). Every time the sun passes through an equinox (Dec ÷ 0), day and night have
(approximately) the same length (12 h), regardless oI the observer's position (Lat. aequae noctes ÷ equal nights).
32
GHA SHA GHA
Aries
RA h 24 h
SHA °
15
SHA ° 360° 15RA h
The declinations oI the planets and the moon are also inIluenced by the inclinations oI their own orbits to the ecliptic.
The plane oI the moon's orbit, Ior example, is inclined to the ecliptic by approx. 5° and makes a tumbling movement
(precession, see below) with a cycle oI 18.6 years (Saros cycle). As a result, the declination oI the moon varies between
approx. 28.5° and ¹28.5° at the beginning and at the end oI the Saros cycle, and between approx. 18.5° and ¹18.5° in
the middle oI the Saros cycle.
Further, sidereal hour angles and declinations oI all bodies change slowly due to the inIluence oI the precession oI the
earth's polar axis. Precession is a slow, tumbling movement oI the polar axis along the surIace oI an imaginary double
cone. One revolution takes about 26000 years (Platonic year). As a result, the equinoxes move westward along the
celestial equator at a rate oI approx. 50'' per year. Thus, the sidereal hour angle oI each star decreases at about the same
rate. In addition, there is a small elliptical oscillation oI the polar axis, called nutation, which causes the equinoxes to
travel along the celestial equator at a periodically changing rate. Thus we have to distinguish between the Iicticious
mean equinox of date and the true equinox of date (see time measurement). Accordingly, the declination oI each
body oscillates. The same applies to the rate oI change oI the sidereal hour angle and right ascension oI each body.
Even stars are not Iixed in space but move individually, resulting in a slow driIt oI their equatorial coordinates (proper
motion). Finally, the apparent positions oI bodies are inIluenced by other Iactors, e. g., the Iinite speed oI light (light
time, aberration), and annual parallax, the parallax caused by the earth orbiting around the sun 16. The accurate
prediction oI geographic positions oI celestial bodies requires complicated algorithms. Formulas Ior the calculation oI
lowprecision ephemerides oI the sun (accurate enough Ior celestial navigation) are given in chapter 15.
Time measurement in navigation and astronomy
Since the Greenwich hour angle of any celestial body changes rapidly, celestial navigation requires accurate time
measurement, and the instant of an observation should be measured to the second if possible. This is usually done
by means oI a chronometer and a stopwatch (chapter 17). The eIIects oI time errors are dicussed in chapter 16. On the
other hand, the earth's rotation with respect to celestial bodies provides an important basis Ior astronomical time
measurement.
Coordinates tabulated in the Nautical Almanac reIer to Universal Time, UT. UT has replaced Greenwich Mean
Time, GMT, the traditional basis Ior civil time keeping. Conceptually, UT (like GMT) is the hour angle oI the
Iictitious mean sun, expressed in hours, with respect to the lower branch oI the Greenwich meridian (mean solar time,
Fig. 36).
33
UT is calculated using the Iollowing Iormula:
UT h GMT h
GHA
Mean Sun
°
15
12
(II UT is greater than 24 h, subtract 24 hours.)
By definition, the GHA of the mean sun increases by exactly 15° per hour, completing a 360° cycle in 24 hours.
The unit Ior UT (and GMT) is 1 solar day, the time interval between two consecutive meridian transits oI the mean sun.
The rate oI change oI the GHA oI the apparent (observable) sun varies periodically and is sometimes slightly greater,
sometimes slightly smaller than 15°/h during the course oI a year. This behavior is caused by the eccentricity oI the
earth's orbit and by the obliquity oI the ecliptic. The time measured by the hour angle oI the apparent sun with respect
to the lower branch oI the Greenwich meridian is called Greenwich Apparent Time, GAT. A sundial located at the
Greenwich meridian would indicate GAT. The diIIerence between GAT and UT (GMT) at a given instant is called
equation of time, EoT:
EoT varies periodically between approx. 14 and ¹17 minutes (Fig. 37). Predicted values Ior EoT Ior each day oI the
year (at 0:00 and 12:00 UT) are given in the Nautical Almanac (grey background indicates negative EoT). EoT is
needed when calculating times oI sunrise and sunset, or determining a noon longitude (chapter 6). Formulas Ior the
calculation oI EoT are given in chapter 15.
The hour angle oI the mean sun with respect to the lower branch oI the local meridian (the upper branch going through
the observer's position) is called Local Mean Time, LMT. LMT and UT are linked through the Iollowing Iormula:
The instant oI the mean sun passing through the upper branch oI the local meridian is called Local Mean Noon, LMN.
A zone time is the local mean time with respect to a longitude being a multiple oI +15°. Thus, zone times diIIer by an
integer number oI hours. In the US, Ior example, Eastern Standard Time (UT5h) is LMT at 75° longitude, PaciIic
Standard Time (UT8h) is LMT at 120 ° longitude. Central European Time (UT¹1h) is LMT at ¹15° longitude.
The hour angle oI the apparent sun with respect to the lower branch oI the local meridian is called Local Apparent
Time, LAT:
The instant oI the apparent sun crossing the upper branch oI the local meridian is called Local Apparent Noon, LAN.
Time measurement by the earth's rotation does not necessarily require the sun as the reIerence point in the sky.
Greenwich Apparent Sidereal Time, GAST, is a time scale based upon the Greenwich hour angle (upper branch) oI
the true vernal equinox oI date, GHA
Aries
(see Fig. 33).
34
EoT GAT UT
LMT h UT h
Lon°
15
LAT h GAT h
Lon °
15
The values Ior GHA
Aries
tabulated in the Nautical Almanac reIer to the true equinox oI date.
GAST is easily measured by the Greenwich meridian transit oI stars since GAST and the right ascension oI the
observed star are numerically equal at the moment oI meridian transit.
The Greenwich hour angle oI the imaginary mean vernal equinox oI date (traveling along the celestial equator at a
constant rate) deIines Greenwich Mean Sidereal Time, GMST. The diIIerence to GAST at a given instant is called
equation of the equinoxes, EQ, or nutation in right ascension. EQ can be predicted precisely. It varies within approx.
+1s.
GMST is oI some importance Ior it is the actual basis Ior UT. Since time measurement by meridian transit oI the sun is
not accurate enough Ior many scientiIic applications, Universal Time is by deIinition calculated Irom GMST. This is
possible because there is a close correlation between GMST and mean solar time at Greenwich. The time thus obtained
is called UT0. Applying a correction (milliseconds) Ior the small eIIect oI polar motion (a tumbling movement oI the
polar axis with respect to the earth's crust) yields UT1, commonly called UT.
Due to the earth's revolution around the sun, a mean sidereal day (the time interval between two consecutive meridian
transits oI the mean equinox) is slightly shorter than a mean solar day:
In analogy with LMT and LAT, there is a Local Mean Sidereal Time, LMST, and a Local Apparent Sidereal Time,
LAST:
Solar time and sidereal time are both linked to the earth's rotation. The earth's rotating speed, however, decreases slowly
(tidal Iriction) and, moreover, Iluctuates in an unpredictable manner due to random movements oI matter within the
earth's body (magma) and on the surIace (water, air). ThereIore, neither oI both time scales is strictly linear. Many
astronomical applications, however, require an exactly linear time scale. One example is the calculation oI ephemerides
since the motions oI celestial bodies in space are independent oI the earth's rotation.
International Atomic Time, TAI, is the most precise time standard presently available. It is obtained by statistical
analysis oI data supplied by a great number oI atomic clocks all over the world. Among others, two important time
scales are derived Irom TAI:
Today, civil liIe is mostly determined by Coordinated Universal Time, UTC, which is the basis Ior time signals
broadcast by radio stations, e. g., WWV or WWVH. UTC is controlled by TAI. Due to the varying rotating speed oI the
earth, UT tends to driIt away Irom UTC. This is undesirable since the cycle oI day and night is linked to UT. ThereIore,
UTC is synchronized to UT, iI necessary, by inserting (or omitting) leap seconds at certain times (June 30 and
December 31) in order to avoid that the diIIerence, UT, exceeds the speciIied maximum value oI +0.9 s.
N is the cumulative number oI leap seconds inserted until now (N ÷ 34 in 2009.0). Due to the occasional leap seconds,
UTC is not a continuous time scale! Predicted values Ior UT (÷ DUT1 ÷ UT1UTC) are published by the IERS
Rapid Service 15 on a weekly basis (IERS Bulletin A). The IERS also announces the insertion (or omission) oI leap
seconds in advance (IERS Bulletins A ¹ C).
Terrestrial Time, TT (Iormerly called Terrestrial Dynamical Time, TDT), is another derivative oI TAI:
TT has replaced Ephemeris Time, ET. The oIIset oI 32.184 s with respect to TAI is necessary to ensure a seamless
continuation oI ET. TT is used in astronomy (calculation oI ephemerides) and space Ilight. The diIIerence between TT
and UT is called T:
At the beginning oI the year 2009, T was ¹65.8s.
35
GAST h
GHA
Aries
°
15
EQ GAST GMST
24 h Mean Sidereal Time 23h56m 4.090524s Mean Solar Time
LMST h GMST h
Lon °
15
LAST h GAST h
Lon°
15
UT UTC UT UTC TAI N
TT TAI 32.184s
T TT UT
T is oI some importance since computer almanacs require TT (TDT) as time argument (programs using UT calculate
on the basis oI interpolated and extrapolated T values). A precise longterm prediction oI T is impossible. ThereIore,
computer almanacs using only UT as time argument may become less accurate in the long term. T values Ior the near
Iuture can be calculated with the Iollowing Iormula:
Current values Ior UT1UTC and TAIUTC (cumulative number oI leap seconds) are published in the IERS Bulletin A.
Bulletin A also contains shortterm predictions Ior UT1UTC.
The GMT problem:
The term GMT has become ambigous since it is often used as a synonym for UTC now. Moreover, astronomers
used to reckon GMT from the upper branch of the Greenwich meridian until 1925 (the time thus obtained is
sometimes called Greenwich Mean Astronomical Time, GMAT). Therefore, the term GMT should be avoided in
scientific publications, except when used in a historical context.
The Nautical Almanac
Predicted values Ior GHA and Dec oI sun, moon and the navigational planets with reIerence to UT are tabulated Ior
each integer hour oI the year on the daily pages oI the Nautical Almanac, N.A., and similar publications 12, 13.
GHA
Aries
is tabulated in the same manner.
Listing GHA and Dec oI all 57 Iixed stars used in navigation Ior each integer hour oI the year would require too much
space in a book. ThereIore, sidereal hour angles are tabulated instead oI Greenwich hour angles. Since declinations and
sidereal hour angles oI stars change only slowly, tabulated values Ior periods oI 3 days are accurate enough Ior celestial
navigation. GHA is obtained by adding the SHA oI the respective star to the current value oI GHA
Aries
.
GHA and Dec Ior each second oI the year are obtained using the interpolation tables at the end oI the N.A. (printed on
tinted paper), as explained in the Iollowing directions:
1.
We note the exact time oI observation (UT), determined with a chronometer and a stopwatch. II UT is not available, we
can use UTC. The resulting error is tolerable in most cases.
2.
We look up the day oI observation in the N.A. (two pages cover a period oI three days).
3.
We go to the nearest integer hour preceding the time oI observation and note GHA and Dec oI the observed body. In
case oI a Iixed star, we Iorm the sum oI GHA Aries and the SHA oI the star, and note the tabulated declination. When
observing planets, we note the and Iactors given at the bottom oI the appropriate column. For the moon, we take v
and d Ior the nearest integer hour preceding the time oI observation.
The quantity v is necessary to apply an additional correction to the Iollowing interpolation oI the GHA oI moon and
planets. It is not required Ior stars. The sun does not require a v Iactor since the correction has been incorporated in the
tabulated values Ior the sun's GHA.
The quantity d, which is negligible Ior stars, is the rate oI change oI Dec, measured in arcminutes per hour. It is needed
Ior the interpolation oI Dec. The sign oI d is critical!
4.
We look up the minute oI observation in the interpolation tables (1 page Ior each 2 minutes oI the hour), go to the
second oI observation, and note the increment Irom the respective column.
We enter one oI the three columns to the right oI the increment columns with the v and d Iactors and note the
corresponding corr(ection) values (vcorr and dcorr).
The sign oI dcorr depends on the trend oI declination at the time oI observation. It is positive iI Dec at the integer hour
Iollowing the observation is greater than Dec at the integer hour preceding the observation. Otherwise it is negative.
v corr is negative Ior Venus. Otherwise, it is always positive.
36
T 32.184s TAI UTC UT1 UTC
5.
We Iorm the sum oI Dec and dcorr (iI applicable).
6.
We Iorm the sum oI GHA (or GHA Aries and SHA in case oI a star), increment, and vcorr (iI applicable).
SHA values tabulated in the Nautical Almanac reIer to the true vernal equinox oI date.
Interactive Computer Ephemeris
Interactive Computer Ephemeris, ICE, is a computer almanac developed by the U.S. Naval Observatory (successor
oI the Floppy Almanac) in the 1980s.
ICE is FREEWARE (no longer supported by USNO), compact, easy to use, and provides a vast quantity oI accurate
astronomical data Ior a time span oI almost 250 (!) years. In spite oI the old design (DOS program), ICE is still a useIul
tool Ior navigators and astronomers.
Among many other Ieatures, ICE calculates GHA and Dec Ior a given body and time as well as altitude and azimuth oI
the body Ior an assumed position (see chapter 4) and, moreover, sextant altitude corrections. Since the navigation data
are as accurate as those tabulated in the Nautical Almanac (approx. 0.1'), the program makes an adequate alternative,
although a printed almanac (and sight reduction tables) should be kept as a backup in case oI a computer Iailure. The
Iollowing instructions reIer to the Iinal version (0.51). Only program Ieatures relevant to navigation are explained.
1. Installation
Copy the program Iiles to a chosen directory on the hard drive, Iloppy disk, USB stick, or similar storage device.
ICE.EXE is the executable program Iile.
2. Getting Started
DOS users: Change to the program directory and enter "ice" or "ICE ". Windows users can run ICE in a DOS box.
Linux users: Install the DOSemulator "DOSBox". Copy the ICE Iiles to a directory oI your choice in your personal
Iolder. ICE is started through the command 'dosbox', Iollowed by a blank space and the path to the program Iile:
dosbox /home/·username~/·program directory~/ICE.EXE (Note that Linux is casesensitive.)
AIter the program has started, the main menu appears.
Use the Iunction keys F1 to F10 to navigate through the submenus. The program is more or less selIexplanatory. Go to
the submenu INITIAL VALUES (F1). Follow the directions on the screen to enter date and time oI observation (F1),
assumed latitude (F2), assumed longitude (F3), and your local time zone (F6). Assumed latitude and longitude deIine
your assumed position.
Use the correct data Iormat, as shown on the screen (decimal Iormat Ior latitude and longitude). AIter entering the
above data, press F7 to accept the values displayed. To change the deIault values permanently, edit the Iile ice.dIt with
a text editor (aIter making a backup copy) and make the desired changes. Do not change the data Iormat. The numbers
have to be in columns 2140. An output Iile can be created to store calculated data. Go to the submenu FILE OUTPUT
(F2) and enter a chosen Iile name, e.g., OUTPUT.TXT.
3. Calculation of Navigational Data
From the main menu, go to the submenu NAVIGATION (F7). Enter the name oI the body. The program displays GHA
and Dec oI the body, GHA and Dec oI the sun (iI visible), and GHA oI the vernal equinox Ior the date and time (UT)
stored in INITIAL VALUES.
Hc (computed altitude) and Zn (azimuth) mark the apparent position oI the body as observed Irom the assumed
position. Approximate altitude corrections (reIraction, SD, PA), based upon Hc, are also displayed (Ior lower limb oI
body). The semidiameter oI the moon includes augmentation. The coordinates calculated Ior Venus and Mars do not
include phase correction. ThereIore, the upper or lower limb (iI visible) should be observed. T is TT(TDT)UT, the
predicted diIIerence between terrestrial time and UT Ior the given date. The T value Ior 2009.0 predicted by ICE is
71.8s, the actual value is 65.8s (see below).
37
Horizontal parallax and semidiameter oI a body can be extracted Irom the submenu POSITIONS (F3). Choose
APPARENT GEOCENTRIC POSITIONS (F1) and enter the name oI the body (sun, moon, planets).
The last column shows the distance oI the center oI the body Irom the center oI the earth, measured in astronomical
units (1 AU ÷ 149.6
.
10
6
km). HP and SD are calculated as Iollows:
r
E
is the equatorial radius oI the earth (6378 km). r
B
is the radius oI the body (Sun: 696260 km, Moon: 1378 km, Venus:
6052 km, Mars: 3397 km, Jupiter: 71398 km, Saturn: 60268 km).
The apparent geocentric positions reIer to TT (TDT), but the diIIerence between TT and UT has no signiIicant eIIect on
HP and SD.
To calculate times oI rising and setting oI a body, go to the submenu RISE & SET TIMES (F6) and enter the name oI
the body. The columns on the right display the time oI rising, meridian transit, and setting Ior the assumed location
(UT¹xh, according to the time zone speciIied).
The increasing error oI T values predicted by ICE may lead to reduced precision when calculating navigation data in
the Iuture. The coordinates oI the moon are particularly sensitive to errors oI T. UnIortunately, ICE has no option Ior
editing and modiIying the internal T algorithm. The highprecision part oI ICE, however, is not aIIected since TT
(TDT) is the time argument.
To circumvent the T problem, extract GHA and Dec using the Iollowing procedure:
1. Compute GAST using SIDEREAL TIME (F5). The time argument is UT.
2. Edit date and time at INITIAL VALUES (F1). Now, the time argument is TT (UT¹T). Compute RA and Dec
using POSITIONS (F3) and APPARENT GEOCENTRIC POSITIONS (F1).
3. Use the Iollowing Iormula to calculate GHA Irom GAST and RA (RA reIers to the true vernal equinox oI
date):
Add or subtract 360° iI necessary.
Highprecision GHA and Dec values thus obtained can be used as an internal standard to crosscheck mediumprecision
data obtained through NAVIGATION (F7).
38
HP arcsin
r
E
km
distance km
SD arcsin
r
B
km
distance km
GHA ° 15GAST h 24h RA h
Chapter 4
Finding One's Position (Sight Reduction)
Lines of Position
Any geometrical or physical line passing through the observer's (still unknown) position and accessible through
measurement or observation is called a position line or line of position, LOP. Examples are circles oI equal altitude,
meridians, parallels oI latitude, bearing lines (compass bearings) oI terrestrial objects, coastlines, rivers, roads, railroad
tracks, power lines, etc. A single position line indicates an infinite series of possible positions. The observer's actual
position is marked by the point oI intersection oI at least two position lines, regardless oI their nature. A position thus
Iound is called fix in navigator's language. The concept oI the position line is essential to modern navigation.
Sight Reduction
Finding a line oI position by observation oI a celestial object is called sight reduction. Although some background in
mathematics is required to comprehend the process completely, knowing the basic concepts and a Iew equations is
suIIicient Ior most practical applications. The geometrical background (law oI cosines, navigational triangle) is given in
chapter 10 and chapter 11. In the Iollowing, we will discuss the semigraphic methods developed by Sumner and St.
Hilaire. Both methods require relatively simple calculations only and enable the navigator to plot lines oI position on a
nautical chart or plotting sheet (chapter 13).
Knowing altitude and geographic position oI a body, we also know the radius oI the corresponding circle oI equal
altitude (our circular line oI position) and the position oI its center. As mentioned in chapter 1 already, plotting circles
oI equal altitude on a chart is usually impossible due to their large dimensions and the distortions caused by map
projection. However, Sumner and St. Hilaire showed that only a short arc oI each circle oI equal altitude is needed to
Iind one's position. Since this arc is comparatively short, it can be represented by a secant or a tangent oI the circle.
Local Meridian, Local Hour Angle and Meridian Angle
The meridian passing through a given position, usually that oI the observer, is called local meridian. In celestial
navigation, the angle between the hour circle oI the observed body (upper branch) and the local meridian (upper
branch) plays a Iundamental role. On the analogy oI the Greenwich hour angle, we can measure this angle westward
Irom the local meridian (0°...¹360°). In this case, the angle is called local hour angle, LHA. It is also possible to
measure the angle westward (0°...¹180°) or eastward (0°...180°) Irom the local meridian in which case it is called
meridian angle, t. In most navigational Iormulas, LHA and t can be substituted Ior each other since the trigonometric
Iunctions return the same results. For example, the cosine oI ¹315° is the same as the cosine oI 45 °.
Like LHA, t is the algebraic sum oI the Greenwich hour angle oI the body, GHA, and the observer's geographic
longitude, Lon. To make sure that the obtained angle is in the desired range, the Iollowing rules have to be applied
when Iorming the sum oI GHA and Lon:
In all calculations, the sign oI Lon and t, respectively, has to be observed careIully. The sign convention (important!) is:
Eastern longitude: positive
Western longitude: negative
Eastern meridian angle: negative
Western meridian angle: positive
For reasons oI symmetry, we will reIer to the meridian angle in the Iollowing considerations (the meridian angle ¹t
results in the same altitude as the meridian angle t), although the local hour angle would lead to the same results.
41
t
GHA + Lon iI GHA + Lon 180°
GHA + Lon 360° iI GHA + Lon > 180°
LHA
GHA + Lon iI 0°GHA +Lon 360°
GHA + Lon + 360° iI GHA + Lon 0°
GHA + Lon  360° iI GHA + Lon > 360°
Fig. 41 illustrates the various angles involved in the sight reduction process. The spherical triangle Iormed by GP, AP,
and the north pole is called navigational (or nautical) triangle (chapter 11). AP is the observer's position (see intercept
method).
Sumner`s Method
In December 1837, Thomas H Sumner, an American sea captain, was on a voyage Irom South Carolina to Greenock,
Scotland. When approaching St. George´s Channel between Ireland and Wales, he managed to measure a single altitude
oI the sun aIter a longer period oI bad weather. Using the time sight Iormula (see chapter 6), he calculated a longitude
Irom his estimated latitude. Since he was doubtIul about his estimate, he repeated his calculations with two slightly
diIIerent latitudes. To his surprise, the three positions thus obtained were on a straight line. Accidentally, the line passed
through the position oI a lighthouse oII the coast oI Wales (Small´s Light). By intuition, Sumner steered his ship along
this line and soon aIter, Small´s Light came in sight. Sumner concluded that he had Iound a ''line oI equal altitude''. The
publication oI his method in 1843 marked the beginning oI 'modern' celestial navigation 18. Although rarely used
today, it is still an interesting alternative. It is easy to comprehend and the calculations to be done are quite simple.
Fig. 42 illustrates the points where a circle oI equal altitude intersects two chosen parallels oI latitude.
An observer being between Lat
1
and Lat
2
is either on the arc AB or on the arc CD. With a rough estimate oI the
longitude oI his position, the observer can easily Iind on which oI both arcs he is, Ior example, AB. The arc thus Iound
is the relevant part oI his line oI position, the other arc is discarded. On a chart, we can approximate the line oI position
by drawing a straight line through A and B which is a secant oI the circle oI equal altitude. This secant is called
Sumner line. BeIore plotting the Sumner line on our chart, we have to Iind the respective longitude oI each point oI
intersection, A, B, C, and D.
Procedure:
1.
We choose a parallel oI latitude (Lat
1
) north oI our estimated latitude. PreIerably, Lat
1
should be marked by the nearest
horizontal grid line on our chart or plotting sheet.
2.
From Lat
1
, Dec, and the observed altitude, Ho, we calculate the meridian angle, t, using the Iollowing Iormula:
42
t arccos
sin Ho  sin Lat sin Dec
cos Lat cos Dec
The equation is derived Irom the navigational triangle (chapter 10 & chapter 11). It has two solutions, ¹t and t, since
the cosine oI ¹t equals the cosine oI t. Geometrically, this corresponds with the Iact that the circle oI equal altitude
intersects the parallel oI latitude at two points. Using the Iollowing Iormulas and rules, we obtain the longitudes oI
these two points oI intersection, Lon and Lon':
Comparing the longitudes thus obtained with our estimate, we select the most probable longitude and discard the other
one. This method oI Iinding one's longitude is called time sight (chapter 6).
3.
We choose a parallel oI latitude (Lat
2
) south oI our estimated latitude. The diIIerence between Lat
1
and Lat
2
should not
exceed 1 or 2 degrees. We repeat steps 1 and 2 with the second latitude, Lat
2
.
4.
On our plotting sheet, we mark both remaining longitudes, either one on its corresponding parallel oI latitude, and plot
the Sumner line through the points thus located (LOP1, see Fig. 43 ).
Using the same parallels oI latitude, we repeat steps 1 through 4 with the declination and observed altitude oI a second
body. The point where the Sumner line thus obtained, LOP2, intersects LOP1 is our Iix.
II we have only a very rough estimate oI our latitude, the point oI intersection may be outside the interval deIined by
both parallels, but the Iix is still correct. Any Iix obtained with Sumner´s method has a small error caused by neglecting
the curvature oI the circles oI equal altitude. We can improve the Iix by iteration. For this purpose, we take a chart with
a larger scale, choose a new pair oI assumed latitudes, nearer to the Iix, and repeat the procedure. Ideally, the horizontal
angular distance between both bodies should be 90° (30°...150° is tolerable). Otherwise, the Iix would become
indistinct. Further, neither oI the bodies should be near the local meridian (see time sight, chapter 6). Sumner´s method
has the (small) advantage that no protractor is needed to plot lines oI position.
The Intercept Method
This procedure was developed by the French navy oIIicer St. Hilaire and others and was Iirst published in 1875. AIter
that, it gradually became the standard Ior sight reduction since it avoids some oI the restrictions oI Sumner´s method.
Although the background is more complicated than with Sumner´s method, the practical application is very convenient.
Theory:
For any given position oI the observer, the geocentric altitude oI a celestial body is solely a Iunction oI the observer's
latitude, the declination oI the body, and the meridian angle (or local hour angle).
43
Lon ¸t¸GHA
Lon´ 360° ¸t¸ GHA
II Lon 180°  Lon +360°
II Lon´ 180°  Lon´ +360°
II Lon´ >+180°  Lon´ 360°
H f ( Lat , Dec , t )
The altitude formula is obtained by applying the law oI cosine Ior sides to the navigational triangle (chapter 10 & 11):
We choose an arbitrary point on our nautical chart which is not too Iar Irom our estimated position. PreIerably this is
the nearest point where two grid lines on the chart intersect. This point is called assumed position, AP (Fig. 44).
Using the above Iormula, we calculate the altitude oI the body resulting Irom Lat
AP
and Lon
AP
, the geographic
coordinates oI AP. The altitude thus obtained is called computed or calculated altitude, Hc.
Usually, Hc will slightly diIIer Irom the actually observed altitude, Ho (chapter 2). The diIIerence, H, is called
intercept.
Ideally, Ho and Hc will be identical iI the observer is exactly at AP.
In the Iollowing, we will discuss which possible positions oI the observer would result in the same intercept, H. For
this purpose, we assume that the intercept is an inIinitesimal quantity and denote it by dH. The general Iormula is:
This diIIerential equation has an inIinite number oI solutions. Since dH and both diIIerential coeIIicients are constant, it
can be reduced to a linear equation oI the general Iorm:
Thus, the graph is a straight line, and it is suIIicient to dicuss two special cases, dt÷0 and dLat÷0, respectively.
In the Iirst case, the observer is on the same meridian as AP, and the small change dH is solely caused by a small
variation oI latitude, dLat, whereas t is constant (dt ÷ 0). We diIIerentiate the altitude Iormula with respect to Lat:
Adding dLat to Lat
AP
, we obtain the point P1, as illustrated in Fig.44. P1 is on the circle oI equal altitude.
In the second case, the observer is on the same parallel oI latitude as AP, and dH is solely caused by a small change oI
the meridian angle, dt, whereas Lat is constant (dLat÷0). Again, we begin with the altitude Iormula:
44
H arcsin(sin Lat sin Dec + cos Lat cos Deccost )
AH Ho  Hc
dH
cH
c Lat
d Lat +
c H
ct
dt
d Lat a + bdt
sin H sin Lat sin Dec + cos Lat cos Dec cos t
d (sin H) (cos Lat sin Dec sin Lat cos Dec cos t )d Lat
cos H dH (cos Lat sin Dec sin Lat cos Dec cos t )d Lat
d Lat
cos H
cos Lat sin Dec sin Lat cos Dec cos t
dH
sin H sin Lat sin Dec + cos Lat cos Dec cos t
DiIIerentiating with respect to t, we get
Adding dt (corresponding with an equal change oI longitude, dLon) to LonAP, we obtain the point P2 which is on the
same circle oI equal altitude. Thus, we would measure Ho at P1 and P2, respectively. Knowing P1 and P2, we can now
plot a straight line passing through these positions. This line, a tangent oI the circle oI equal altitude, is our line of
position, LOP. The great circle passing through AP and GP is represented by a straight line perpendicular to the line oI
position, called azimuth line. The arc between AP and GP is the radius oI the circle oI equal altitude. The distance
between AP and the point where the azimuth line intersects the line oI position is the intercept, dH. The angle Iormed
by the azimuth line and the local meridian oI AP is called azimuth angle, Az. The same angle is measured between the
line oI position and the parallel oI latitude passing through AP (Fig. 44).
There are several ways to derive Az and the true azimuth, Az
N
, Irom the right (plane) triangle deIined by the vertices
AP, P1, and P2:
Timealtitude azimuth:
Alternatively, this Iormula can be derived Irom the navigational triangle (law oI sines and cosines, chapter 10 & chapter
11). Az is not necessarily identical with the true azimuth, Az
N
, since the arccos Iunction returns angles between 0° and
¹180°, whereas Az
N
is measured Irom 0° to ¹360°. To obtain Az
N
, we have to apply the Iollowing rules aIter
calculating Az with the Iormula Ior timealtitude azimuth:
Time azimuth:
The Iactor cos Lat is the relative circumIerence oI the parallel oI latitude on which AP is located (equator ÷ 1).
The time azimuth Iormula is also derived Irom the navigational triangle (law oI cotangents, chapter 10 & chapter 11).
Knowing the altitude is not necessary. This Iormula requires a diIIerent set oI rules to obtain Az
N
:
Calculating the time azimuth is more convenient with the arctan2 (÷ atan2) Iunction. The latter is part oI many
programming languages and spreadsheet programs and eliminates the quadrant problem. Thus, no conversion rules are
required to obtain Az
N
. In an OpenOIIice spreadsheet, Ior example, the equation would have the Iollowing Iormat:
45
d (sin H) cos Lat cos Decsint dt
cos H dH cos Lat cos Dec sint dt
dt 
cos H
cos Lat cos Decsint
dH
cos A:
dH
d Lat
cos Lat sin Dec  sin Lat cos Deccost
cos H
A: arccos
cos Lat sin Dec  sin Lat cos Deccost
cos H
A:
N
A: iI t 0° (180° LHA 360°)
360°  A: iI t > 0° (0° LHA 180°)
tan A:
d Lat
cos Lat dt
sin t
sin Lat cost cos Lat tan Dec
A: arctan
sint
sin Lat cost  cos Lat tan Dec
A:
N
A: iI numerator 0 AND denominator 0
A: +360° iI numerator >0 AND denominator 0
A: +180° iI denominator > 0
A:
N
DEGREES( Pi () + ATAN2 ( denominator , numerator) )
Altitude azimuth:
This Iormula, preIerred by many navigators, is directly derived Irom the navigational triangle (cosine law, chapter 10 &
chapter 11) without using diIIerential calculus.
As with the Iormula Ior timealtitude azimuth, Az
N
is obtained through application oI these rules:
In contrast to dH, H is a measurable quantity. Further, the position line is curved. Fig. 45 shows a macroscopic view
oI the line oI position, the azimuth line, and the circles oI equal altitude.
Procedure:
Although the theory oI the intercept method may look complicated at Iirst glance, the practical application is very
simple and does not require any background in diIIerential calculus. The procedure comprises the Iollowing steps:
1.
We choose an assumed position, AP (see Fig. 41), as near as possible to our estimated position. PreIerably, AP
should be deIined by an integer number oI degrees or arcminutes Ior Lat
AP
and Lon
AP
, respectively, depending on the
scale oI the chart. Our estimated position itselI may be used as well, but plotting a position line is easier when putting
AP on a point where two grid lines intersect.
2.
We calculate the meridian angle, t
AP
, (or local hour angle, LHA
AP
) Irom GHA and Lon
AP
, as shown above.
3.
We calculate the geocentric altitude oI the observed body as a Iunction oI Lat
AP
, t
AP
, and Dec (computed altitude):
4.
Using one oI the azimuth Iormulas stated above, we calculate the true azimuth oI the body, Az
N
, Irom Hc, Lat
AP
, t
AP
,
and Dec, Ior example:
46
cos A:
sin Dec sin H sin Lat
cos Hcos Lat
A: arccos
sin Dec  sin Hsin Lat
cos Hcos Lat
A:
N
A: iI t 0 (180° LHA 360°)
360°  A: iI t >0 (0° LHA 180°)
Hc arcsin (sin Lat
AP
sin Dec +cos Lat
AP
cos Deccos t
AP
)
A: arccos
sin Dec  sin Hcsin Lat
AP
cos Hccos Lat
AP
5.
We calculate the intercept, H, the diIIerence between observed altitude, Ho (chapter 2), and computed altitude, Hc.
The intercept, which is directly proportional to the diIIerence between the radii oI the corresponding circles oI equal
altitude, is usually expressed in nautical miles:
6.
We take a chart or plotting sheet with a convenient scale (depending on the respective scenario), and draw a suitable
length oI the azimuth line through AP (Fig. 46). On this line, we measure the intercept, H, Irom AP (towards GP iI
H~0, away Irom GP iI H·0) and draw a perpendicular through the point thus located. This perpendicular to the
azimuth line is our approximate line oI position (the red line in Fig. 46).
7.
To obtain our position, we need at least one additional position line. We repeat the procedure with altitude and GP oI a
second celestial body or oI the same body at a diIIerent time oI observation (Fig. 47). The point where both position
lines (tangents) intersect is our Iix. The second observation does not necessarily require the same AP to be used.
As mentioned above, the intercept method ignores the curvatures oI the actual position lines. ThereIore, the obtained Iix
is not our exact position but rather an improved position (compared with AP). The residual error remains tolerable as
long as the radii oI the circles oI equal altitude are not too small and AP is not too Iar Irom our actual position (chapter
16). The geometric error inherent to the intercept method can be decreased by iteration, i.e., substituting the obtained
Iix Ior AP and repeating the calculations (same altitudes and GP's). This will result in a more accurate position. II
necessary, we can reiterate the procedure until the obtained position remains virtually constant.
47
A:
N
A: iI t 0 (180° LHA 360°)
360°  A: iI t >0 (0° LHA 180°)
AH nm 60( Ho °  Hc ° )
Since an estimated position is usually nearer to our true position than an assumed position, the latter may require a
greater number oI iterations. Accuracy is also improved by observing three bodies instead oI two. Theoretically, the
position lines should intersect each other at a single point. Since no observation is entirely Iree oI errors, we will usually
obtain three points oI intersection Iorming an error triangle (Fig. 48).
Area and shape oI the triangle give us a rough estimate oI the quality oI our observations (chapter 16). Our most
probable position, MPP, is approximately represented by the 'center oI gravity' oI the error triangle (the point where
the bisectors oI the three angles oI the error triangle meet).
When observing more than three bodies, the resulting position lines will Iorm the corresponding polygons.
Direct Computation
II we do not want to plot lines oI position to determine our Iix, we can calculate the most probable position directly
Irom an unlimited number oI observations, n (n ~ 1). The Nautical Almanac provides an averaging procedure. First, the
auxiliary quantities A, B, C, D, E, and G have to be calculated:
48
A
_
i 1
n
cos
2
A:
i
B
_
i 1
n
sin A:
i
cos A:
i
C
_
i 1
n
sin
2
A:
i
D
_
i 1
n
(AH )
i
cos A:
i
E
_
i 1
n
(AH )
i
sin A:
i
G AC  B
2
In these Iormulas, Az
i
denotes the true azimuth oI the respective body. The H values are measured in degrees (same
unit as Lon and Lat). The geographic coordinates oI the observer's MPP are then obtained as Iollows:
The method does not correct Ior the geometric errors caused by the curvatures oI position lines. These are eliminated, iI
necessary, by iteration. For this purpose, we substitute the calculated MPP Ior AP. For each body, we calculate new
values Ior t (or LHA), Hc, H , and Az
N
. With these values, we recalculate A, B, C, D, E, G, Lon, and Lat.
Upon repeating this procedure, the resulting positions will converge rapidly. In the majority oI cases, less than two
iterations will be suIIicient, depending on the distance between AP and the true position.
Combining Different Lines of Position
Since the point oI intersection oI any two position lines, regardless oI their nature, marks the observer's geographic
position, one celestial LOP may suIIice to Iind one's position iI another LOP oI a diIIerent kind is available.
In the desert, Ior instance, we can determine our current position by Iinding the point on the map where a position line
obtained by observation oI a celestial object intersects the dirt road we are using (Fig. 49).
We can as well Iind our position by combining our celestial LOP with the bearing line oI a distant mountain peak or
any other prominent landmark (Fig. 410). B is the compass bearing oI the terrestrial object (corrected Ior magnetic
declination).
Both examples demonstrate the versatility oI position line navigation.
49
Lon Lon
AP
+
AE  BD
Gcos Lat
AP
Lat Lat
AP
+
C D  BE
G
Chapter 5
Finding the Position of a Moving Vessel
Celestial navigation on a moving vessel requires a correction Ior the change oI position between subsequent
observations unless the latter are perIormed in rapid succession or, better, simultaneously by a second observer.
II the navigator knows the speed oI the vessel, v, and the course over ground, C (the angle Iormed by the vector oI
motion and the local meridian), position line navigation provides a simple graphic solution.
Assuming that we make our Iirst observation at the time T
1
and our second observation at T
2
, the distance, d, traveled
during the time interval T
2
T
1
is
1 kn (knot) ÷ 1 nm/h
Although we have no knowledge oI our absolute position yet, we know our second position relative to the Iirst one,
deIined by C and d.
To Iind the absolute position, we plot both position lines in the usual manner, as illustrated in chapter 4. Next, we
choose an arbitrary point on the Iirst position line, LOP1, (resulting Irom the observation at T
1
) and advance this point
according to the motion vector with the length d and the direction C. Finally, we draw a parallel oI the Iirst position line
through the point thus located. The point where this advanced position line intersects the second line oI position
(resulting Irom the observation at T
2
) marks our position at the time T
2
. A position obtained in this Iashion is called
running fix (Fig. 51).
In a similar manner, we can obtain our position at T
1
by retiring the second position line, LOP2. In this case we have
to substitute C + 180° Ior C (Fig. 52).
51
d
Fig. 51
to GP2
AP
LOP1 (advanced)
LOP2
Fix
to GP1
d
LOP1
d
Fig. 52
to GP2
AP
LOP2
(retired) LOP2
Fix
to GP1
d
LOP1
d nm
T
2
h T
1
h
v kn
Terrestrial lines oI position may be advanced or retired in the same way as astronomical position lines.
It is also possible to choose two diIIerent assumed positions. AP1 should be close to the estimated position at T1, AP2
close to the estimated position at T2 (Fig. 53).
A running Iix is not as accurate as a stationary Iix. For one thing, course and speed over ground can only be estimated
since the eIIects oI current and wind (driIt) are not exactly known in most cases.
Further, there is a geometrical error inherent to the method. The latter is based on the assumption that each point oI the
circle oI equal altitude, representing a possible position oI the vessel, travels the same distance, d, along the rhumb line
(chapter 12) deIined by the course, C. The result oI such an operation, however, is a slightly distorted circle.
Consequently, an advanced or retired LOP is not exactly parallel to the original LOP. The resulting position error
usually increases as the distance, d, increases 19. The procedure gives Iairly accurate results when the distance
traveled between the observations is smaller than approx. 60 nm.
52
Fig. 53
AP2
LOP1
(advanced)
LOP2
Fix
to GP2
d
LOP1
to GP1
AP1
d
Chapter 6
Determination of Latitude and Longitude, Finding a Position by Direct Calculation
Latitude by Polaris
The geocentric altitude oI a celestial object being vertically above the geographic north pole would be numerically
equal to the latitude oI the observer (Fig. 61).
This is nearly the case with Polaris, the pole star. However, since the declination oI Polaris is not exactly 90° (89° 16.0'
in 2000.0), the altitude oI Polaris is inIluenced by the local hour angle. The altitude oI Polaris is also aIIected, to a
lesser degree, by nutation. To obtain the accurate latitude Irom the observed altitude, several corrections have to be
applied:
The corrections a
0
, a
1
, and a
2
, respectively, depend on LHA
Aries
, the observer's estimated latitude, and the number oI the
current month. They are given in the Polaris Tables oI the Nautical Almanac 12. To extract the data, the observer has
to know his approximate position and the approximate time.
When using a computer almanac instead oI the N. A., we can calculate Lat with the Iollowing simple procedure. Lat
E
is
our estimated latitude, Dec is the declination oI Polaris, and t is the meridian angle oI Polaris (calculated Irom GHA and
our estimated longitude). Hc is the computed altitude, Ho is the observed altitude (chapter 4).
Adding the altitude diIIerence, H, to the estimated latitude, we obtain the improved latitude:
The error oI Lat is smaller than 0.1' when Lat
E
is smaller than 70° and when the error oI Lat
E
is smaller than 2°,
provided the exact longitude is known. In polar regions, the algorithm becomes less accurate. However, the result can
be improved by iteration (substituting Lat Ior Lat
E
and repeating the calculation). Latitudes greater than 85° should be
avoided because a greater number oI iterations might be necassary. The method may lead to erratic results when the
observer is close to the north pole (Lat
E
Dec
Polaris
). An error in Lat resulting Irom an error in longitude is not
decreased by iteration. However, this error is always smaller than 1' when the error in longitude is smaller than 1°.
61
Lat Ho 1° + a
0
+ a
1
+ a
2
Hc arcsin (sin Lat
E
sin Dec + cos Lat
E
cos Dec cos t )
AH Ho  Hc
Lat
improved
 Lat
E
+ AH
Noon Latitude (Latitude by Maximum Altitude)
This is a very simple method enabling the observer to determine the latitude by measuring the maximum altitude oI the
sun (or any other object). A very accurate time measurement is not required. The altitude oI the sun passes through a
Ilat maximum approximately (see noon longitude) at the moment oI upper meridian passage (local apparent noon,
LAN) when t equals 0 and the GP oI the sun is either north or south oI the observer, depending on the declination oI the
sun and observer`s geographic latitude. The observer`s latitude is easily calculated by Iorming the algebraic sum or
diIIerence oI the declination and observed zenith distance z (90°Ho) oI the sun, depending on whether the sun is north
or south oI the observer (Fig. 62).
1. Sun south oI observer (Fig. 62a): Lat Dec + : Dec  Ho +90°
2. Sun north oI observer (Fig. 62b): Lat Dec  : Dec + Ho 90°
Northern declination is positive, southern declination negative.
BeIore starting the observations, we need a rough estimate oI our current longitude to know the time oI meridian
transit. We look up the time (UT) oI Greenwich meridian transit oI the sun on the daily page oI the Nautical Almanac
and add 4 minutes Ior each degree oI western longitude or subtract 4 minutes Ior each degree oI eastern longitude. To
determine the maximum altitude, we start observing the sun approximately 15 minutes beIore meridian transit. We
Iollow the increasing altitude oI the sun with the sextant, note the maximum altitude when the sun starts descending
again, and apply the usual corrections.
We look up the declination oI the sun at the approximate time (UT) oI local meridian passage on the daily page oI the
Nautical Almanac and apply the appropriate Iormula.
Historically, noon latitude and latitude by Polaris are among the oldest methods oI celestial navigation.
ExMeridian Sight
Sometimes, it may be impossible to measure the maximum altitude oI the sun. For example, the sun may be obscured
by a cloud at this instant. II we have a chance to measure the altitude oI the sun a Iew minutes beIore or aIter meridian
transit, we are still able to Iind our exact latitude by reducing the observed altitude to the meridian altitude, provided we
know our exact longitude (see below) and approximate latitude. The method is similar to the one used with the pole
star. First, we need the time (UT) oI local meridian transit (eastern longitude is positive, western longitude negative):
The meridian angle oI the sun, t, is calculated Irom the time oI observation (GMT):
62
T
Transit
h 12  EoT h 
Lon °
15
t ° 15
(
T
Observation
h T
Transit
h
)
Starting with our estimated Latitude, Lat
E
, we calculate the altitude oI the sun at the time oI observation. We use the
altitude Iormula Irom chapter 4:
Dec reIers to the time oI observation. We calculate the diIIerence between observed and calculated altitude:
We calculate an improved latitude, Lat
improved
:
(sun north oI observer: ¹H, sun south oI observer: H)
The exact latitude is obtained by iteration, i. e., we substitute Lat
improved
Ior Lat
E
and repeat the calculations until the
obtained latitude is virtually constant. Usually, no more than one or two iterations are necessary. The method has a Iew
limitations and requires critical judgement. The meridian angle should be small compared with the zenith distance oI
the sun. Otherwise, a greater number oI iterations may be necessary. The method may yield erratic results iI Lat
E
is
similar to Dec. A sight should be discarded when the observer is not sure iI the sun is north or south oI his position.
The inIluence oI a longitude error on the latitude thus obtained is not decreased by iteration.
Latitude by two altitudes
Even iI no estimated longitude is available, the exact latitude can still be Iound by observation oI two celestial bodies.
The required quantities are Greenwich hour angle, declination, and observed altitude oI each body 7.
The calculations are based upon spherical triangles (see chapter 10 & chapter 11). In Fig. 63, P
N
denotes the north
pole, O the observer`s unknown position, GP
1
the geographic position oI the Iirst body, and GP
2
the position oI the
second body.
First, we consider the spherical triangle GP
1
, P
N
, GP
2
. Fig. 63 shows only one oI several possible conIigurations. O
may as well be outside the triangle GP
1
, P
N
, GP
2
. We Iorm the diIIerence oI both Greenwich hour angles, GHA:
Using the law oI cosines Ior sides (chapter 10), we calculate d, the great circle distance between GP
1
and GP
2
:
Now we solve the same triangle Ior the angle , the horizontal distance between P
N
and GP
2
, measured at GP
1
:
For the spherical triangle GP
1
, O, GP
2
, we calculate the angle , the horizontal distance between O and GP
2
, measured
at GP
1
.
63
Hc arcsin (sin Lat
E
sin Dec + cos Lat
E
cos Dec cos t )
AH Ho  Hc
Lat
improved
 Lat
E
AH
AGHA ¸GHA
2
GHA
1
¸
cos d sin Dec
1
sin Dec
2
+cos Dec
1
cos Dec
2
cos(AGHA )
d arccos
sin Dec
1
sin Dec
2
+cos Dec
1
cos Dec
2
cos(AGHA )
coso
sin Dec
2
sin Dec
1
cosd
cos Dec
1
sind
o arccos
sin Dec
2
sin Dec
1
cos d
cos Dec
1
sin d
cos j
sin H
2
sin H
1
cos d
cos H
1
sin d
We calculate the angle , the horizontal distance between P
N
and O, measured at GP
1
. There are two solutions (
1
and
2
) since cos ÷ cos ( ):
The circles oI equal altitude intersect each other at two points. The corresponding positions are on opposite sides oI the
great circle going through GP
1
and GP
2
(not shown in Fig. 63). Using the law oI cosines Ior sides again, we solve the
spherical triangle GP
1
, P
N
, O Ior Lat. Since we have two solutions Ior , we obtain two possible latitudes, Lat
1
and
Lat
2
.
We choose the value nearest to our estimated latitude. The other one is discarded. II both solutions are very similar and
a clear distinction is not possible, one oI the sights should be discarded, and a body with a more Iavorable position
should be chosen.
Although the method requires more complicated calculations than, e. g., a latitude by Polaris, it has the advantage that
measuring two altitudes usually takes less time than Iinding the maximum altitude oI a single body. Moreover, iI Iixed
stars are observed, even a chronometer error oI several hours has no signiIicant inIluence on the resulting latitude since
GHA and both declinations change very slowly in this case.
When the horizontal distance between the observed bodies is in the vicinity oI 0° or 180°, the observer's position is
close to the great circle going through GP
1
and GP
2
. In this case, the two solutions Ior latitude are similar, and Iinding
which one corresponds with the actual latitude may be diIIicult (depending on the quality oI the estimate). The resulting
latitudes are also close to each other when the observed bodies have approximately the same Greenwich hour angle.
Noon Longitude (Longitude by Equal Altitudes)
Since the earth rotates with an angular velocity oI exactly 15° per hour with respect to the mean sun, the time oI local
meridian transit (local apparent noon) oI the sun, T
Transit
, can be used to calculate the observer's longitude:
T
Transit
is measured as UT (decimal Iormat). The correction Ior EoT at the time oI meridian transit, EoT
Transit
, has to be
made because the apparent sun, not the mean sun, is observed (see chapter 3). Since the Nautical Almanac contains
only values Ior EoT (see chapter 3) at 0:00 UT and 12:00 UT oI each day, EoT
Transit
has to be Iound by interpolation.
Since the altitude oI the sun  like the altitude oI any celestial body  passes through a rather Ilat maximum, the time oI
peak altitude is diIIicult to measure. The exact time oI meridian transit can be derived, however, Irom the times oI two
equal altitudes oI the sun.
Assuming that the sun moves along a symmetrical arc in the sky, T
Transit
is the mean oI the times corresponding with a
chosen pair oI equal altitudes oI the sun, one occurring beIore LAN, T
1
, the other past LAN, T
2
(Fig. 64).
64
j arccos
sin H
2
sin H
1
cosd
cos H
1
sin d
µ
1
¸o j¸ µ
2
o+j
sin Lat
1
sin H
1
sin Dec
1
+cos H
1
cos Dec
1
cos µ
1
Lat
1
arcsin (sin H
1
sin Dec
1
+ cos H
1
cos Dec
1
cosµ
1
)
sin Lat
2
sin H
1
sin Dec
1
+ cos H
1
cos Dec
1
cosµ
2
Lat
2
arcsin (sin H
1
sin Dec
1
+cos H
1
cos Dec
1
cos µ
2
)
Lon° 15
(
12  T
Transit
h  EoT
Transit
h
)
T
Transit
T
1
+T
2
2
In practice, the times oI equal altitudes oI the sun are measured as Iollows:
In the morning, the observer records the time T
1
corresponding with a chosen altitude, H. In the aIternoon, the time T
2
is recorded when the descending sun passes through the same altitude again. Since only times oI equal altitudes are
measured, no altitude correction is required. The interval T
2
T
1
should be greater than approx. 2 hours.
UnIortunately, the arc oI the sun is only symmetrical with respect to T
Transit
iI the sun's declination is constant during
the observation interval. This is approximately the case around the times oI the solstices. During the rest oI the year,
particularly at the times oI the equinoxes, T
Transit
diIIers signiIicantly Irom the mean oI T
1
and T
2
due to the changing
declination oI the sun. Fig. 65 shows the altitude oI the sun as a Iunction oI time and illustrates how the changing
declination aIIects the apparent path oI the sun in the sky, resulting in a time diIIerence, T.
The blue line shows the path oI the sun Ior a given, constant declination, Dec
1
. The red line shows how the path would
look with a diIIerent declination, Dec
2
. In both cases, the apparent path oI the sun is symmetrical with respect to T
Transit
.
However, iI the sun's declination varies Irom Dec
1
at T
1
to Dec
2
at T
2
, the path shown by the green line will result.
Now, T
1
and T
2
are no longer symmetrical to T
Transit
. The sun's meridian transit occurs beIore (T
1
¹T
2
)/2 iI the sun's
declination changes toward the observer's parallel oI latitude, like shown in Fig. 65. Otherwise, the meridian transit
occurs aIter (T
1
¹T
2
)/2. Since time and local hour angle (or meridian angle) are proportional to each other, a systematic
error in longitude results.
The error in longitude is negligible around the times oI the solstices when Dec is almost constant, and is greatest (up to
several arcminutes) at the times oI the equinoxes when the rate oI change oI Dec is greatest (approx. 1'/h). Moreover,
the error in longitude increases with the observer's latitude and may be quite dramatic in polar regions.
The obtained longitude can be improved, iI necessary, by application oI the equation of equal altitudes 5:
t is the change in the meridian angle, t, which cancels the change in altitude resulting Irom a small change in
declination, Dec. Lat is the observer's latitude. II the accurate latitude is not known, an estimated latitude may be used.
t
2
is the meridian angle oI the sun at T
2
. Since we do not know the exact value Ior t
2
initially, we start our calculations
with an approximate value calculated Irom T
1
and T
2
:
65
AT 
(
tan Lat
sint
2

tan Dec
2
tant
2
)
A Dec ADec Dec
2
 Dec
1
t
2
° 
15
(
T
2
h T
1
h
)
2
We denote the improved value Ior T
2
by T
2
*.
At T
2
*, the sun would pass through the same altitude as measured at T
1
iI Dec did not change during the interval oI
observation. Accordingly, the improved time oI meridian transit is:
The residual error resulting Irom the initial error oI t
2
is usually not signiIicant. It can be decreased, iI necessary, by
iteration. Substituting T
2
* Ior T
2
, we get the improved meridian angle, t
2
*:
With the improved meridian angle t
2
*, we calculate the improved correction t *:
Finally, we obtain a more accurate time value, T
2
**:
And, accordingly:
The error oI Dec should be as small as possible. Calculating Dec with a highprecision computer almanac is
preIerable to extracting it Irom the Nautical Almanac. When using the Nautical Almanac, Dec should be calculated
Irom the daily change oI declination to keep the rounding error as small as possible.
Although the equation oI equal altitudes is strictly valid only Ior an inIinitesimal change oI Dec, dDec, it can be used
Ior a measurable change, Dec, (up to several arcminutes) as well without sacriIicing much accuracy. Accurate time
measurement provided, the residual error in longitude rarely exceeds +0.1'.
Theory of the Equation of Equal Altitudes
The equation oI equal altitudes is derived Irom the altitude Iormula (see chapter 4) using differential calculus:
First, we need to know how a small change in declination would aIIect sin H. We Iorm the partial derivative with
respect to Dec:
66
T
2
* h T
2
h AT h T
2
h 
At °
15
T
Transit
T
1
+T
2
*
2
t
2
*° 
15
(
T
2
* h T
1
h
)
2
At * 
(
tan Lat
sin t
2
*

tan Dec
2
tan t
2
*
)
ADec
T
2
** h T
2
h 
At * °
15
T
Transit
T
1
+T
2
**
2
sin H sin Lat sin Dec + cos Lat cos Dec cos t
c(sin H)
c Dec
sin Lat cos Dec  cos Lat sin Dec cos t
Thus, the change in sin H caused by an inIinitesimal change in declination, d Dec, is:
Now, we Iorm the partial derivative with respect to t in order to Iind out how a small change in the meridian angle
would aIIect sin H:
The change in sin H caused by an inIinitesimal change in the meridian angle, dt, is:
Since we want both eIIects to cancel each other, the total diIIerential has to be zero:
Longitude Measurement on a Moving Vessel
On a moving vessel, we have to take into account not only the inIluence oI varying declination but also the eIIects oI
changing latitude and longitude on the altitude oI the body during the observation interval. DiIIerentiating sin H
(altitude Iormula) with respect to Lat, we get:
Again, the total diIIerential is zero because the combined eIIects oI latitude and meridian angle cancel each other with
respect to their inIluence on sin H:
67
c(sin H)
c Dec
d Dec (sin Lat cos Dec cos Lat sin Deccost )d Dec
c(sin H)
ct
cos Lat cos Decsint
c(sin H)
ct
d t cos Lat cos Dec sint d t
c(sin H)
c Dec
d Dec +
c(sin H)
ct
d t 0

c(sin H)
ct
d t
c(sin H )
cDec
d Dec
cos Lat cos Decsint d t (sin Lat cos Dec  cos Lat sin Deccost )d Dec
d t
sin Lat cos Dec cos Lat sin Dec cos t
cos Lat cos Decsin t
d Dec
d t
(
tan Lat
sin t

tan Dec
tan t
)
d Dec
At 
(
tan Lat
sin t

tan Dec
tan t
)
ADec
c(sin H)
cLat
cos Lat sin Dec sin Lat cos Deccos t
c(sinH)
cLat
d Lat +
c(sinH)
ct
dt 0
In analogy with a change in declination, we obtain the Iollowing Iormula Ior a small change in latitude:
The correction Ior the combined variations in Dec, Lat, and Lon is:
Lat and Lon are the small changes in latitude and longitude corresponding with the path oI the vessel traveled
between T
1
and T
2
. The meridian angle, t
2
, has to include a correction Ior Lon:
Lat and Lon are calculated Irom the course over ground, C, the velocity over ground, v, and the time elapsed.
C is measured clockwise Irom true north (0°...360°). Again, the corrected time oI equal altitude is:
The longitude calculated Irom T
Transit
reIers to the observer's position at T1. The longitude at T2 is Lon¹Lon.
The longitude error caused by a change in latitude can be dramatic and requires the navigator's particular attention,
even iI the vessel travels at a moderate speed. The above considerations clearly demonstrate that determining one's
exact longitude by equal altitudes oI the sun is not as simple as it seems to be at Iirst glance, particularly on a moving
vessel. It is thereIore quite natural that with the development oI position line navigation (including simple graphic
solutions Ior a traveling vessel), longitude by equal altitudes became less important.
The Meridian Angle of the Sun at Maximum Altitude
Fig. 65 shows that the maximum altitude oI the sun is slightly diIIerent Irom the altitude at the moment oI meridian
passage iI the declination changes. At maximum altitude, the rate oI change oI altitude caused by the changing
declination cancels the rate oI change oI altitude caused by the changing meridian angle.
The equation oI equal altitude enables us to calculate the meridian angle oI the sun at this moment. We divide each side
oI the equation by the inIinitesimal time interval dT:
68
dt
(
tan Dec
sin t

tan Lat
tant
)
d Lat
ALat ' v kn cosC
(
T
2
h  T
1
h
)
Lat
2
Lat
1
+ ALat
ALon ' v kn
sinC
cos Lat
(
T
2
h T
1
h
)
Lon
2
Lon
1
+ALon
1kn(knot ) 1 nm h
T
2
* h T
2
h 
At °
15
T
Transit
T
1
+T
2
*
2
dt
d T
(
tan Lat
sin t

tan Dec
tan t
)
d Dec
dT
At 
(
tan Lat
2
sint
2

tan Dec
2
tan t
2
)
ADec +
(
tan Dec
2
sint
2

tan Lat
2
tan t
2
)
ALat ALon
t
2
° 
15
(
T
2
h T
1
h
)
ALon°
2
Measuring the rate oI change oI t and Dec in arcminutes per hour we get:
Since t is a very small angle, we can substitute tan t Ior sin t:
Now, we can solve the equation Ior tan t:
Since a small angle (in radians) is nearly equal to its tangent, we get:
Measuring t in arcminutes, the equation is stated as:
dDec/dT is the rate oI change oI declination measured in arcminutes per hour.
The maximum altitude occurs aIter meridian transit iI t is positive, and beIore meridian transit iI t is negative.
For example, at the time oI the spring equinox (Dec ≈ 0, dDec/dT ≈ ¹1'/h) an observer being at ¹80° (N) latitude would
observe the maximum altitude oI the sun at t ≈ ¹21.7', i. e., 86.8 seconds aIter local meridian transit (LAN). An
observer at ¹45° latitude, however, would observe the maximum altitude at t ≈ +3.82', i. e., only 15.3 seconds aIter
meridian transit.
The Maximum Altitude of the Sun
We can use the last equation to evaluate the systematic error oI a noon latitude. The latter is based upon the maximum
altitude oI the sun, not on the altitude at the moment oI meridian transit. Following the above example, the observer at
80° latitude would observe the maximum altitude 86.7 seconds aIter meridian transit. During this interval, the
declination oI the sun would have changed Irom 0 to ¹1.445'' (assuming that Dec is 0 at the time oI meridian transit).
Using the altitude Iormula (chapter 4), we get:
In contrast, the calculated altitude at meridian transit would be exactly 10°. Thus, the error oI the noon latitude would
be 0.72''.
In the same way, we can calculate the maximum altitude oI the sun observed at 45° latitude:
In this case, the error oI the noon latitude would be only 0.13''.
The above examples show that even at the times oI the equinoxes, the systematic error oI a noon latitude caused by the
changing declination oI the sun is not signiIicant because it is much smaller than other observational errors, e. g., the
errors in dip or reIraction. A measurable error in latitude can only occur iI the observer is very close to one oI the poles
(tan Lat!). Around the times oI the solstices, the error in latitude is practically nonexistent.
69
[ ]
[ ] h T d
Dec d Dec Lat
t
'
900
tan tan
tan ⋅
−
≈
900' h
(
tanLat
sin t

tan Dec
tan t
)
d Dec '
d T h
900 
tan Lat  tan Dec
tant
d Dec '
d T h
t °
n
180

tan Lat  tan Dec
900
d Dec '
d T h
t '  3.82(tan Lat  tan Dec)
d Dec '
dT h
Hc arcsin (sin80°sin1.445' ' +cos80°cos1.445' ' cos 21.7' ) 10° 0' 0.72' '
Hc arcsin (sin45° sin0.255' ' + cos45°cos0.255' ' cos3.82' ) 45° 0' 0.13' '
Time Sight
The process oI deriving the longitude Irom a single altitude oI a body (as well as the observation made Ior this purpose)
is called time sight. However, this method requires knowledge oI the exact latitude, e. g., a noon latitude. Solving the
navigational triangle (chapter 11) Ior the meridian angle, t, we get:
The equation has two solutions, ¹t and t, since cos t ÷ cos (t). Geometrically, this corresponds with the Iact that the
circle oI equal altitude intersects the parallel oI latitude at two points.
Using the Iollowing Iormulas and rules, we obtain the longitudes oI these points oI intersection, Lon
1
and Lon
2
:
Even iI we do not know the exact latitude, we can still use a time sight to derive a line oI position Irom an assumed
latitude. AIter solving the time sight, we plot the assumed parallel oI latitude and the calculated meridian.
Next, we calculate the azimuth oI the body with respect to the position thus obtained (azimuth Iormulas, chapter 4) and
plot the azimuth line. Our line oI position is the perpendicular oI the azimuth line going through the calculated position
(Fig. 66).
The latter method is oI historical interest only. The modern navigator will certainly preIer the intercept method (chapter
4) which can be used without any restrictions regarding meridian angle (local hour angle), latitude, and declination (see
below).
A time sight is not reliable when the body is close to the meridian. Using diIIerential calculus, we can demonstrate that
the error oI the meridian angle, dt, resulting Irom an altitude error, dH, varies in proportion with 1/sin t:
Moreover, dt varies inversely with cos Lat and cos Dec. ThereIore, high latitudes and declinations should be avoided as
well. The same restrictions apply to Sumner´s method which is based upon two time sights.
610
t arccos
sinHo sin Lat sinDec
cos Lat cos Dec
Lon
1
t  GHA
Lon
2
360° t GHA
If Lon
1
180°  Lon
1
+360°
If Lon
2
180°  Lon
2
+360°
If Lon
2
>+180°  Lon
2
360°
dt 
cos Ho
cos Lat cos Decsint
dH
Direct Computation of Position
II we know the exact time, the observations Ior a latitude by two altitudes even enable us to calculate our position
directly, without any graphic plot. AIter obtaining our latitude, Lat, Irom two altitudes (see above), we use the time
sight Iormula to calculate the meridian angle oI one oI the bodies. In case oI the Iirst body, Ior example, we calculate t
1
Irom the quantities Lat, Dec
1
, and H
1
(see Fig. 63). Two possible longitudes result Irom the meridian angle thus
obtained. We choose the one nearest to our estimated longitude. This is a rigorous method, not an approximation.
Direct computation was rarely used in the past since the calculations are more complicated than those required Ior
graphic solutions. OI course, in the age oI computers the complexity oI the method does not pose a problem anymore.
611
Chapter 7
Finding Time and Longitude by Lunar Distances
In celestial navigation, time and longitude are interdependent. Finding one`s longitude at sea or in unknown terrain is
impossible without knowing the exact time and vice versa. ThereIore, oldtime navigators were basically restricted to
latitude sailing on long voyages, i. e., they had to sail along a chosen parallel oI latitude until they came in sight oI the
coast. Since there was no reliable estimate oI the time oI arrival, many ships ran ashore during periods oI darkness or
bad visibility. Spurred by heavy losses oI men and material, scientists tried to solve the longitude problem by using
astronomical events as time marks. In principle, such a method is only suitable when the observed time oI the event is
virtually independent oI the observer`s geographic position.
Measuring time by the apparent movement oI the moon with respect to the background oI Iixed stars was suggested in
the 15
th
century already (Regiomontanus) but proved impracticable since neither reliable ephemerides Ior the moon nor
precise instruments Ior measuring angles were available at that time.
Around the middle oI the 18
th
century, astronomy and instrument making had Iinally reached a stage oI development
that made time measurement by lunar observations possible. Particularly, deriving the time Irom a socalled lunar
distance, the angular distance oI the moon Irom a chosen reIerence body, became a popular method. Although the
procedure is rather cumbersome, it became an essential part oI celestial navigation and was used Iar into the 19
th
century, long aIter the invention oI the mechanical chronometer (Harrison, 1736). This was mainly due to the limited
availability oI reliable chronometers and their exorbitant price. When chronometers became aIIordable around the
middle oI the 19
th
century, lunar distances gradually went out oI use. Until 1906, the Nautical Almanac included lunar
distance tables showing predicted geocentric angular distances between the moon and selected bodies in 3hour
intervals.* AIter the tables were dropped, lunar distances Iell more or less into oblivion. Not much later, radio time
signals became available worldwide, and the longitude problem was solved once and Ior all. Today, lunar distances are
mainly oI historical interest. The method is so ingenious, however, that a detailed study is worthwhile.
The basic idea oI the lunar distance method is easy to comprehend. Since the moon moves across the celestial sphere at
a rate oI about 0.5° per hour, the angular distance between the moon, M, and a body in her path, B, varies at a similar
rate and rapidly enough to be used to measure the time. The time corresponding with an observed lunar distance can be
Iound by comparison with tabulated values.
Tabulated lunar distances are calculated Irom the geocentric equatorial coordinates oI M and B using the cosine law:
or
D is the geocentric lunar distance. These Iormulas can be used to set up one`s own table with the aid oI the
Nautical Almanac or any computer almanac iI a lunar distance table is not available.
*Almost a century aIter the original Lunar Distance Tables were dropped, Steven Wepster resumed the tradition.
His tables are presently (2004) available through the internet 14.
Clearing the lunar distance
BeIore a lunar distance measured by the observer can be compared with tabulated values, it has to be reduced to the
corresponding geocentric angle by clearing it Irom the eIIects oI reIraction and parallax. This essential process is called
clearing the lunar distance. Numerous procedures have been developed, among them rigorous and 'quick¨ methods.
In the Iollowing, we will discuss the almost identical methods by Dunthorne (1766) and Young (1856). They are
rigorous Ior a spherical model oI the earth.
71
cos D sin Dec
M
sin Dec
B
+ cos Dec
M
cos Dec
B
cos(GHA
M
GHA
B
)
cos D sin Dec
M
sin Dec
B
+ cos Dec
M
cos Dec
B
cos
15
(
RA
M
h  RA
B
h
)
Fig. 71 shows the positions oI the moon, M, and a reIerence body, B, in the coordinate system oI the horizon. We
denote the apparent positions oI the centers oI the moon and the reIerence body by M
app
and B
app
, respectively. Z is the
zenith.
The side D
app
oI the spherical triangle B
app
ZM
app
is the apparent lunar distance. The altitudes oI Mapp and Bapp
(obtained aIter applying the corrections Ior index error, dip, and semidiameter) are H
Mapp
and H
Bapp
, respectively. The
vertical circles oI both bodies Iorm the angle , the diIIerence between the azimuth oI the moon, Az
M
, and the azimuth
oI the reIerence body, Az
B
:
The position oI each body is shiIted along its vertical circle by atmospheric reIraction and parallax in altitude. AIter
correcting HMapp and HBapp Ior both eIIects, we obtain the geocentric positions M and B. We denote the altitude oI M by
HM and the altitude oI B by HB. HM is always greater than HMapp because the parallax oI the moon is always greater
than reIraction. The angle is neither aIIected by reIraction nor by the parallax in altitude:
The side D oI the spherical triangle BZM is the unknown geocentric lunar distance. II we knew the exact value Ior ,
calculation oI D would be very simple (cosine law). UnIortunately, the navigator has no means Ior measuring
precisely. It is possible, however, to calculate D solely Irom the Iive quantities Dapp, HMapp, HM, HBapp, and HB.
Applying the cosine Iormula to the spherical triangle Iormed by the zenith and the apparent positions, we get:
Repeating the procedure with the spherical triangle Iormed by the zenith and the geocentric positions, we get:
coso
cos D sin H
M
sin H
B
cos H
M
cos H
B
72
o Az
M
 Az
B
Az
M
Az
Mapp
Az
B
Az
Bapp
cos D
app
sin H
Mapp
sin H
Bapp
+ cos H
Mapp
cos H
Bapp
coso
coso
cos D
app
sin H
Mapp
sinH
Bapp
cosH
Mapp
cos H
Bapp
cos D sin H
M
sin H
B
+cos H
M
cos H
B
coso
Since is constant, we can combine both azimuth Iormulas:
Thus, we have eliminated the unknown angle . Now, we subtract unity Irom both sides oI the equation:
Using the addition Iormula Ior cosines, we have:
Solving Ior cos D, we obtain Dunthornes Iormula Ior clearing the lunar distance:
Adding unity to both sides oI the equation instead oI subtracting it, leads to Youngs Iormula:
Procedure
Deriving UT Irom a lunar distance comprises the Iollowing steps:
1.
We measure the altitude oI the upper or lower limb oI the moon, whichever is visible, and note the time oI the
observation indicated by our watch, WT1
LMapp
.
We apply the corrections Ior index error and dip (iI necessary) and get the apparent altitude oI the limb, H1
LMapp
. We
repeat the procedure with the reIerence body and obtain the watch time WT1
Bapp
and the altitude H1
Bapp
.
2.
We measure the angular distance between the limb oI the moon and the reIerence body, D
Lapp
, and note the
corresponding watch time, WT
D
. The angle D
Lapp
has to be measured with the greatest possible precision. It is
recommended to measure a Iew D
Lapp
values and their corresponding WT
D
values in rapid succession and calculate the
respective average value. When the moon is almost Iull, it is not quite easy to distinguish the limb oI the moon Irom the
terminator (shadow line). In general, the limb has a sharp appearance whereas the terminator is slightly indistinct.
73
cos D sinH
M
sin H
B
cos H
M
cos H
B
cosD
app
sin H
Mapp
sinH
Bapp
cos H
Mapp
cosH
Bapp
cos D sinH
M
sin H
B
cos H
M
cos H
B
1
cos D
app
sin H
Mapp
sin H
Bapp
cos H
Mapp
cos H
Bapp
1
cos D  sinH
M
sin H
B
cos H
M
cos H
B

cos H
M
cos H
B
cos H
M
cos H
B
cos D
app
sin H
Mapp
sin H
Bapp
cos H
Mapp
cos H
Bapp

cos H
Mapp
cos H
Bapp
cos H
Mapp
cos H
Bapp
cos D sinH
M
sin H
B
 cos H
M
cos H
B
cos H
M
cos H
B
cos D
app
sin H
Mapp
sinH
Bapp
 cos H
Mapp
cos H
Bapp
cos H
Mapp
cos H
Bapp
cos D  cos(H
M
 H
B
)
cos H
M
cos H
B
cos D
app
 cos( H
Mapp
 H
Bapp
)
cos H
Mapp
cos H
Bapp
cos D
cos H
M
cos H
B
cos H
Mapp
cos H
Bapp
cos D
app
cos( H
Mapp
 H
Bapp
)
+cos( H
M
 H
B
)
cos D
cosH
M
cos H
B
cosH
Mapp
cos H
Bapp
cos D
app
+ cos( H
Mapp
+ H
Bapp
)
cos( H
M
+ H
B
)
3.
We measure the altitudes oI both bodies again, as described above. We denote them by H2
LMapp
and H2
Bapp
, and note
the corresponding watch times oI observation, WT2
LMapp
and WT2
Bapp
.
4.
Since the observations are only a Iew minutes apart, we can calculate the altitude oI the respective body at the moment
oI the lunar distance observation by linear interpolation:
5.
We correct the altitude oI the moon and the angular distance D
Lapp
Ior the augmented semidiameter oI the moon, SD
aug
.
The latter can be calculated directly Irom the altitude oI the upper or lower limb oI the moon:
The altitude correction is:
The rules Ior the lunar distance correction are:
The above procedure is an approximation since the augmented semidiameter is a Iunction oI the altitude corrected Ior
reIraction. Since reIraction is a small quantity and since the total augmentation between 0° and 90° altitude is only
approx. 0.3`, the resulting error is very small and may be ignored.
The sun, when chosen as reIerence body, requires the same corrections Ior semidiameter. Since the sun does not show a
measurable augmentation, we can use the geocentric semidiameter tabulated in the Nautical Almanac or calculated with
a computer program.
6.
We correct both altitudes, H
Mapp
and H
Bapp
, Ior atmospheric reIraction, R.
R
i
is subtracted Irom the respective altitude. The reIraction Iormula is only accurate Ior altitudes above approx. 10°.
Lower altitudes should be avoided anyway since reIraction may become erratic and since the apparent disk oI the moon
(and sun) assumes an oval shape caused by an increasing diIIerence in reIraction Ior upper and lower limb. This
distortion would aIIect the semidiameter with respect to the reIerence body in a complicated way.
74
H
LMapp
H1
LMapp
+( H2
LMapp
 H1
LMapp
)
WT
D
 WT1
LMapp
WT2
LMapp
WT1
LMapp
H
Bapp
H1
Bapp
+( H2
Bapp
 H1
Bapp
)
WT
D
WT1
Bapp
WT2
Bapp
WT1
Bapp
tan SD
aug
k
.
1
sin
2
HP
M
( cos H
LMapp
k)
2
 sinH
LMapp
k 0.2725
upper limb: cos H
LMapp
 k lower limb: cos H
LMapp
+ k
Lower limb: H
Mapp
H
LMapp
+ SD
aug
Upper limb: H
Mapp
H
LMapp
 SD
aug
Limb of moon towards reference body: D
app
D
Lapp
+ SD
aug
Limb of moon away from reference body: D
app
D
Lapp
 SD
aug
R
i
'
p mbar
1010
283
T °C + 273
(
0.97127
tan H
i

0.00137
tan
3
H
i
)
i Mapp , Bapp H
i
> 10°
7.
We correct the altitudes Ior the parallax in altitude:
We apply the altitude corrections as Iollows:
The correction Ior parallax is not applied to the altitude oI a Iixed star (HPB ÷ 0).
8.
With D
app
, H
Mapp
, H
M
, H
Bapp
, and H
B
, we calculate D using Dunthornes or Youngs Iormula.
9.
The time corresponding with the geocentric distance D is Iound by interpolation. Lunar distance tables show D as a
Iunction oI time, T (UT). II the rate oI change oI D does not vary too much (less than approx. 0.3` in 3 hours), we can
use linear interpolation. However, in order to Iind T, we have to consider T as a Iunction oI D (inverse interpolation).
TD is the unknown time corresponding with D. D
1
and D
2
are tabulated lunar distances. T
1
and T
2
are the corresponding
time (UT) values (T
2
÷ T
1
¹ 3h). D is the geocentric lunar distance calculated Irom D
app
. D has to be between D
1
and
D
2
.
II the rate oI change oI D varies signiIicantly, more accurate results are obtained with methods Ior nonlinear
interpolation, Ior example, with 3point Lagrange interpolation. Choosing three pairs oI tabulated values, (T
1
, D
1
), (T
2
,
D
2
), and (T
3
, D
3
), T
D
is calculated as Iollows:
D may have any value between D
1
and D
3
.
There must not be a minimum or maximum oI D in the time interval T
1
, T
3
. This problem does not occur with a
properly chosen body having a suitable rate oI change oI D. Near a minimum or maximum oI D, D/T would be very
small, and the observation would be erratic anyway.
AIter Iinding T
D
, we can calculate the watch error, T.
T is the diIIerence between our watch time at the moment oI observation, WT
D
, and the time Iound by interpolation,
T
D
.
Subtracting the watch error Irom the watch time, WT, results in UT.
75
sinP
M
sin HP
M
cos( H
Mapp
 R
Mapp
) sin P
B
sin HP
B
cos( H
Bapp
 R
Bapp
)
H
M
H
Mapp
 R
Mapp
+ P
M
H
B
H
Bapp
 R
Bapp
+ P
B
T
D
T
1
+(T
2
T
1
)
D  D
1
D
2
 D
1
T
D
T
1
( D  D
2
)( D D
3
)
( D
1
 D
2
)( D
1
 D
3
)
+ T
2
( D  D
1
)( D D
3
)
(D
2
 D
1
)( D
2
 D
3
)
+T
3
( D  D
1
)( D  D
2
)
( D
3
 D
1
)( D
3
 D
2
)
T
2
T
1
+ 3h T
3
T
2
+ 3h D
1
D
2
D
3
or D
1
> D
2
> D
3
AT WT
D
T
D
UT WT  AT
Improvements
The procedures described so Iar reIer to a spherical earth. In reality, however, the earth has approximately the shape oI
an ellipsoid Ilattened at the poles. This leads to small but measurable eIIects when observing the moon, the body
nearest to the earth. First, the parallax in altitude diIIers slightly Irom the value calculated Ior a spherical earth. Second,
there is a small parallax in azimuth which would not exist iI the earth were a sphere (see chapter 9). II no correction is
applied, D may contain an error oI up to approx. 0.2`. The Iollowing Iormulas reIer to an observer on the surIace oI the
reIerence ellipsoid (approximately at sea level).
The corrections require knowledge oI the observer`s latitude, Lat, the true azimuth oI the moon, Az
M
, and the true
azimuth oI the reIerence body, Az
B
.
Since the corrections are small, the three values do not need to be very accurate. Errors oI a Iew degrees are tolerable.
Instead oI the azimuth, the compass bearing oI each body, corrected Ior magnetic declination, may be used.
Parallax in altitude:
This correction is applied to the parallax in altitude and is used to calculate H
M
with higher precision beIore clearing the
lunar distance.
I is the Ilattening oI the earth: f
1
298.257
Parallax in azimuth:
The correction Ior the parallax in azimuth is applied aIter calculating H
M
and D. The Iollowing Iormula is a Iairly
accurate approximation oI the parallax in azimuth, Az
M
:
In order to Iind how Az
M
aIIects D, we go back to the cosine Iormula:
We diIIerentiate the equation with respect to :
76
AP
M
 f HP
M
sin( 2Lat )cos Az
M
sinH
Mapp
sin
2
Lat cos H
Mapp
P
M , improved
P
M
+ AP
M
H
M
H
mapp
 R
Mapp
+ P
M , improved
AAz
M
 f HP
M
sin(2Lat )sin Az
M
cos H
M
cos D sin H
M
sin H
B
+cos H
M
cos H
B
coso
d( cos D)
do
cos H
M
cos H
B
sino
d (cos D) sin Dd D
sinDd D cos H
M
cos H
B
sinod o
d D
cos H
M
cos H
B
sino
sin D
d o
Since d o d Az
M
, the change in D caused by an inIinitesimal change in Az
M
is:
With a small but measurable change in Az
M
, we have:
Combining the Iormulas Ior Az
M
and D, we get:
In most cases, D
improved
will be accurate to 0.1'' (provided the measurements are errorIree). The correction Iormula is
less accurate when the topocentric (~ apparent) positions oI moon and reIerence body are close (· 5°) together. The
Iormula should not be applied when the reIerence body is less than about Iour semidiameters (~1°) away Irom the
center oI the moon.
Accuracy
According to modern requirements, the lunar distance method is rather inaccurate. In the 18
th
and early 19
th
century,
however, this was generally accepted because a longitude with an error oI 0.5°1° was still better than no longitude
measurement at all. Said error is the approximate result oI an error oI only 1` in the measurement oI D
Lapp
, not
uncommon Ior a sextant reading under practical conditions. ThereIore, D
Lapp
should be measured with greatest care.
The altitudes oI both bodies do not quite require the same degree oI precision because a small error in the apparent
altitude leads to about the same error in the geocentric altitude. Since both errors cancel each other to a large extent, the
resulting error in D is comparatively small. An altitude error oI a Iew arcminutes is tolerable in most cases. ThereIore,
measuring two altitudes oI each body and Iinding the altitude at the moment oI the lunar distance observation by
interpolation is not absolutely necessary. Measuring a single altitude oI each body shortly beIore or aIter the lunar
distance measurement is suIIicient iI a small loss in accuracy is accepted.
The position oI the reIerence body with respect to the moon is crucial. The rate oI change oI D should not be too low. It
becomes zero when D passes through a minimum or maximum, making an observation useless. This can be checked
with lunar distance tables. Since the plane oI the lunar orbit Iorms a relatively small angle (approx. 5°) with the ecliptic,
bright bodies in the vicinity oI the ecliptic are most suitable (sun, planets, selected stars).
The stars generally recommended Ior the lunar distance method are Aldebaran, Altair, Antares, Fomalhaut, Hamal,
Markab, Pollux, Regulus, and Spica, but other stars close to the ecliptic may be used as well, e. g., Nunki. The lunar
distance tables oI the Nautical Almanac contained only D values Ior those bodies having a Iavorable position with
respect to the moon on the day oI observation.
77
d D
cosH
M
cos H
B
sin o
sinD
d Az
M
AD 
cos H
M
cos H
B
sino
sin D
AAz
M
D
improved
 D + AD
D
improved
 D + f HP
M
cos H
B
sin(2Lat )sin Az
M
sin( Az
M
 Az
B
)
sin D
Chapter 8
Rise, Set, Twilight
General Conditions for Visibility
For the planning oI observations, it is useIul to know the times during which a certain body is above the horizon as well
as the times oI sunrise, sunset, and twilight.
A body can be always above the horizon, always below the horizon, or above the horizon during a part oI the day,
depending on the observer's latitude and the declination oI the body.
A body is circumpolar (always above the celestial horizon) when the zenith distance is smaller than 90° at the moment
oI lower meridian passage, i. e., when the body is on the lower branch oI the local meridian (Fig 81a). This is the case
iI
A body is continually below the celestial horizon when the zenith distance is greater than 90° at the instant oI upper
meridian passage (Fig 81b). This is the case iI
A celestial body being on the same hemisphere as the observer is either sometimes above the horizon or circumpolar. A
body being on the opposite hemisphere is either sometimes above the horizon or permanently invisible, but never
circumpolar.
The sun provides a good example oI how the visibility oI a body is aIIected by latitude and declination. At the time oI
the summer solstice (Dec ÷ ¹23.5°), the sun is circumpolar to an observer being north oI the arctic circle (Lat ~
¹66.5°). At the same time, the sun remains below the celestial horizon all day iI the observer is south oI the antarctic
circle (Lat · −66.5°). At the times oI the equinoxes (Dec ÷ 0°), the sun is circumpolar only at the poles. At the time oI
the winter solstice (Dec ÷ 23.5°), the sun is circumpolar south oI the antarctic circle and invisible north oI the arctic
circle. II the observer is between the arctic and the antarctic circle, the sun is visible during a part oI the day all year
round.
Rise and Set
The events oI rise and set can be used to determine latitude, longitude, or time. One should not expect very accurate
results, however, since atmospheric reIraction may be erratic iI the body is on or near the horizon.
81
¸Lat + Dec¸ > 90°
¸Lat  Dec¸ > 90°
The geometric rise or set oI a body occurs when the center oI the body passes through the celestial horizon (H ÷ 0°).
Due to the inIluence oI atmospheric reIraction, all bodies except the moon appear above the visible and sensible
horizon at this instant. The moon is not visible at the moment oI her geometric rise or set since the depressing eIIect oI
the horizontal parallax (∼1°) is greater than the elevating eIIect oI atmospheric reIraction.
The approximate apparent altitudes (reIerring to the sensible horizon) at the moment oI the astronomical rise or set are:
Sun (lower limb): 15'
Stars: 29'
Planets: 29' − HP
When measuring these altitudes with reIerence to the sea horizon, we have to add the dip oI horizon (chapter 2) to the
above values. For example, the altitude oI the lower limb oI the rising or setting sun is approx. 20' iI the height oI eye is
8m.
We begin with the wellknown altitude Iormula (chapter 4).
Solving the equation Ior the meridian angle, t, we get :
The equation has no solution iI the argument oI the inverse cosine is smaller than −1 or greater than 1. In the Iirst case,
the body is circumpolar, in the latter case, the body remains continuously below the horizon. Otherwise, the arccos
Iunction returns values in the range Irom 0° through 180°.
Due to the ambiguity oI the arccos Iunction, the equation has two solutions, one Ior rise and one Ior set. For the
calculations below, we have to observe the Iollowing rules:
II the body is rising (body eastward Irom the observer), t is treated as a negative quantity.
II the body is setting (body westward Irom the observer), t is treated as a positive quantity.
II we know our latitude and the time oI rise or set, we can calculate our longitude:
GHA is the Greenwich hour angle oI the body at the moment oI rise or set. The sign oI t has to be observed careIully
(see above). II the resulting longitude is smaller than −180°, we add 360°.
Knowing our position, we can calculate the times oI sunrise and sunset:
The times oI sunrise and sunset obtained with the above Iormula are not quite accurate since Dec and EoT are variable.
Since we do not know the exact time oI rise or set at the beginning, we have to use estimated values Ior Dec and EoT
initially. The time oI rise or set can be improved by iteration (repeating the calculations with Dec and EoT at the
calculated time oI rise or set). Further, the times thus calculated are inIluenced by the irregularities oI atmospheric
reIraction near the horizon. ThereIore, a time error oI 2 minutes is not unusual.
Accordingly, we can calculate our longitude Irom the time oI sunrise or sunset iI we know our latitude:
Again, this is not a very precise method, and an error oI several arcminutes in longitude is not unlikely.
82
cos t 
sin Lat sin Dec
cos Lat cos Dec
t arccos (tan Lat tan Dec)
Lon GHA t
UT
Sunrise, Sunset
12  EoT 
Lon°
15
t °
15
Lon ° 15(12  GMT
Sunrise, Sunset
 EoT) t °
sin H sin Lat sin Dec +cos Lat cos Deccos t 0
Knowing our longitude, we are able to determine our approximate latitude Irom the time oI sunrise or sunset:
In navigation, rise and set are defined as the moments when the upper limb of a body is on the visible horizon.
These events can be observed without a sextant. Now, we have to take into account the eIIects oI reIraction, horizontal
parallax, dip, and semidiameter. These quantities determine the altitude (Ho) oI a body with respect to the celestial
horizon at the instant oI the visible rise or set.
By deIinition, the standard reIraction Ior a body being on the sensible horizon, R
H
, is 34' (it is subject to random
variations in reality).
When observing the upper limb oI the sun, we get:
Ho is negative. II we reIer to the upper limb oI the sun and the sensible horizon (Dip÷0), the meridian angle at the time
oI sunrise or sunset is:
Azimuth and Amplitude
The azimuth angle oI a rising or setting body is calculated with the azimuth Iormula (see chapter 4):
With H÷0, we get:
Az is ¹90° (rise) and −90° (set) iI the declination oI the body is zero, regardless oI the observer's latitude. Accordingly,
the sun rises in the east and sets in the west at the times oI the equinoxes (geometric rise and set).
With H
center
÷ −50' (upper limb oI the sun on the sensible horizon), we have:
The true azimuth oI the rising or setting body is:
The azimuth oI a body at the moment oI rise or set can be used to Iind the magnetic declination at the observer's
position (compare with chapter 13).
83
t ° Lon °  15(12 GMT
Sunrise , Sunset
 EoT )
Lat arctan
(

cost
tan Dec
)
t arccos
sinHo sin Lat sinDec
cos Lat cos Dec
Ho HP  SD  R
H
 Dip
Ho 0.15' 16'  34'  Dip  50'  Dip
t arccos
0.0145  sinLat sin Dec
cos Lat cos Dec
Az arccos
sin Dec  sinH sin Lat
cos Hcos Lat
Az arccos
sin Dec
cos Lat
Az arccos
sin Dec + 0.0145sinLat
0.9999cos Lat
Az
N
Az if t 0
360° Az if t > 0
The horizontal angular distance oI a rising or setting body Irom the east (rising) or west (setting) point on the horizon is
called amplitude and can be calculated Irom the azimuth. An amplitude oI E45°N, Ior instance, means that the body
rises 45° north oI the east point on the horizon.
Twilight
At sea, twilight is important Ior the observation oI stars and planets since it is the only time when these bodies and the
horizon are visible. By deIinition, there are three kinds oI twilight. The altitude, H, reIers to the center oI the sun and
the celestial horizon and marks the beginning (morning) and the end (evening) oI the respective twilight.
Civil twilight: H ÷ −6°
Nautical twilight: H ÷ −12°
Astronomical twilight: H ÷ −18°
In general, an altitude oI the sun between −3° and −9° is recommended Ior astronomical observations at sea (best
visibility oI brighter stars and sea horizon). However, exceptions to this rule are possible, depending on the actual
weather conditions and the brightness oI the observed body.
The meridian angle Ior the sun at −6° altitude (center) is:
Using this Iormula, we can Iind the approximate time Ior our observations (in analogy to sunrise and sunset).
As mentioned above, the simultaneous observation oI stars and the horizon is possible during a limited time interval
only.
To calculate the length oI this interval, AT, we use the altitude Iormula and diIIerentiate sin H with respect to the
meridian angle, t:
Substituting cosH
.
dH Ior d(sinH) and solving Ior dt, we get the change in the meridian angle, dt, as a Iunction oI a
change in altitude, dH:
With H ÷ −6° and dH AH ÷ 6° (H ÷ −3°...−9°), we get:
Converting the change in the meridian angle to a time span (measured in minutes) and ignoring the sign, the equation is
stated as:
84
t arccos
0.10453  sinLat sin Dec
cos Lat cos Dec
d(sin H)
dt
 cos Lat cos Dec sin t
d (sinH)  cos Lat cos Decsint dt
dt 
cos H
cos Lat cos Decsint
dH
At °  
5.97
cos Lat cos Dec sint
AT m 
24
cos Lat cos Decsint
The shortest possible time interval Ior our observations (Lat ÷ 0, Dec ÷ 0, t ÷ 96°) lasts approx. 24 minutes. As the
observer moves northward or southward Irom the equator, cos Lat and sin t decrease (t~90°). Accordingly, the duration
oI twilight increases. When t is 0° or 180°, AT is inIinite.
This is conIirmed by the wellknown Iact that the duration oI twilight is shortest in equatorial regions and longest in
polar regions.
We would obtain the same result when calculating t Ior H ÷ −3° and H ÷ −9°, respectively:
The Nautical Almanac provides tabulated values Ior the times oI sunrise, sunset, civil twilight and nautical twilight Ior
latitudes between −60° and ¹72° (reIerring to an observer being at the Greenwich meridian). In addition, times oI
moonrise and moonset are given.
85
AT m 4
(
t
9°
°  t
3°
°
)
Chapter 9
Geodetic Aspects of Celestial Navigation
The Ellipsoid
Celestial navigation is based upon the assumption that the earth is a sphere. Accordingly, calculations are based on the
laws oI spherical trigonometry. In reality, the shape oI the earth rather resembles an oblate spheroid (ellipsoid)
resulting Irom two Iorces, gravitation and centrifugal force, acting on the viscous body oI the earth. While gravitation
alone would Iorce the earth to assume the shape oI a sphere, the state oI lowest potential energy, the centriIugal Iorce
caused by the earth's rotation contracts the earth along the axis oI rotation (polar axis) and stretches it along the plane oI
the equator. The local vector sum oI both Iorces is called gravity.
There are several reIerence ellipsoids in use to describe the shape oI the earth, Ior example the World Geodetic System
ellipsoid oI 1984 (WGS 84). An important characteristic oI the WGS 84 ellipsoid is that its center conincides with the
mass center oI the earth. There are special reIerence ellipsoids whose centers are not identical with the mass center. OII
center ellipsoids are constructed to obtain a better Iit Ior a particular region. The Iollowing considerations reIer to the
WGS 84 ellipsoid which gives the best universal Iit and is accurate enough Ior the purpose oI navigation in most cases.
Fig.91 shows a meridional section oI the ellipsoid.
Earth data (WGS 84 ellipsoid) :
Equatorial radius
r
e
6378137.0 m
Polar radius
r
p
6356752.3142 m
Flattening
I ÷ (r
e
 r
p
) / r
e
1/298.25722
Due to the Ilattening oI the earth, we have to distinguish between geodetic and geocentric latitude oI a given position.
The geodetic latitude, Lat, is the angle between the local normal (perpendicular) to the surIace oI the reIerence ellipsoid
and the line oI intersection Iormed by the plane oI the equator and the plane oI the local meridian. The geocentric
latitude, Lat', is the angle Iormed by the local radius vector and said line oI intersection. Geodetic and geocentric
latitude are interrelated as Iollows:
II the earth were a sphere (I ÷ 0), geodetic and geocentric latitude would be the same. With the spheroid, both
quantities are equal only at the poles and on the equator. At all other places, the absolute value oI the geocentric latitude
is smaller than the absolute value oI the geodetic latitude. Due to the rotational symmetry of the ellipsoid, geodetic
and geocentric longitude are equal. Maps are usually based upon geodetic coordinates which are also reIerred to as
geographic coordinates 1*. In this context it should be mentioned that the term 'geographic position', applied to a
celestial body, is misleading, since Greenwich hour angle and declination are geocentric coordinates (see chapter 3).
*In other publications, e. g., 10, astronomical coordinates (see below) and geographic coordinates are considered as identical.
91
tan Lat ' (1f )
2
tan Lat
In the Iollowing, we will discuss the eIIects oI the oblateness (Ilattening) oI the earth on celestial navigation.
Any zenith distance (and corresponding altitude) measured by the navigator reIers to the local direction of gravity
(plumb line) which points to the astronomical nadir and thus deIines the astronomical zenith which is exactly
opposite to the nadir. Even the visible sea horizon is deIined by the astronomical zenith since the plane tangent to the
water surIace at the observer's position is perpendicular to the local direction oI gravity.
Under the assumption that the mass distribution inside the ellipsoid is in a hydrostatic equilibrium, the plumb line
coincides with the local normal to the ellipsoid which passes through the geodetic zenith. Thus, astronomical and
geodetic zenith are identical. Accordingly the astronomical coordinates oI a terrestrial position (obtained by
astronomical observations) are equal to the geodetic (geographic) coordinates in this hypothetical case. As
demonstrated in Fig. 91 Ior example, the altitude oI the celestial north pole, P
N
, with respect to the geoidal horizon
equals the geodetic, not the geocentric latitude. A noon latitude, calculated Irom the (geocentric) declination and the
zenith distance with respect to the astronomical zenith would lead to the same result.
The geocentric zenith is deIined as the point where a straight line originating Irom the center oI the earth and passing
through the observer's position intersects the celestial sphere. The angle between this line and the local normal to the
reIerence ellipsoid is called angle of the vertical, v. The angle oI the vertical lies on the plane oI the local meridian and
is a Iunction oI the geodetic latitude. The Iollowing Iormula was proposed by Smart 9:
The coeIIicients oI the above Iormula reIer to the proportions oI the WGS 84 ellipsoid.
The angle oI the vertical at a given position equals the diIIerence between geodetic and geocentric latitude (Fig. 91):
The maximum value oI v, occurring at 45° geodetic latitude, is approx. 11.5'. Thus, the geocentric latitude oI an
observer being at 45°geodetic latitude is only 44° 48.5'.
The navigator, oI course, wants to know iI the oblateness oI the earth causes signiIicant errors due to the Iact that
calculations oI celestial navigation are based on the laws oI spherical trigonometry. According to the above values Ior
polar radius and equatorial radius oI the WGS 84 ellipsoid, the great circle distance oI one arcminute is 1849 m at the
poles and 1855 m at the equator. This small diIIerence does not produce a signiIicant error when plotting lines oI
position. It is thereIore suIIicient to use the adopted mean value (1 nautical mile ÷ 1.852 km). However, when
calculating the great circle distance (chapter 11) oI two locations thousands oI nautical miles apart, the error caused by
the oblateness oI the earth can increase to several nautical miles. II high precision is required, the Iormulas Ior geodetic
distance should be used 2. The shortest path between two points on the surIace oI an ellipsoid is called geodesic line.
It is the equivalent to the arc oI a great circle on the surIace oI a sphere.
The Geoid
In reality, the earth is not in a state oI hydrostatic equilibrium and is thereIore not exactly an oblate ellipsoid. The shape
oI the earth is more accurately described by the geoid, an equipotential surIace oI gravity.
The geoid has local anomalies in the Iorm oI elevations and depressions caused by geographic Ieatures and a non
uniIorm mass distribution. Elevations occur at local accumulations oI matter (mountains, ore deposits), depressions at
local deIiciencies oI matter (large water bodies, valleys, caverns). The elevation or depression oI each point oI the geoid
with respect to the reIerence ellipsoid is Iound by gravity measurement.
On the slope oI an elevation or depression oI the geoid, the direction oI gravity (the normal to the geoid) does not
coincide with the normal to the reIerence ellipsoid, i. e., the astronomical zenith diIIers Irom the geodetic zenith in such
places. The small angle between the local direction oI gravity and the local normal to the reIerence ellipsoid, i. e., the
angular distance between astronomical and geodetic zenith, is called deflection of the vertical. The latter is composed
oI a meridional (northsouth) component and a zonal (westeast) component.
The deIlection oI the vertical is usually negligible at sea. In the vicinity oI mountain ranges, however, signiIicant
deIlections oI the vertical (up to approx. 1 arcminute) have been reported (chapter 2). Thus, an astronomical position
may be measurably diIIerent Irom the geodetic (geographic) position. This is important to surveying and mapmaking.
ThereIore, local corrections Ior the meridional and zonal component may have to be applied to an astronomical
position, depending on the required precision.
92
v ' '  692.666sin(2Lat ) 1.163sin( 4Lat ) + 0.026sin(6Lat )
v Lat  Lat '
The Parallax of the Moon
Due to the oblateness oI the earth, the distance between geoidal and celestial horizon is not constant but can assume any
value between r
p
and r
e
, depending on the observer's latitude. This has a measurable eIIect on the parallax in altitude oI
the moon since tabulated values Ior HP reIer to the equatorial radius, r
e
. The apparent position oI the moon is Iurther
aIIected by the Iact that usually the local direction oI gravity does not pass through the center oI the ellipsoid. This
displacement oI the plumb line Irom the earth's center causes a small (usually negligible) parallax in azimuth unless the
moon is on the local meridian. In the Iollowing, we will calculate the eIIects oI the oblateness oI the earth on the
parallax oI the moon with the exact Iormulas oI spherical astronomy 9. The eIIect oI the oblateness oI the earth on the
apparent position oI other bodies is negligible.
Fig. 92 shows a projection oI the astronomical zenith, Z
a
, the geocentric zenith, Z
c
, and the geographic position oI the
moon, M, on the celestial sphere, an imaginary hollow sphere oI inIinite diameter with the earth at its center.
The geocentric zenith, Z
c
, is the point where a straight line Irom the earth's center through the observer's position
intersects the celestial sphere. The astronomical zenith, Z
a
, is the point at which the plumb line going through the
observer's position intersects the celestial sphere. Z
a
and Z
c
are on the same celestial meridian. M is the projected
geocentric position oI the moon deIined by Greenwich hour angle and declination.
M' is the point where a straight line Irom the observer through the moon's center intersects the celestial sphere. Z
c
, M,
and M' are on a great circle. The zenith distance measured by the observer is z
a
' because the astronomical zenith is the
only reIerence available. The quantity we want to know is z
a
, the astronomical zenith distance corrected Ior parallax in
altitude. This is the angular distance oI the moon Irom the astronomical zenith, measured by a Iictitious observer at the
earth's center.
The known quantities are v, A
a
', and z
a
'. In contrast to the astronomer, the navigator is usually not able to measure A
a
'
precisely. For navigational purposes, the calculated azimuth (see chapter 4) may be substituted Ior A
a
'.
We have three spherical triangles, Z
a
Z
c
M', Z
a
Z
c
M, and Z
a
MM'. First, we calculate z
c
' Irom z
a
', v, and A
a
' using the law
of cosines for sides (see chapter 10):
93
cos z
c
l
cos z
a
l
cos v +sin z
a
l
sinvcos (180° A
a
l
)
z
c
l
arccos
(
cos z
a
l
cos v sin z
a
l
sinvcos A
a
l
)
To obtain z
c
, we Iirst have to calculate the relative length (r
e
÷ 1) oI the radius vector, r, and the geocentric parallax, p
c
:
HP is the equatorial horizontal parallax. The geocentric zenith distance corrected Ior parallax is:
Using the cosine Iormula again, we calculate A
c
, the azimuth angle oI the moon with respect to the geocentric zenith:
The astronomical zenith distance corrected Ior parallax is:
Thus, the parallax in altitude (astronomical) is:
The small angle between M and M', measured at Z
a
, is the parallax in azimuth, p
az
:
The parallax in azimuth does not exist when the moon is on the local meridian. It is Iurther nonexistant when the
observer is at one oI the poles or on the equator (v ÷ 0) but greatest when the observer is at medium latitudes. As a
consequence oI the parallax in azimuth, the horizontal direction oI the moon observed Irom the surIace oI the ellipsoid
is always a little closer to the elevated pole (the celestial pole above the horizon) than the horizontal direction observed
Irom the center oI the ellipsoid. The parallax in azimuth does not exceed +I HP when the moon is on the horizon but
increases with increasing altitude. In most cases, particularly at sea, the navigator will not notice the inIluence oI the
Ilattening oI the earth. Traditionally, the apparent altitude oI a body is reduced to the geocentric altitude through the
established altitude correction procedure (including the correction Ior parallax in altitude). The intercept method
(chapter 4) compares the observed altitude thus obtained with the geocentric altitude calculated Irom the assumed
geodetic (geographic) coordinates oI the observer and the geocentric equatorial coordinates (chapter 3) oI the observed
body. The diIIerence between observed and calculated altitude is the intercept. The calculated azimuth is geocentric. A
correction Ior the parallax in azimuth (see above) is usually omitted since such a degree oI precision can not be
reproduced when plotting position lines on a nautical chart. On land, however, more accurate altitude measurement is
possible, and the navigator or surveyor may wish to use reIined methods Ior the calculation oI his position when
observing the moon.
Mediumprecision method
During the course oI altitude corrections, we calculate the parallax in altitude, P, with the Iormulas Ior spherical bodies
(chapter 2). AIter doing this, we calculate the approximate correction Ior the Ilattening oI the earth, P:
Adding P to P, we get the improved parallax in altitude which we use Ior our Iurther calculations instead oI P:
94
p
c
arcsin
(
jsin HPsinz
c
l
)
j
r
r
e
.
1 ( 2e
2
e
4
)sin
2
Lat
1  e
2
sin
2
Lat
e
2
1 
r
p
2
r
e
2
z
c
z
c
l
 p
c
A
c
arccos
cos z
a
l
 cosz
c
l
cosv
sinz
c
l
sin v
z
a
arccos
(
cos z
c
cosv + sin z
c
sinvcos A
c
)
PA z
a
l
 z
a
p
az
arccos
cos p
c
 cos z
a
cos z
a
l
sinz
a
sin z
a
l
AP  f HP sin( 2Lat )cos A:
N
sin H sin
2
Lat cos H
P
improved
P +A P
As a result, we obtain a more accurate intercept (chapter 4). The above correction Iormula is accurate to a Iraction oI a
second oI arc.
The approximate parallax in azimuth is obtained through a simple Iormula:
The topocentric true azimuth is
The Iormula Ior the parallax in azimuth is also accurate to a Iraction oI an arcsecond. It becomes less accurate as the
altitude approaches 90°. Observing bodies with such altitudes, however, is diIIicult and usually avoided.
Rigorous method`
For even more accurate results, we use the topocentric equatorial coordinates oI the observed body Ior sight
reduction. Instead oI the center oI the earth, the observer`s position is the origin oI this coordinate system. The plane oI
the topocentric equator is parallel to the geocentric equator. The plane oI the local meridian remains the same. The
values Ior altitude and true azimuth calculated Irom the topocentric coordinates oI the observed body are topocentric as
well. There is neither a parallax in altitude nor a parallax in azimuth, so we have to skip the parallax correction and
have to correct Ior the topocentric (augmented) semidiameter oI the body when perIorming the altitude corrections.
The topocentric equatorial coordinates oI a celestial body are obtained Irom the geocentric ones through coordinate
transIormation. The given quantities are:
Geographic latitude oI the observer Lat
Geocentric meridian angle t
Geocentric declination Dec
Equatorial horizontal parallax HP
Polar radius oI the earth r
p
Equatorial radius oI the earth r
e
To be calculated:
Topocentric meridian angle t'
Topocentric declination Dec'
First, we calculate a number oI auxiliary quantities:
Eccentricity oI the ellipsoid, distance between center and Iocal point oI a meridional section (r
e
÷ 1):
Local radius ( r
e
÷ 1):
Geocentric latitude oI the observer:
*The Iormulas are rigorous Ior an observer on the surIace oI a reIerence ellipsoid the center oI which coincides with the mass center oI the earth.
95
e
.
1
(
r
p
r
e
)
2
j
r
p
r
e
1
.1e
2
cos
2
Lat ´
Lat ' arctan
(
r
p
r
e
)
2
tan Lat
arctan (1f )
2
tan Lat
AA:
N
 f HP
sin( 2Lat )sin A:
N , geocentric
cos H
geocentric
A:
N , topocentric
 A:
N , geocentric
A A:
N
The topocentric coordinates oI the body, t' and Dec', are calculated as Iollows. t is the parallax in hour angle:
96
At arctan
jcos Lat ´sin HPsin t
cos Dec  jcos Lat ´sin HPcost
t ´ t +At
Dec ´ arctan
(sin Dec  jsin Lat ´sin HP)cosAt
cos Dec jcos Lat ´sin HPcos t
Chapter 10
Spherical Trigonometry
The earth is usually regarded as a sphere in celestial navigation although an oblate spheroid would be a better
approximation. Otherwise, navigational calculations would become too diIIicult Ior practical use. The position error
introduced by the spherical earth model is usually very small and stays within the "statistical noise" caused by other
omnipresent errors like, e.g., abnormal reIraction, rounding errors, etc. Although it is possible to perIorm navigational
calculations solely with the aid oI tables (H.O. 229, H.O. 211, etc.) and with little mathematics, the principles oI
celestial navigation can not be comprehended without knowing the elements oI spherical trigonometry.
The Oblique Spherical Triangle
Like any triangle, a spherical triangle is characterized by three sides and three angles. However, a spherical triangle is
part oI the surIace oI a sphere, and the sides are not straight lines but arcs oI great circles (Fig. 101).
A great circle is a circle on the surIace oI a sphere whose plane passes through the center oI the sphere (see chapter 3).
Any side oI a spherical triangle can be regarded as an angle  the angular distance between the adjacent vertices,
measured at the center oI the sphere. The interrelations between angles and sides oI a spherical triangle are described by
the law of sines, the law of cosines for sides, the law of cosines for angles, the law of sines and cosines, the law of
cotangents, Napier's analogies, and Gauss' formulas (apart Irom other Iormulas).
Law of sines:
Law of cosines for sides:
Law of cosines for angles:
101
sin A
1
sin s
1
sin A
2
sin s
2
sin A
3
sin s
3
cos s
1
cos s
2
coss
3
sins
2
sins
3
cos A
1
cos s
2
coss
1
coss
3
sins
1
sin s
3
cos A
2
cos s
3
cos s
1
cos s
2
sins
1
sin s
2
cos A
3
cos A
1
cos A
2
cos A
3
sin A
2
sin A
3
cos s
1
cos A
2
cos A
1
cos A
3
sin A
1
sin A
3
coss
2
cos A
3
cos A
1
cos A
2
sin A
1
sin A
2
cos s
3
Law of sines and cosines:
Law of cotangents:
Napier's analogies:
Gauss' formulas:
These Iormulas and others derived thereoI enable any quantity (angle or side) oI a spherical triangle to be calculated iI
three other quantities are known.
Particularly the law of cosines for sides is of interest to the navigator.
102
tan
A
1
A
2
2
tan
A
3
2
cos
s
1
s
2
2
cos
s
1
s
2
2
tan
A
1
A
2
2
tan
A
3
2
sin
s
1
s
2
2
sin
s
1
s
2
2
tan
s
1
s
2
2
tan
s
3
2
cos
A
1
A
2
2
cos
A
1
A
2
2
tan
s
1
s
2
2
tan
s
3
2
sin
A
1
A
2
2
sin
A
1
A
2
2
sin
A
1
A
2
2
cos
A
3
2
cos
s
1
s
2
2
cos
s
3
2
cos
A
1
A
2
2
sin
A
3
2
cos
s
1
s
2
2
cos
s
3
2
sin
A
1
A
2
2
cos
A
3
2
sin
s
1
s
2
2
sin
s
3
2
cos
A
1
A
2
2
sin
A
3
2
sin
s
1
s
2
2
sin
s
3
2
sin s
1
cos A
2
cos s
2
sin s
3
sin s
2
cos s
3
cos A
1
sin A
1
cot A
2
cot s
2
sin s
3
cos s
3
cos A
1
sin s
2
cos A
3
cos s
3
sin s
1
sin s
3
cos s
1
cos A
2
sin s
3
cos A
1
cos s
1
sin s
2
sin s
1
cos s
2
cos A
3
sin s
1
cos A
3
cos s
3
sin s
2
sin s
3
cos s
2
cos A
1
sin s
2
cos A
1
cos s
1
sin s
3
sin s
1
cos s
3
cos A
2
sin A
1
cot A
3
cot s
3
sin s
2
cos s
2
cos A
1
sin A
2
cot A
3
cot s
3
sin s
1
cos s
1
cos A
2
sin A
2
cot A
1
cot s
1
sin s
3
cos s
3
cos A
2
sin A
3
cot A
1
cot s
1
sin s
2
cos s
2
cos A
3
sin s
3
cos A
2
cos s
2
sin s
1
sin s
2
cos s
1
cos A
3
sin A
3
cot A
2
cot s
2
sin s
1
cos s
1
cos A
3
The Right Spherical Triangle
Solving a spherical triangle is less complicated when it contains a right angle (Fig. 102). Using rules of
circular parts, any quantity can be calculated iI only two other quantities (apart Irom the right angle) are known.
We arrange the sides Iorming the right angle (s
1
, s
2
) and the complements oI the remaining angles (A
1
, A
2
) and
opposite side (s
3
) in the Iorm oI a circular diagram consisting oI Iive sectors, called "parts" (in the same order as they
appear in the triangle). The right angle itselI is omitted (Fig. 103):
According to Napier´s rules, the sine oI any part oI the diagram equals the product oI the tangents oI the adjacent parts
and the product oI the cosines oI the opposite parts:
In a simpler Iorm, these equations are stated as:
There are several applications Ior the right spherical triangle in navigation, Ior example Ageton's sight reduction
tables (chapter 11) and great circle navigation (chapter 13).
103
sins
1
tans
2
tan 90° A
2
cos90° A
1
cos90° s
3
sins
2
tan90° A
1
tan s
1
cos 90° s
3
cos 90° A
2
sin90° A
1
tan 90° s
3
tan s
2
cos90° A
2
cos s
1
sin 90° s
3
tan90° A
2
tan 90° A
1
cos s
1
cos s
2
sin90° A
2
tans
1
tan90°s
3
cos s
2
cos 90° A
1
sin s
1
tan s
2
cot A
2
sin A
1
sin s
3
sin s
2
cot A
1
tan s
1
sin s
3
sin A
2
cos A
1
cot s
3
tan s
2
sin A
2
cos s
1
cos s
3
cot A
2
cot A
1
cos s
1
cos s
2
cos A
2
tan s
1
cot s
3
cos s
2
sin A
1
Chapter 11
The Navigational Triangle
The navigational (nautical) triangle is the (usually) oblique spherical triangle Iormed by the north pole, P
N
, the
observer's assumed position, AP, and the geographic position oI the celestial object, GP (Fig. 111). All common sight
reduction procedures are based upon the navigational triangle.
When using the intercept method, the latitude oI the assumed position, Lat
AP
, the declination oI the observed celestial
body, Dec, and the meridian angle, t, or the local hour angle, LHA, (calculated Irom the longitude oI AP and the GHA
oI the object), are the known quantities.
The Iirst step is calculating the side z oI the navigational triangle by using the law oI cosines Ior sides:
Since cos (90°x) equals sin x and vice versa, the equation can be written in a simpler Iorm:
The side z is not only the great circle distance between AP and GP but also the zenith distance oI the celestial object
and the radius oI the circle of equal altitude (see chapter 1).
Substituting the altitude H Ior z, we get:
Solving the equation Ior H leads to the altitude Iormula known Irom chapter 4:
The altitude thus obtained Ior a given position is called computed altitude, Hc.
111
cos z cos 90° Lat
AP
cos90° Dec sin 90° Lat
AP
sin90° Deccos t
cos z sin Lat
AP
sin Dec cos Lat
AP
cos Dec cos t
sin H sin Lat
AP
sinDec cos Lat
AP
cos Deccos t
H arcsin
sin Lat
AP
sinDec cos Lat
AP
cos Deccos t
The azimuth angle oI the observed body is also calculated by means oI the law oI cosines Ior sides:
Using the computed altitude instead oI the zenith distance results in the Iollowing equation:
Solving the equation Ior Az Iinally yields the azimuth Iormula Irom chapter 4:
The arccos Iunction returns angles between 0° and 180°. ThereIore, the resulting azimuth angle is not necessarily
identical with the true azimuth, Az
N
(0°... 360°, measured clockwise Irom true north) commonly used in navigation. In
all cases where t is negative (GP east oI AP) , Az
N
equals Az. Otherwise (t positive, GP westward Irom AP as shown in
Fig. 111), Az
N
is obtained by subtracting Az Irom 360°.
When the meridian angle, t, (or the local hour angle, LHA) is the quantity to be calculated (time sight, Sumner´s
method), Dec, Lat
AP
(the assumed latitude), and z (or H) are the known quantities.
Again, the law oI cosines Ior sides is applied:
The obtained meridian angle, t (or LHA), is then used as described in chapter 4 and chapter 5.
When observing a celestial body at the time oI meridian passage (e. g., Ior determining one's latitude), the local hour
angle is zero, and the navigational triangle becomes inIinitesimally narrow. In this case, the Iormulas oI spherical
trigonometry are not needed, and the sides oI the spherical triangle can be calculated by simple addition or subtraction.
The Divided Navigational Triangle
An alternative method Ior solving the navigational triangle is based upon two right spherical triangles obtained by
constructing a great circle passing through GP and intersecting the local meridian perpendicularly at X (Fig. 112).
112
cos90° Dec cos90° Lat
AP
cos z sin 90° Lat
AP
sinzcos Az
sin Dec sin Lat
AP
cosz cos Lat
AP
sinzcos Az
sin Dec sin Lat
AP
sin Hc cos Lat
AP
cos Hccos Az
Az arccos
sin Dec sin Lat
AP
sinHc
cos Lat
AP
cos Hc
cos z cos90° Lat
AP
cos90° Dec sin90° Lat
AP
sin90° Deccos t
sin H sin Lat
AP
sin Dec cos Lat
AP
cos Dec cos t
cos t
sin H sin Lat
AP
sinDec
cos Lat
AP
cos Dec
t arccos
sin H sinLat
AP
sin Dec
cos Lat
AP
cos Dec
The Iirst right triangle is Iormed by P
N
, X, and GP, the second one by GP, X, and AP. The auxiliary parts R and K are
intermediate quantities used to calculate z (or Hc) and Az. K is the geographic latitude oI X. Both triangles are solved
using Napier's rules of circular parts (see chapter 9). Fig. 113 illustrates the corresponding circular diagrams:
According to Napier's rules, Hc and Az are calculated by means oI the Iollowing Iormulas:
Substitute 180°−K Ior K in the Iollowing equation iI t ~ 90° (or 90° · LHA · 270°).
For Iurther calculations, substitute 180°−Az Ior Az iI K and Lat have opposite signs or iI ,K, < ,Lat,.
113
sin R sint cos Dec R arcsinsint cos Dec
sin Dec cos Rsin K sin K
sin Dec
cos R
K arcsin
sin Dec
cos R
sin Hc cos Rcos K Lat
AP
Hc arcsin
cos Rcos K Lat
AP
sin R cos Hc sin Az sin Az
sin R
cos Hc
Az arcsin
sin R
cos Hc
To obtain the true azimuth, Az
N
(0°... 360°), the Iollowing rules have to be applied:
The divided navigational triangle is oI considerable importance since it Iorms the theoretical background Ior a number
oI sight reduction tables, e.g., the Ageton Tables (see below). It is also used Ior great circle navigation (chapter 12).
Using the secant and cosecant Iunction (sec x ÷ 1/cos x, csc x ÷ 1/sin x), we can write the equations Ior the divided
navigational triangle in the Iollowing Iorm:
Substitute 180°−K Ior K in the Iollowing equation iI t ~ 90°:
Substitute 180°−Az Ior Az iI K and Lat have opposite signs or iI K<Lat.
In logarithmic Iorm, these equations are stated as:
With the logarithms oI the secants and cosecants oI angles arranged in the Iorm oI a suitable table, we can solve a sight
by a sequence oI simple additions and subtractions. Apart Irom the table itselI, the only tools required are a sheet oI
paper and a pencil.
The Ageton Tables (H.O. 211), Iirst published in 1931, are based upon the above Iormulas and provide a very eIIicient
arrangement oI angles and their log secants and log cosecants on 36 pages. Since all calculations are based on absolute
values, certain rules included in the instructions have to be observed.
Sight reduction tables were developed many years beIore electronic calculators became available in order to simpliIy
calculations necessary to reduce a sight. Still today, sight reduction tables are preIerred by people who do not want to
deal with the Iormulas oI spherical trigonometry. Moreover, they provide a valuable backup method iI electronic
devices Iail.
Two modiIied versions oI the Ageton Tables are available at: http://www.celnav.de/page3.htm
114
Az
N
Az if Lat
AP
0 N AND t 0 180° LHA 360°
360° Az if Lat
AP
0 N AND t 0 0° LHA 180°
180° Az if Lat
AP
0 S
csc R csc t sec Dec
csc K
csc Dec
sec R
csc Hc sec Rsec K Lat
csc Az
csc R
sec Hc
log csc R log csct log sec Dec
log csc K log csc Dec log sec R
log csc Hc log sec R log sec K Lat
log csc Az log csc R log sec Hc
Chapter 12
General Formulas for Navigation
Although the Iollowing Iormulas are not part oI celestial navigation, they are indispensible because they are necessary
to calculate distance and direction (course) Irom the point of departure, A, to the point of arrival, B, as well as to
calculate the position oI B Irom the position oI A iI course and distance are known. The true course, C, is the angle
made by the vector oI motion and the local meridian. It is measured Irom true north (clockwise through 360°).
Knowing the coordinates oI A, Lat
A
and Lon
A
, and the coordinates oI B, Lat
B
and Lon
B
, the navigator has the principal
choice between rhumb line navigation (simple procedure but longer distance) and great circle navigation (shortest
possible distance on a sphere). Combinations oI both methods are possible.
Rhumb Line Navigation
A rhumb line, also called loxodrome, is a line on the surIace oI the earth intersecting all meridians at a constant angle,
C. Thus, a rhumb line is represented by a straight line on a Mercartor chart (see chapter 13) which makes voyage
planning quite simple. On a globe, a rhumb line Iorms a spherical spiral extending Irom pole to pole unless it is
identical with a meridian (C ÷ 0° or 180°) or a parallel oI latitude (C ÷ 90° or 270°). A vessel steering a constant course
travels along a rhumb line, provided there is no driIt. Rhumb line course, C, and distance, d, are calculated as shown
below. First, we imagine traveling the inIinitesimal distance dx Irom the point oI departure, A, to the point oI arrival, B.
Our course is C ():
The distance, dx, is the vector sum oI a northsouth component, dLat, and a westeast component, dLon
.
cos Lat. The
Iactor cos Lat is the relative circumIerence oI the respective parallel oI latitude (equator ÷ 1):
II the distance between A (deIined by Lat
A
and Lon
A
) and B (deIined by Lat
B
and Lon
B
) is a measurable quantity, we
have to integrate:
121
tan C
d Loncos Lat
d Lat
d Lat
cos Lat
1
tan C
d Lon

Lat
A
Lat
B
d Lat
cos Lat
1
tan C

Lon
A
Lon
B
d Lon
ln
tan
(
Lat
B
2
+
n
4
)
ln
tan
(
Lat
A
2
+
n
4
)
Lon
B
 Lon
A
tanC
Measuring angles in degrees and solving Ior C, we get:
The term Lon
B
Lon
A
has to be in the range between 180° tand +180°. If it is outside this range, we have to add
or subtract 360° before entering the rhumb line course formula.
The arctan function returns values between 90° and +90°. To obtain the true course (0°...360°), we apply the
following rules:
To Iind the total length oI the rhumb line track, we calculate the inIinitesimal distance dx:
The total length d is Iound through integration:
Finally, we get:
II both positions have the same latitude, the distance can not be calculated using the above Iormulas. In this case, the
Iollowing Iormulas apply (C is either 90° or 270°):
122
tanC
Lon
B
 Lon
A
ln
tan
(
Lat
B
2
+
n
4
)
tan
(
Lat
A
2
+
n
4
)
C arctan
Lon
B
 Lon
A
ln
tan
(
Lat
B
2
+ 45°
)
tan
(
Lat
A
2
+ 45°
)
C 
C if Lat
B
> Lat
A
AND Lon
B
> Lon
A
360°+C if Lat
B
> Lat
A
AND Lon
B
Lon
A
180°+C if Lat
B
Lat
A
dx
d Lat
cosC
d
1
cosC

Lat
A
Lat
B
d Lat
Lat
B
 Lat
A
cosC
d km
40031.6
360
Lat
B
 Lat
A
cos C
d nm 60
Lat
B
 Lat
A
cosC
d km
40031.6
360
( Lon
B
 Lon
A
)cos Lat d nm 60( Lon
B
 Lon
A
)cos Lat
Great Circle Navigation
Great circle distance, d
AB
, and course, C
A
, are calculated on the analogy oI zenith distance and azimuth. For this
pupose, we consider the navigational triangle (see chapter 11) and substitute A Ior GP, B Ior AP, d
AB
Ior z, and
Lon
AB
(diIIerence oI longitude) Ior LHA ():
Northern latitude and eastern longitude are positive, southern latitude and western longitude negative. A great circle
distance has the dimension oI an angle. To measure it in distance units, we multiply it by 40031.6/360 (distance in km)
or by 60 (distance in nm).
II the term sin (Lon
B
Lon
A
) is negative, we replace C
A
with 360°C
A
in order to obtain the true course (0°... 360°
clockwise Irom true north).
In , C
A
is the initial great circle course, C
B
the final great circle course. Since the angle between the great
circle and the respective local meridian varies as we progress along the great circle (unless the great circle coincides
with the equator or a meridian), we can not steer a constant course as we would when Iollowing a rhumb line.
Theoretically, we have to adjust the course continually. This is possible with the aid oI navigation computers and
autopilots. II such means are not available, we have to calculate an updated course at certain intervals (see below).
Great circle navigation requires more careIul voyage planning than rhumb line navigation. On a Mercator chart (see
chapter 13), a great circle track appears as a line bent towards the equator. As a result, the navigator may need more
inIormation about the intended great circle track in order to veriIy iI it leads through navigable areas.
With the exception oI the equator, every great circle has two vertices, the points Iarthest Irom the equator. The vertices
have the same absolute value oI latitude (with opposite sign) but are 180° apart in longitude. At each vertex (also called
apex), the great circle is tangent to a parallel oI latitude, and C is either 90° or 270° (cos C ÷ 0). Thus, we have a right
spherical triangle Iormed by the north pole, P
N
, the vertex,V, and the point oI departure, A (Fig. 123):
123
d
AB
arccos
(
sin Lat
A
sinLat
B
+ cos Lat
A
cos Lat
B
cos ALon
AB
)
ALon
AB
Lon
B
 Lon
A
C
A
arccos
sin Lat
B
 sin Lat
A
cosd
AB
cos Lat
A
sin d
AB
To derive the Iormulas needed Ior the Iollowing calculations, we use Napier's rules oI circular parts (Fig. 124). The
right angle is at the bottom oI the circular diagram. The Iive parts are arranged clockwise.
First, we need the latitude oI the vertex, Lat
V
:
Solving Ior Lat
V
, we get:
The absolute value oI sin C
A
is used to make sure that Lat
V
does not exceed +90° (the arccos Iunction returns values
between 90° and 180° Ior negative arguments). The equation has two solutions, according to the number oI vertices.
Only the vertex lying ahead oI us is relevant to voyage planning. It is Iound using the Iollowing modiIied Iormula:
sng(x) is the signum Iunction:
II V is located between A and B (like shown in Fig. 123), our latitude passes through an extremum at the instant we
reach V. This does not happen iI B is between A and V.
124
cos Lat
V
sinC
A
cos Lat
A
Lat
V
arccos
(
¸sin C
A
¸cos Lat
A
)
Lat
V
sgn
(
cosC
A
)
arccos
(
¸ sinC
A
¸cos Lat
A
)
sgn( x)
1 if x0
0 if x0
+1 if x>0
Knowing Lat
V
, we are able to calculate the longitude oI V. Again, we apply Napier's rules:
Solving Ior Lon
AV
, we get:
The longitude oI V is
(Add or subtract 360° iI necessary.)
The term sng (sin C
A
) in the above Iormula provides an automatic correction Ior the sign oI Lon
AV
.
Knowing the position oI V (deIined by Lat
V
and Lon
V
), we are now able to calculate the position oI any chosen point,
X, on the intended great circle track (substituting X Ior A in the right spherical triangle). Using Napier's rules once
more, we get:
Further, we can calculate the course at the point X:
Alternatively, C
X
can be calculated Irom the oblique spherical triangle Iormed by X, P
N
, and B.
The above Iormulas enable us to establish suitably spaced waypoints on the great circle and connect them by straight
lines on the Mercator chart. The series oI legs thus obtained, each one being a rhumb line track, is a practical
approximation oI the intended great circle track. Further, we are now able to see beIorehand iI there are obstacles in our
way.
Mean latitude
Because oI their simplicity, the mean latitude Iormulas are oIten used in everyday navigation. Mean latitude is a good
approximation Ior rhumb line navigation Ior short and medium distances between A and B. The method is less suitable
Ior polar regions (convergence oI meridians).
Course:
The true course is obtained by applying the same rules to C
as to the rhumb line course (see above).
Distance:
125
cosALon
AV
tan Lat
A
tan Lat
V
ALon
AV
ALon
V
 ALon
A
ALon
AV
arccos
tan Lat
A
tan Lat
V
Lon
V
Lon
A
+sgn
(
sinC
A)
arccos
tan Lat
A
tan Lat
V
tan Lat
X
cosALon
XV
tan Lat
V
ALon
XV
Lon
V
 Lon
X
Lat
X
arctan
(
cos ALon
XV
tan Lat
V
)
cosC
X
sinALon
XV
sinLat
V
C
X
arccos
(
sinALon
XV
sin Lat
V
)
if sinC
A
> 0
arccos
(
sinALon
XV
sin Lat
V
)
+ 180° if sinC
A
0
C arctan
(
cos Lat
M
Lon
B
 Lon
A
Lat
B
 Lat
A
)
Lat
M
Lat
A
+ Lat
B
2
d km
40031.6
360
Lat
B
 Lat
A
cos C
d nm 60
Lat
B
 Lat
A
cosC
II C ÷ 90° or C ÷ 270°, we have use the Iollowing Iormulas:
Dead Reckoning
Dead reckoning is the navigational term Ior extrapolating one's new position, B, Irom the previous position, A, the
course C, and the distance, d (calculated Irom the vessel's average speed and time elapsed). The position thus obtained
is called a dead reckoning position, DRP.
Since a DRP is only an approximate position (due to the inIluence oI driIt, etc.), the mean latitude method (see above)
provides suIIicient accuracy. On land, dead reckoning is oI limited use since it is usually not possible to steer a constant
course (apart Irom driving in large, entirely Ilat areas like, e.g., salt Ilats).
At sea, the DRP is needed to choose a suitable AP Ior the intercept method. II celestial observations are not possible
and electronic navigation aids not available, dead reckoning may be the only way oI keeping track oI one's position.
Apart Irom the very simple graphic solution, there are two Iormulas Ior the calculation oI the DRP.
Calculation oI new latitude:
Calculation oI new longitude:
II the resulting longitude is greater than ¹180°, we subtract 360°. II it is smaller than 180°, we add 360°.
II our movement is composed oI several components (including driIt, etc.), we have to replace the terms dcos C and
dsin C with
_
d
i
cosC
i
and
_
d
i
sinC
i
, respectively.
126
Lat
B
° Lat
A
° +
360
40031.6
d kmcosC Lat
B
° Lat
A
° +
d nmcosC
60
Lon
B
° Lon
A
° +
360
40031.6
d kmsinC
cosLat
M
Lon
B
° Lon
A
° +
d nmsinC
60cos Lat
M
d km
40031.6
360
(
Lon
B
 Lon
A
)
cos Lat d nm 60
(
Lon
B
 Lon
A
)
cos Lat
Chapter 13
Charts and Plotting Sheets
Mercator Charts
Sophisticated navigation is not possible without the use oI a map (chart), a projection oI a certain area oI the earth's
surIace with its geographic Ieatures on a plane. Among the numerous types oI map projection, the Mercator
projection, named aIter the FlemishGerman cartographer Gerhard Kramer (Latin: Gerardus Mercator), is mostly used
in navigation because it produces charts with an orthogonal grid which is most convenient Ior measuring directions and
plotting lines oI position. Further, rhumb lines appear as straight lines on a Mercator chart. Great circles do not, apart
Irom meridians and the equator which are also rhumb lines.
In order to construct a Mercator chart, we have to remember how the grid printed on a globe looks. At the equator, an
area oI, e. g., 2 by 2 degrees looks almost like a square, but it appears as a narrow trapezoid when we place it near one
oI the poles. While the distance between two adjacent parallels oI latitude is constant, the distance between two
meridians becomes progressively smaller as the latitude increases because the meridians converge to the poles. An area
with the inIinitesimal dimensions dLat and dLon would appear as an oblong with the dimensions dx and dy on our
globe (Fig. 131):
dx contains the Iactor cos Lat since the circumIerence oI a parallel oI latitude is proportional to cos Lat. The constant c'
is the scale Iactor oI the globe (measured in, e. g., mm/°).
Since we require any rhumb line to appear as a straight line intersecting all meridians at a constant angle, meridians
have to be equally spaced vertical lines on our chart, and any inIinitesimal oblong deIined by dLat and dLon must have
the same aspect ratio as on the globe (dy/dx ÷ const.) at a given latitude (conformality). ThereIore, iI we transIer the
oblong deIined by dLat and dLon Irom the globe to our chart, we get the dimensions:
The new constant c is the scale Iactor oI the chart. Now, dx remains constant (parallel meridians), but dy is a Iunction
oI the latitude at which our small oblong is located. To obtain the smallest distance Irom any point with the latitude
Lat
P
to the equator, we integrate:
Y is the distance oI the respective parallel oI latitude Irom the equator. In the above equation, angles are given in
circular measure (radians). II we measure angles in degrees, the equation is stated as:
131
dx c' d Loncos Lat
dy c' d Lat
dx cd Lon
dy c
d Lat
cos Lat
Y
0
Y
dy c
0
Lat
P
d Lat
cos Lat
cln tan
Lat
P
2
4
Y cln tan
Lat
P
°
2
45°
The distance oI any point Irom the Greenwich meridian (Lon ÷ 0°) varies proportionally with the longitude oI the
point, Lon
P
. X is the distance oI the respective meridian Irom the Greenwich meridian:
Fig. 132 shows an example oI the resulting graticule (10° spacing). While meridians oI longitude appear as equally
spaced vertical lines, parallels oI latitude are horizontal lines drawn Iarther apart as the latitude increases. Y would be
inIinite at 90° latitude.
Mercator charts have the disadvantage that geometric distortions increase as the distance Irom the equator increases.
The Mercator projection is thereIore not suitable Ior polar regions. A circle oI equal altitude, Ior example, would
appear as a distorted ellipse at higher latitudes. Areas near the poles, e. g., Greenland, appear much greater on a
Mercator map than on a globe.
It is oIten said that a Mercator chart is obtained by projecting each point oI the surIace oI a globe Irom the center oI the
globe to the inner surIace oI a hollow cylinder tangent to the globe at the equator. This is only a rough approximation.
As a result oI such a (purely geometrical) projection, Y would be proportional to tan Lat, and conIormality would not
be achieved.
Plotting Sheets
II we magniIy a small part oI a Mercator chart, e. g., an area oI 30' latitude by 40' longitude, we will notice that the
spacing between the parallels oI latitude now seems to be almost constant. An approximated Mercator grid oI such a
small area can be constructed by drawing equally spaced horizontal lines, representing the parallels oI latitude, and
equally spaced vertical lines, representing the meridians.
The spacing oI the parallels oI latitude, y, deIines the scale oI our chart, e. g., 5mm/nm. The spacing oI the meridians,
x, is a Iunction oI the mean latitude, Lat
M
:
A sheet oI otherwise blank paper with such a simpliIied Mercator grid is called a small area plotting sheet and is a
very useIul tool Ior plotting lines oI position.
132
X
0
Lon
P
dx cLon
P
x ycos Lat
M
Lat
M
Lat
min
Lat
max
2
II a calculator or trigonometric table is not available, the meridian lines can be constructed with the graphic method
shown in Fig. 133:
We take a sheet oI blank paper and draw the required number oI equally spaced horizontal lines (parallels oI latitude).
A spacing oI 3  10 mm per nautical mile is recommended Ior most applications.
We draw an auxiliary line intersecting the parallels oI latitude at an angle numerically equal to the mean latitude. Then
we mark the map scale (deIined by the spacing oI the parallels) periodically on this line, and draw the meridian lines
through the points thus located. Compasses can be used to transIer the map scale Irom the chosen meridian to the
auxiliary line.
Small area plotting sheets are available at nautical book stores.
Gnomonic Charts
For great circle navigation, the gnomonic projection oIIers the advantage that any great circle appears as a straight
line. Rhumb lines, however, are curved. A gnomonic chart is obtained by projecting each point on the earth's surIace
Irom the earth's center to a plane tangent to the surIace. Since the distance oI a projected point Irom the point oI
tangency varies in proportion with the tangent oI the angular distance oI the original point Irom the point oI tangency, a
gnomonic chart covers less than a hemisphere, and distortions increase rapidly with increasing distance Irom the point
oI tangency. In contrast to the Mercator projection, the gnomonic projection is nonconIormal (not anglepreserving).
There are three types oI gnomonic projection:
II the plane oI projection is tangent to the earth at one oI the poles (polar gnomonic chart), the meridians appear as
straight lines radiating Irom the pole. The parallels oI latitude appear as concentric circles. The spacing oI the latter
increases rapidly as the polar distance increases.
II the point oI tangency is on the equator (equatorial gnomonic chart), the meridians appear as straight lines parallel
to each other. Their spacing increases rapidly as their distance Irom the point oI tangency increases. The equator
appears as a straight line perpendicular to the meridians. All other parallels oI latitude (small circles) are lines curved
toward the respective pole. Their curvature increases with increasing latitude.
In all other cases (oblique gnomonic chart), the meridians appear as straight lines converging at the elevated pole.
The equator appears as a straight line perpendicular to the central meridian (the meridian going through the point oI
tangency). Parallels oI latitude are lines curved toward the poles.
133
Fig. 134 shows an example oI an oblique gnomonic chart.
Fig. 134
A gnomonic chart is a useIul graphic tool Ior longdistance voyage planning. The intended great circle track is plotted
as a straight line Irom A to B. Obstacles, iI existing, become visible at once. The coordinates oI the chosen waypoints
(preIerably those lying on meridian lines) are then read Irom the graticule and transIerred to a Mercator chart, where the
waypoints are connected by rhumb line tracks.
134
Chapter 14
Magnetic Declination
Since the magnetic poles oI the earth do not coincide with the geographic poles and due to other irregularities oI the
earth's magnetic Iield, the horizontal component oI the magnetic Iield at a given position, called magnetic meridian,
usually Iorms an angle with the local geographic meridian. This angle is called magnetic declination or, in mariner's
language, magnetic variation. Accordingly, the needle oI a magnetic compass, aligning itselI with the local magnetic
meridian, does not exactly indicate the direction oI true north ().
Magnetic declination depends on the observer's geographic position and can exceed 30° or even more in some areas.
Knowledge oI the local magnetic declination is thereIore necessary to avoid dangerous navigation errors. Although
magnetic declination is oIten given in the legend oI topographic maps, the inIormation may be outdated because
magnetic declination varies with time (up to several degrees per decade). In some places, magnetic declination may
even diIIer Irom oIIicial statements due to local distortions oI the magnetic Iield caused by deposits oI lava,
Ierromagnetic ores, etc.
The azimuth Iormulas described in chapter 4 provide a very useIul tool to determine the magnetic declination at a given
position. II the observer does not know his exact position, an estimate will suIIice in most cases. A sextant is not
required Ior the simple procedure:
1. We choose a celestial body being low in the sky or on the visible horizon, preIerably sun or moon. We measure the
magnetic compass bearing, B, oI the center oI the body and note the time. The vicinity oI cars, steel objects,
magnets, DC power cables, etc. has to be avoided since they distort the magnetic Iield locally.
2. We extract GHA and Dec oI the body Irom the Nautical Almanac or calculate these quantities with a computer
almanac.
3. We calculate the meridian angle, t (or the local hour angle, LHA), Irom GHA and our longitude (see chapter 4).
4. We calculate the true azimuth, Az
N
, oI the body Irom Lat, Dec, and t. The time sight Iormula (chapter 4) with its
accompanying rules is particularly suitable Ior this purpose since an observed or computed altitude is not needed.
5. Magnetic declination, MD, is obtained by subtracting Az
N
Irom the compass bearing, B.
(Add 360° iI the angle thus obtained is smaller than 180°. Subtract 360° iI the angle is greater than ¹180°.)
Eastern declination (shown in ) is positive (0°...¹180°), western declination negative (0°...180°).
141
MD B Az
N
Chapter 15
Ephemerides of the Sun
The sun is probably the most Irequently observed body in celestial navigation. Greenwich hour angle and declination oI
the sun as well as GHA
Aries
and EoT can be calculated using the algorithms listed below 8,17. The Iormulas are
relatively simple and useIul Ior navigational calculations with programmable pocket calculators (10 digits
recommended).
First, the time variable, T, has to be calculated Irom year, month, and day. T is the time, measured in days and Iractions
oI a day, beIore or aIter Jan 1, 2000, 12:00:00 UT:
y is the number oI the year (4 digits), m is the number oI the month, and d the number oI the day in the respective
month. UT is Universal Time in decimal Iormat (e.g., 12h 30m 45s ÷ 12.5125). For May 17, 1999, 12:30:45 UT, Ior
example, T is 228.978646. The equation is valid Irom March 1, 1900 through February 28, 2100.
Iloor(x) is the greatest integer smaller than x. For example, Iloor(3.8) ÷ 3, Iloor(2.2) ÷ 3. The floor Iunction is part oI many
programming languages, e.g., JavaScript. In general, it is identical with the int Iunction used in other languages. However, there
seem to be diIIerent deIinitions Ior the int Iunction. This should be checked beIore programming the above Iormula.
Mean anomaly oI the sun
*
:
Mean longitude oI the sun
*
:
True longitude oI the sun
*
:
Obliquity oI the ecliptic:
Declination of the sun:
Right ascension oI the sun (in degrees)
*
:
151
T 367y floor
1.75
y floor
m 9
12
floor
275
m
9
d
UT h
24
730531.5
g ° 0.9856003T 2.472
L
T
° L
M
° 1.915sin g 0.02sin2g
L
M
° 0.9856474T 79.53938
° 23.439 4T 10
7
Dec ° arcsin
sin L
T
sin
RA° 2arctan
cossin L
T
cos Dec cos L
T
GHA
Aries
`
:
Greenwich hour angle of the sun
`
:
*
These quantities have to be within the range between 0° and 360°. II necessary, add or subtract 360° or multiples thereoI. This can
be achieved using the Iollowing algorithm which is particularly useIul Ior programmable calculators:
Equation of Time:
(II GAT ~ 24h, subtract 24h.)
EoT h GAT h UT h
(II EoT ~ ¹0.3h, subtract 24h. II EoT · 0.3h, add 24h.)
Semidiameter and Horizontal Parallax
Due to the excentricity oI the earth's orbit, semidiameter and horizontal parallax oI the sun change periodically during
the course oI a year. The SD oI the sun varies inversely with the distance earthsun, R:
(1 AU ÷ 149.6
.
10
6
km)
SD '
16.0
R AU
The mean horizontal parallax oI the sun is approx. 0.15'. The periodic variation oI HP is too small to be oI practical signiIicance.
Accuracy
The maximum error oI GHA and Dec is about ±0.6'. Results have been crosschecked with Interactive Computer
Ephemeris 0.51 (accurate to approx. 0.1'). Between the years 1900 and 2049, the error was smaller than ±0.3' in most
cases (100 dates chosen at random). EoT was accurate to approx. ±2s. In comparison, the maximum error oI GHA and
Dec extracted Irom the Nautical Almanac is approx. ±0.25' (Ior the sun) when using the interpolation tables. The error
oI SD is smaller than +0.1'.
152
GHA
Aries
° 0.9856474T 15UT h 100.46062
GHA ° GHA
Aries
° RA °
y 360
x
360
floor
x
360
GAT h
GHA°
15
12 h
R AU 1.00014 0.01671cos g 0.00014cos 2g
Chapter 16
Navigational Errors
Altitude errors
Apart Irom systematic errors which can be corrected to a large extent (see chapter 2), observed altitudes always contain
random errors caused by ,e.g., heavy seas, abnormal atmospheric reIraction, and the limited optical resolution oI the
human eye. Although good sextants are manuIactured to a mechanical precision oI ca. 0.1' 0.3', the standard
deviation oI an altitude measured with a marine sextant is approximately 1' under Iair working conditions. The
standard deviation may increase to several arcminutes due to disturbing Iactors or iI a bubble sextant or a plastic sextant
is used. Altitudes measured with a theodolite are considerably more accurate (0.1' 0.2').
Due to the inIluence oI random observation errors, lines oI position are more or less indistinct and are better considered
as bands of position.
Two intersecting bands oI position deIine an area of position (ellipse oI uncertainty). Fig. 161 illustrates the
approximate size and shape oI the ellipse oI uncertainty Ior a given pair oI position lines. The standard deviations (+x
Ior the Iirst altitude, +y Ior the second altitude) are indicated by grey lines.
The area oI position is smallest iI the angle between the bands is 90°. The most probable position is at the center oI the
area, provided the error distribution is symmetrical. Since position lines are perpendicular to their corresponding
azimuth lines, objects should be chosen whose azimuths diIIer by approx. 90° Ior best accuracy. An angle between 30°
and 150°, however, is tolerable in most cases.
When observing more than two bodies, the azimuths should have a roughly symmetrical distribution (bearing spread).
With multiple observations, the optimum horizontal angle between two adjacent bodies is obtained by dividing 360° by
the number oI observed bodies (3 bodies: 120°, 4 bodies: 90°, 5 bodies: 72°, 6 bodies: 60°, etc.).
A symmetrical bearing spread not only improves geometry but also compensates Ior systematic errors like, e.g., index
error.
Moreover, there is an optimum range oI altitudes the navigator should choose to obtain reliable results. Low altitudes
increase the inIluence oI abnormal reIraction (random error), whereas high altitudes, corresponding to circles oI equal
altitude with small diameters, increase geometric errors due to the curvature oI position lines. The generally
recommended range to be used is 20°  70°, but exceptions are possible.
161
Time errors
The time error is as important as the altitude error since the navigator usually presets the instrument to a chosen altitude
and records the time at the instant the image oI the body coincides with the reIerence line visible in the telescope. The
accuracy oI time measurement is usually in the range between a Iraction oI a second and several seconds, depending on
the rate oI change oI altitude and other Iactors. Time error and altitude error are closely interrelated and can be
converted to each other, as shown below (Fig. 162):
The GP oI any celestial body travels westward with an angular velocity oI approx. 0.25' per second. This is the rate oI
change oI the local hour angle oI the observed body caused by the earth's rotation. The same applies to each circle oI
equal altitude surrounding GP (tangents shown in Fig. 62). The distance between two concentric circles oI equal
altitude (with the altitudes H
1
and H
2
) passing through AP in the time interval dt, measured along the parallel oI latitude
going through AP is:
dx is also the eastwest displacement oI a LoP caused by the time error dt. The letter d indicates a small (inIinitesimal)
change oI a quantity (see mathematical literature). The cosine oI Lat
AP
is the ratio oI the circumIerence oI the parallel oI
latitude going through AP to the circumIerence oI the equator (Lat ÷ 0).
The corresponding diIIerence in altitude (the radial distance between both circles oI equal altitude) is:
Thus, the rate oI change oI altitude is:
dH/dt is greatest when the observer is on the equator and decreases to zero as the observer approaches one oI the poles.
Further, dH/dt is greatest iI GP is exactly east oI AP (dH/dt positive) or exactly west oI AP (dH/dt negative). dH/dt is
zero iI the azimuth is 0° or 180°. This corresponds to the Iact that the altitude oI the observed body passes through a
minimum or maximum at the instant oI meridian transit (dH/dt ÷ 0).
The maximum or minimum oI altitude occurs exactly at meridian transit only iI the declination oI a body is constant.
Otherwise, the highest or lowest altitude is observed shortly beIore or aIter meridian transit (chapter 6). The
phenomenon is particularly obvious when observing the moon whose declination changes rapidly.
A chronometer error is a systematic time error (chapter 17). It inIluences each line oI position in such a way that only
the longitude oI a Iix is aIIected whereas the latitude remains unchanged, provided the declination does not change
signiIicantly (moon!).
162
dx nm 0.25cos Lat
AP
dt s
dH ' sin Az
N
dx nm
dH '
dt s
0.25sin Az
N
cos Lat
AP
A chronometer being 1 s Iast, Ior example, displaces a Iix by 0.25' to the west, a chronometer being 1 s slow displaces
the Iix by the same amount to the east. II we know our position, we can calculate the chronometer error Irom the
diIIerence between our true longitude and the longitude Iound by our observations. II we do not know our longitude,
the approximate chronometer error can be Iound by lunar observations (chapter 7).
Ambiguity
Poor geometry may not only decrease accuracy but may even result in an entirely wrong Iix. As the observed horizontal
angle (diIIerence in azimuth) between two objects approaches 180°, the distance between the points oI intersection oI
the corresponding circles oI equal altitude becomes very small (at exactly 180°, both circles are tangent to each other).
Circles oI equal altitude with small diameters resulting Irom high altitudes also contribute to a short distance. A small
distance between both points oI intersection, however, increases the risk oI ambiguity (Fig. 163).
In cases where due to a horizontal angle near 180° and/or very high altitudes the distance between both points oI
intersection is too small, we can not be sure that the assumed position is always close enough to the actual position.
II AP is close to the actual position, the Iix obtained by plotting the LoP's (tangents) will be almost identical with the
actual position. The accuracy oI the Iix decreases as the distance oI AP Irom the actual position becomes greater. The
distance between Iix and actual position increases dramatically as AP approaches the line going through GP1 and GP2
(draw the azimuth lines and tangents mentally). In the worst case, a position error oI several hundred or even thousand
nm may result !
II AP is exactly on the line going through GP1 and GP2, i.e., equidistant Irom the actual position and the second point
oI intersection, the horizontal angle between GP1 and GP2, as viewed Irom AP, will be 180°. In this case, both LoP's
are parallel to each other, and no Iix can be Iound.
As AP approaches the second point oI intersection, a Iix more or less close to the latter is obtained. Since the actual
position and the second point oI intersection are symmetrical with respect to the line going through GP1 and GP2, the
intercept method can not detect which oI both theoretically possible positions is the right one.
Iterative application oI the intercept method can only improve the Iix iI the initial AP is closer to the actual position
than to the second point oI intersection. Otherwise, an "improved" wrong position will be obtained.
Each navigational scenario should be evaluated critically before deciding if a fix is reliable or not. The distance
from AP to the observer's actual position has to be considerably smaller than the distance between actual
position and second point of intersection. This is usually the case if the above recommendations regarding
altitude, horizontal angle, and distance between AP and actual position are observed.
163
A simple method to improve the reliability of a fix
Each altitude measured with a sextant, theodolite, or any other device contains systematic and random errors which
inIluence the Iinal result (Iix). Systematic errors are more or less eliminated by careIul calibration oI the instrument
used. The inIluence oI random errors decreases iI the number oI observations is suIIiciently large, provided the error
distribution is symmetrical. Under practical conditions, the number oI observations is limited, and the error
distribution is more or less unsymmetrical, particularly iI an outlier, a measurement with an abnormally large error, is
present. ThereIore, the average result may diIIer signiIicantly Irom the true value. When plotting more than two lines oI
position, the experienced navigator may be able to identiIy outliers by the shape oI the error polygon and remove the
associated LoP's. However, the method oI least squares, producing an average value, does not recognize outliers and
may yield an inaccurate result.
The Iollowing simple method takes advantage oI the Iact that the median oI a number oI measurements is much less
inIluenced by outliers than the mean value:
1. We choose a celestial body and measure a series oI altitudes. We calculate azimuth and intercept Ior each
observation oI said body. The number oI measurements in the series has to be odd (3, 5, 7...). The reliability oI the
method increases with the number oI observations.
2. We sort the calculated intercepts by magnitude and choose the median (the central value in the array oI intercepts
thus obtained) and its associated azimuth. We discard all other observations oI the series.
3. We repeat the above procedure with at least one additional body (or with the same body aIter its azimuth has
become suIIiciently diIIerent).
4. We plot the lines oI position using the azimuth and intercept selected Irom each series, or use the selected data to
calculate the Iix with the method oI least squares (chapter 4).
The method has been checked with excellent results on land. At sea, where the observer's position usually changes
continually, the method has to be modiIied by advancing AP according to the path oI travel between the observations oI
each series.
164
Chapter 17
The Marine Chronometer
A marine chronometer is a precise timepiece kept on board as a portable time standard. In Iormer times, the
chronometer time was usually checked (compared with an optical time signal) shortly beIore departure. During the
voyage, the chronometer had to be reliable enough to avoid dangerous longitude errors (chapter 16) even aIter weeks or
months at sea. Today, radio time signals, e. g., WWV, can be received around the world, and the chronometer can be
checked as oIten as desired during a voyage. ThereIore, a common quartz watch oI reasonable quality is suitable Ior
most navigational tasks iI checked periodically, and the marine chronometer serves more or less as a backup.
The Iirst mechanical marine chronometer oI suIIicient precision was built by Harrison in 1736. Due to the exorbitant
price oI early chronometers it took decades until the marine chronometer became part oI the standard navigation
equipment. In the meantime, the longitude oI a ship was mostly determined by lunar distance (chapter 7).
During the second halI oI the 20
th
century quartz chronometers replaced the mechanical ones almost entirely because
they are much more accurate, cheaper to manuIacture, and almost maintenanceIree.
Today, mechanical chronometers are valued collector's items since there are only Iew manuIacturers leIt. Fig. 71
shows a POLJOT () 6MX, a traditional chronometer made in Russia. Note that the timepiece is suspended in
gimbals to reduce the inIluence oI ship movements (torque) on the balance wheel.
Usage
Shortly beIore an astronomical observation, the navigator starts a stopwatch at a chosen integer hour or minute
displayed by the chronometer and makes a note oI the chronometer reading. Also beIore the observation, the sextant or
theodolite is set to a chosen altitude (unless a maximum or minimum altitude is to be observed). During the observation
itselI, the time is stopped at the instant the observed body makes contact with the horizontal reIerence line in the
telescope oI the instrument. This may be the sea horizon (sextant) or the cross hairs (theodolite). The sum oI the
previously noted chronometer time and the time measured with the stopwatch is the chronometer time oI observation.
Using a chronometer, the navigator has to know a number oI quantities since there is no guarantee that time signals are
available at any time.
The chronometer error, CE, is the diIIerence between chronometer time (time displayed by chronometer) and UT at
a given instant (usually the last chronometer check).
The most important individual characteristic oI a chronometer is the chronometer rate, the change oI the chronometer
error during a chosen time interval. The daily rate is measured in seconds per 24 hours. For better accuracy, the daily
rate is usually obtained by measuring the change oI the chronometer error within a 10day period and deviding the
result by 10. The chronometer rate can be positive (chronometer gaining) or negative (chronometer losing).
Knowing the initial error and daily rate, we can extrapolate the current UT Irom the current chronometer time and the
number oI days and hours expired since the last chronometer check.
171
CE T
Chrono
UT
The correction Iormula is:
UT
P
T
Chrono
CE
init
n
m
24
DR
UTP Predicted UT
TChrono Chronometer time
CEinit Initial chronometer error
DR Daily rate
n Integer number oI days expired since last chronometer check
m Additional number oI hours expired
The daily rate oI a chronometer is not constant but subject to systematic changes. ThereIore it should be measured
periodically. At the beginning oI the service liIe, the daily rate oI a chronometer changes and Iinally approaches a more
or less constant value (,running in'). Temperature variations also aIIect the daily rate. The temperature coefficient is
the change oI daily rate caused by a certain temperature variation. Chronometers are temperaturecompensated.
ThereIore, the temperature coeIIicient oI a mechanical marine chronometer is small, typically about +0.1 s/K per day.
Moreover, the daily rate exhibits random Iluctuations, called daily ratevariation, DRV. The latter is the diIIerence
between the daily rates measured on two consecutive days (day i and day i¹1, respectively).
Variations in daily rate are usually small but should be monitored regularly. A daily ratevariation outside the range
speciIied by the manuIacturer may indicate a mechanical problem, e. g., abnormal wear. In such a case, the instrument
needs overhauling or repair. The standard deviation oI a series oI n (usually 10) consecutive daily ratevariations, may
be regarded as a 'quality index', 22.
The Iollowing table shows a series oI measurements made with a mechanical chronometer (POLJOT 6MX #22787) at
room temperature. Like most mechanical marine chronometers, the 6MX has a halIsecond beat. The chronometer time
was compared with UTC (radiocontrolled watch, 1second beat). ThereIore, all values have been rounded to the next
halI second.
Day CE s DR s (1
st
DiII.) DRVs (2
nd
DiII.) DRV
2
s
2

0 41
2
1 39 1 1
3
2 36 0.5 0.25
2.5
3 33.5 0.5 0.25
3
4 30.5 0 0
3
5 27.5 0.5 0.25
2.5
6 25 0.5 0.25
3
7 22 0.5 0.25
2.5
8 19.5 0.5 0.25
3
9 16.5 0 0
3
10 13.5 0.5 0.25
2.5
11 11 Sum: 2.75
The mean daily rate oI the chronometer (Iirst 10 days) is 2.75 s (speciIication: +3.5 s). The maximum variation in daily
rate is 1 s (speciIication: +2.3 s). The 'quality index' is +0.52 s (not speciIied).
172
1
n
i1
n
DRJ
i
2
DRJ
i
DR
i 1
DR
i
Appendix
Literature :
1 Bowditch, The American Practical Navigator, Pub. No. 9, DeIense Mapping Agency
Hydrographic/Topographic Center, Bethesda, MD, USA
2 Jean Meeus, Astronomical Algorithms, WillmannBell, Inc., Richmond, VA, USA 1991
3 Bruce A. Bauer, The Sextant Handbook, International Marine, P.O. Box 220, Camden, ME 04843, USA
4 Charles H. Cotter, A History oI Nautical Astronomy, American Elsevier Publishing Company, Inc., New
York, NY, USA (This excellent book is out oI print. Used examples may be available at
www.amazon.com .)
5 Charles H. Brown, Nicholl's Concise Guide to the Navigation Examinations, Vol.II, Brown, Son & Ferguson,
Ltd., Glasgow, G41 2SG, UK
6 Helmut Knopp, Astronomische Navigation, Verlag Busse ¹ Seewald GmbH, HerIord, Germany (German)
7 Willi Kahl, Navigation Iür Expeditionen, Touren, Törns und Reisen, Schettler Travel Publikationen,
HattorI, Germany (German)
8 KarlRichard Albrand and Walter Stein, Nautische TaIeln und Formeln (German), DSVVerlag, Germany
9 William M. Smart, Textbook on Spherical Astronomy, 6
th
Edition, Cambridge University Press, 1977
10 P. K. Seidelman (Editor), Explanatory Supplement to the Astronomical Almanac, University Science Books,
Sausalito, CA 94965, USA
11 Allan E. Bayless, Compact Sight Reduction Table (modiIied H. O. 211, Ageton's Table), 2
nd
Edition, Cornell
Maritime Press, Centreville, MD 21617, USA
12 The Nautical Almanac (contains not only ephemeral data but also Iormulas and tables Ior sight reduction), US
Government Printing OIIice, Washington, DC 20402, USA
13 Nautisches Jahrbuch oder Ephemeriden und TaIeln (German), Bundesamt Iür SeeschiIIahrt und Hydrographie,
Germany
14 The Lunar Distance Page, http://www.lunardistance.com
15 IERS Rapid Service, http://maia.usno.navy.mil
16 Hannu Karttunen et al., Fundamental Astronomy, 4
th
Ed., Springer Verlag Berlin Heidelberg New York, 2003
17 The Astronomical Almanac Ior the Year 2002, US Government Printing OIIice, Washington, DC 20402, USA
18 Michel Vanvaerenberg and Peter IIland, Line oI Position Navigation, Unlimited Publishing, Bloomingdale,
Indiana, 2003
19 George H. Kaplan, The Motion oI the Observer in Celestial Navigation, Astronomical Applications
Department, U. S. Naval Observatory, Washington DC
20 Ed William's Aviation Formulary, http://williams.best.vwh.net/avIorm.htm
21 William Chauvenet, A Manual oI Spherical and Practical Astronomy, Vol. I ¹ II (reprint oI the edition
published in 1887 by J. B. Lippincot Company), Elibron Classics Series, Adamant Media Corporation, 2005
22 N. Liapin, A Method oI Determining the Mean Accidental Variations in Daily Rates oI a Number oI
Chronometers, Monthly Notices oI the Royal Astronomical Society 1919, Vol.80, p.64
Author's web site : http://www.celnav.de
GNU Free Documentation License
Version 1.3, 3 November 2008
Copyright © 2000, 2001, 2002, 2007, 2008 Free SoItware Foundation, Inc. ·http://IsI.org/~
Everyone is permitted to copy and distribute verbatim copies oI this license document, but changing
it is not allowed.
0. PREAMBLE
The purpose oI this License is to make a manual, textbook, or other Iunctional and useIul document
"Iree" in the sense oI Ireedom: to assure everyone the eIIective Ireedom to copy and redistribute it,
with or without modiIying it, either commercially or noncommercially. Secondarily, this License
preserves Ior the author and publisher a way to get credit Ior their work, while not being considered
responsible Ior modiIications made by others.
This License is a kind oI "copyleIt", which means that derivative works oI the document must
themselves be Iree in the same sense. It complements the GNU General Public License, which is a
copyleIt license designed Ior Iree soItware.
We have designed this License in order to use it Ior manuals Ior Iree soItware, because Iree
soItware needs Iree documentation: a Iree program should come with manuals providing the same
Ireedoms that the soItware does. But this License is not limited to soItware manuals; it can be used
Ior any textual work, regardless oI subject matter or whether it is published as a printed book. We
recommend this License principally Ior works whose purpose is instruction or reIerence.
1. APPLICABILITY AND DEFINITIONS
This License applies to any manual or other work, in any medium, that contains a notice placed by
the copyright holder saying it can be distributed under the terms oI this License. Such a notice
grants a worldwide, royaltyIree license, unlimited in duration, to use that work under the
conditions stated herein. The "Document", below, reIers to any such manual or work. Any member
oI the public is a licensee, and is addressed as "you". You accept the license iI you copy, modiIy or
distribute the work in a way requiring permission under copyright law.
A "ModiIied Version" oI the Document means any work containing the Document or a portion oI it,
either copied verbatim, or with modiIications and/or translated into another language.
A "Secondary Section" is a named appendix or a Irontmatter section oI the Document that deals
exclusively with the relationship oI the publishers or authors oI the Document to the Document's
overall subject (or to related matters) and contains nothing that could Iall directly within that overall
subject. (Thus, iI the Document is in part a textbook oI mathematics, a Secondary Section may not
explain any mathematics.) The relationship could be a matter oI historical connection with the
subject or with related matters, or oI legal, commercial, philosophical, ethical or political position
regarding them.
The "Invariant Sections" are certain Secondary Sections whose titles are designated, as being those
oI Invariant Sections, in the notice that says that the Document is released under this License. II a
section does not Iit the above deIinition oI Secondary then it is not allowed to be designated as
Invariant. The Document may contain zero Invariant Sections. II the Document does not identiIy
any Invariant Sections then there are none.
The "Cover Texts" are certain short passages oI text that are listed, as FrontCover Texts or Back
Cover Texts, in the notice that says that the Document is released under this License. A FrontCover
Text may be at most 5 words, and a BackCover Text may be at most 25 words.
A "Transparent" copy oI the Document means a machinereadable copy, represented in a Iormat
whose speciIication is available to the general public, that is suitable Ior revising the document
straightIorwardly with generic text editors or (Ior images composed oI pixels) generic paint
programs or (Ior drawings) some widely available drawing editor, and that is suitable Ior input to
text Iormatters or Ior automatic translation to a variety oI Iormats suitable Ior input to text
Iormatters. A copy made in an otherwise Transparent Iile Iormat whose markup, or absence oI
markup, has been arranged to thwart or discourage subsequent modiIication by readers is not
Transparent. An image Iormat is not Transparent iI used Ior any substantial amount oI text. A copy
that is not "Transparent" is called "Opaque".
Examples oI suitable Iormats Ior Transparent copies include plain ASCII without markup, TexinIo
input Iormat, LaTeX input Iormat, SGML or XML using a publicly available DTD, and standard
conIorming simple HTML, PostScript or PDF designed Ior human modiIication. Examples oI
transparent image Iormats include PNG, XCF and JPG. Opaque Iormats include proprietary Iormats
that can be read and edited only by proprietary word processors, SGML or XML Ior which the DTD
and/or processing tools are not generally available, and the machinegenerated HTML, PostScript or
PDF produced by some word processors Ior output purposes only.
The "Title Page" means, Ior a printed book, the title page itselI, plus such Iollowing pages as are
needed to hold, legibly, the material this License requires to appear in the title page. For works in
Iormats which do not have any title page as such, "Title Page" means the text near the most
prominent appearance oI the work's title, preceding the beginning oI the body oI the text.
The "publisher" means any person or entity that distributes copies oI the Document to the public.
A section "Entitled XYZ" means a named subunit oI the Document whose title either is precisely
XYZ or contains XYZ in parentheses Iollowing text that translates XYZ in another language. (Here
XYZ stands Ior a speciIic section name mentioned below, such as "Acknowledgements",
"Dedications", "Endorsements", or "History".) To "Preserve the Title" oI such a section when you
modiIy the Document means that it remains a section "Entitled XYZ" according to this deIinition.
The Document may include Warranty Disclaimers next to the notice which states that this License
applies to the Document. These Warranty Disclaimers are considered to be included by reIerence in
this License, but only as regards disclaiming warranties: any other implication that these Warranty
Disclaimers may have is void and has no eIIect on the meaning oI this License.
2. VERBATIM COPYING
You may copy and distribute the Document in any medium, either commercially or
noncommercially, provided that this License, the copyright notices, and the license notice saying
this License applies to the Document are reproduced in all copies, and that you add no other
conditions whatsoever to those oI this License. You may not use technical measures to obstruct or
control the reading or Iurther copying oI the copies you make or distribute. However, you may
accept compensation in exchange Ior copies. II you distribute a large enough number oI copies you
must also Iollow the conditions in section 3.
You may also lend copies, under the same conditions stated above, and you may publicly display
copies.
3. COPYING IN QUANTITY
II you publish printed copies (or copies in media that commonly have printed covers) oI the
Document, numbering more than 100, and the Document's license notice requires Cover Texts, you
must enclose the copies in covers that carry, clearly and legibly, all these Cover Texts: FrontCover
Texts on the Iront cover, and BackCover Texts on the back cover. Both covers must also clearly
and legibly identiIy you as the publisher oI these copies.
The Iront cover must present the Iull title with all words oI the title equally prominent and visible.
You may add other material on the covers in addition. Copying with changes limited to the covers,
as long as they preserve the title oI the Document and satisIy these conditions, can be treated as
verbatim copying in other respects.
II the required texts Ior either cover are too voluminous to Iit legibly, you should put the Iirst ones
listed (as many as Iit reasonably) on the actual cover, and continue the rest onto adjacent pages.
II you publish or distribute Opaque copies oI the Document numbering more than 100, you must
either include a machinereadable Transparent copy along with each Opaque copy, or state in or
with each Opaque copy a computernetwork location Irom which the general networkusing public
has access to download using publicstandard network protocols a complete Transparent copy oI the
Document, Iree oI added material. II you use the latter option, you must take reasonably prudent
steps, when you begin distribution oI Opaque copies in quantity, to ensure that this Transparent
copy will remain thus accessible at the stated location until at least one year aIter the last time you
distribute an Opaque copy (directly or through your agents or retailers) oI that edition to the public.
It is requested, but not required, that you contact the authors oI the Document well beIore
redistributing any large number oI copies, to give them a chance to provide you with an updated
version oI the Document.
4. MODIFICATIONS
You may copy and distribute a ModiIied Version oI the Document under the conditions oI sections 2
and 3 above, provided that you release the ModiIied Version under precisely this License, with the
ModiIied Version Iilling the role oI the Document, thus licensing distribution and modiIication oI
the ModiIied Version to whoever possesses a copy oI it. In addition, you must do these things in the
ModiIied Version:
A. Use in the Title Page (and on the covers, iI any) a title distinct Irom that oI the Document,
and Irom those oI previous versions (which should, iI there were any, be listed in the History
section oI the Document). You may use the same title as a previous version iI the original
publisher oI that version gives permission.
B. List on the Title Page, as authors, one or more persons or entities responsible Ior
authorship oI the modiIications in the ModiIied Version, together with at least Iive oI the
principal authors oI the Document (all oI its principal authors, iI it has Iewer than Iive),
unless they release you Irom this requirement.
C. State on the Title page the name oI the publisher oI the ModiIied Version, as the
publisher.
D. Preserve all the copyright notices oI the Document.
E. Add an appropriate copyright notice Ior your modiIications adjacent to the other
copyright notices.
F. Include, immediately aIter the copyright notices, a license notice giving the public
permission to use the ModiIied Version under the terms oI this License, in the Iorm shown in
the Addendum below.
G. Preserve in that license notice the Iull lists oI Invariant Sections and required Cover Texts
given in the Document's license notice.
H. Include an unaltered copy oI this License.
I. Preserve the section Entitled "History", Preserve its Title, and add to it an item stating at
least the title, year, new authors, and publisher oI the ModiIied Version as given on the Title
Page. II there is no section Entitled "History" in the Document, create one stating the title,
year, authors, and publisher oI the Document as given on its Title Page, then add an item
describing the ModiIied Version as stated in the previous sentence.
J. Preserve the network location, iI any, given in the Document Ior public access to a
Transparent copy oI the Document, and likewise the network locations given in the
Document Ior previous versions it was based on. These may be placed in the "History"
section. You may omit a network location Ior a work that was published at least Iour years
beIore the Document itselI, or iI the original publisher oI the version it reIers to gives
permission.
K. For any section Entitled "Acknowledgements" or "Dedications", Preserve the Title oI the
section, and preserve in the section all the substance and tone oI each oI the contributor
acknowledgements and/or dedications given therein.
L. Preserve all the Invariant Sections oI the Document, unaltered in their text and in their
titles. Section numbers or the equivalent are not considered part oI the section titles.
M. Delete any section Entitled "Endorsements". Such a section may not be included in the
ModiIied Version.
N. Do not retitle any existing section to be Entitled "Endorsements" or to conIlict in title
with any Invariant Section.
O. Preserve any Warranty Disclaimers.
II the ModiIied Version includes new Irontmatter sections or appendices that qualiIy as Secondary
Sections and contain no material copied Irom the Document, you may at your option designate
some or all oI these sections as invariant. To do this, add their titles to the list oI Invariant Sections
in the ModiIied Version's license notice. These titles must be distinct Irom any other section titles.
You may add a section Entitled "Endorsements", provided it contains nothing but endorsements oI
your ModiIied Version by various partiesIor example, statements oI peer review or that the text
has been approved by an organization as the authoritative deIinition oI a standard.
You may add a passage oI up to Iive words as a FrontCover Text, and a passage oI up to 25 words
as a BackCover Text, to the end oI the list oI Cover Texts in the ModiIied Version. Only one
passage oI FrontCover Text and one oI BackCover Text may be added by (or through
arrangements made by) any one entity. II the Document already includes a cover text Ior the same
cover, previously added by you or by arrangement made by the same entity you are acting on behalI
oI, you may not add another; but you may replace the old one, on explicit permission Irom the
previous publisher that added the old one.
The author(s) and publisher(s) oI the Document do not by this License give permission to use their
names Ior publicity Ior or to assert or imply endorsement oI any ModiIied Version.
5. COMBINING DOCUMENTS
You may combine the Document with other documents released under this License, under the terms
deIined in section 4 above Ior modiIied versions, provided that you include in the combination all
oI the Invariant Sections oI all oI the original documents, unmodiIied, and list them all as Invariant
Sections oI your combined work in its license notice, and that you preserve all their Warranty
Disclaimers.
The combined work need only contain one copy oI this License, and multiple identical Invariant
Sections may be replaced with a single copy. II there are multiple Invariant Sections with the same
name but diIIerent contents, make the title oI each such section unique by adding at the end oI it, in
parentheses, the name oI the original author or publisher oI that section iI known, or else a unique
number. Make the same adjustment to the section titles in the list oI Invariant Sections in the license
notice oI the combined work.
In the combination, you must combine any sections Entitled "History" in the various original
documents, Iorming one section Entitled "History"; likewise combine any sections Entitled
"Acknowledgements", and any sections Entitled "Dedications". You must delete all sections
Entitled "Endorsements".
6. COLLECTIONS OF DOCUMENTS
You may make a collection consisting oI the Document and other documents released under this
License, and replace the individual copies oI this License in the various documents with a single
copy that is included in the collection, provided that you Iollow the rules oI this License Ior
verbatim copying oI each oI the documents in all other respects.
You may extract a single document Irom such a collection, and distribute it individually under this
License, provided you insert a copy oI this License into the extracted document, and Iollow this
License in all other respects regarding verbatim copying oI that document.
7. AGGREGATION WITH INDEPENDENT WORKS
A compilation oI the Document or its derivatives with other separate and independent documents or
works, in or on a volume oI a storage or distribution medium, is called an "aggregate" iI the
copyright resulting Irom the compilation is not used to limit the legal rights oI the compilation's
users beyond what the individual works permit. When the Document is included in an aggregate,
this License does not apply to the other works in the aggregate which are not themselves derivative
works oI the Document.
II the Cover Text requirement oI section 3 is applicable to these copies oI the Document, then iI the
Document is less than one halI oI the entire aggregate, the Document's Cover Texts may be placed
on covers that bracket the Document within the aggregate, or the electronic equivalent oI covers iI
the Document is in electronic Iorm. Otherwise they must appear on printed covers that bracket the
whole aggregate.
8. TRANSLATION
Translation is considered a kind oI modiIication, so you may distribute translations oI the Document
under the terms oI section 4. Replacing Invariant Sections with translations requires special
permission Irom their copyright holders, but you may include translations oI some or all Invariant
Sections in addition to the original versions oI these Invariant Sections. You may include a
translation oI this License, and all the license notices in the Document, and any Warranty
Disclaimers, provided that you also include the original English version oI this License and the
original versions oI those notices and disclaimers. In case oI a disagreement between the translation
and the original version oI this License or a notice or disclaimer, the original version will prevail.
II a section in the Document is Entitled "Acknowledgements", "Dedications", or "History", the
requirement (section 4) to Preserve its Title (section 1) will typically require changing the actual
title.
9. TERMINATION
You may not copy, modiIy, sublicense, or distribute the Document except as expressly provided
under this License. Any attempt otherwise to copy, modiIy, sublicense, or distribute it is void, and
will automatically terminate your rights under this License.
However, iI you cease all violation oI this License, then your license Irom a particular copyright
holder is reinstated (a) provisionally, unless and until the copyright holder explicitly and Iinally
terminates your license, and (b) permanently, iI the copyright holder Iails to notiIy you oI the
violation by some reasonable means prior to 60 days aIter the cessation.Moreover, your license
Irom a particular copyright holder is reinstated permanently iI the copyright holder notiIies you oI
the violation by some reasonable means, this is the Iirst time you have received notice oI violation
oI this License (Ior any work) Irom that copyright holder, and you cure the violation prior to 30
days aIter your receipt oI the notice.
Termination oI your rights under this section does not terminate the licenses oI parties who have
received copies or rights Irom you under this License. II your rights have been terminated and not
permanently reinstated, receipt oI a copy oI some or all oI the same material does not give you any
rights to use it.
10. FUTURE REVISIONS OF THIS LICENSE
The Free SoItware Foundation may publish new, revised versions oI the GNU Free Documentation
License Irom time to time. Such new versions will be similar in spirit to the present version, but
may diIIer in detail to address new problems or concerns. See http://www.gnu.org/copyleIt/.
Each version oI the License is given a distinguishing version number. II the Document speciIies that
a particular numbered version oI this License "or any later version" applies to it, you have the
option oI Iollowing the terms and conditions either oI that speciIied version or oI any later version
that has been published (not as a draIt) by the Free SoItware Foundation. II the Document does not
speciIy a version number oI this License, you may choose any version ever published (not as a
draIt) by the Free SoItware Foundation. II the Document speciIies that a proxy can decide which
Iuture versions oI this License can be used, that proxy's public statement oI acceptance oI a version
permanently authorizes you to choose that version Ior the Document.
11. RELICENSING
"Massive Multiauthor Collaboration Site" (or "MMC Site") means any World Wide Web server that
publishes copyrightable works and also provides prominent Iacilities Ior anybody to edit those
works. A public wiki that anybody can edit is an example oI such a server. A "Massive Multiauthor
Collaboration" (or "MMC") contained in the site means any set oI copyrightable works thus
published on the MMC site.
"CCBYSA" means the Creative Commons AttributionShare Alike 3.0 license published by
Creative Commons Corporation, a notIorproIit corporation with a principal place oI business in
San Francisco, CaliIornia, as well as Iuture copyleIt versions oI that license published by that same
organization.
"Incorporate" means to publish or republish a Document, in whole or in part, as part oI another
Document.
An MMC is "eligible Ior relicensing" iI it is licensed under this License, and iI all works that were
Iirst published under this License somewhere other than this MMC, and subsequently incorporated
in whole or in part into the MMC, (1) had no cover texts or invariant sections, and (2) were thus
incorporated prior to November 1, 2008.
The operator oI an MMC Site may republish an MMC contained in the site under CCBYSA on the
same site at any time beIore August 1, 2009, provided the MMC is eligible Ior relicensing.
ADDENDUM: How to use this License for your documents
To use this License in a document you have written, include a copy oI the License in the document
and put the Iollowing copyright and license notices just aIter the title page:
CopyJ1gh1 {C) YEAR Y0bR hAhE.
PeJm1ss1oh 1s gJah1ed 1o copy, d1s1J1bu1e ahd/oJ mod11y 1h1s documeh1
uhdeJ 1he 1eJms o1 1he Chb FJee 0ocumeh1a11oh L1cehse, veJs1oh 1.3
oJ ahy ìa1eJ veJs1oh pubì1shed by 1he FJee So11waJe Fouhda11oh,
w11h ho ThvaJ1ah1 Sec11ohs, ho FJoh1CoveJ Tex1s, ahd ho BackCoveJ Tex1s.
A copy o1 1he ì1cehse 1s 1hcìuded 1h 1he sec11oh eh111ìed "Chb
FJee 0ocumeh1a11oh L1cehse".
II you have Invariant Sections, FrontCover Texts and BackCover Texts, replace the "with .
Texts." line with this:
w11h 1he ThvaJ1ah1 Sec11ohs be1hg LTST ThETR TTTLES, w11h 1he
FJoh1CoveJ Tex1s be1hg LTST, ahd w11h 1he BackCoveJ Tex1s be1hg LTST.
II you have Invariant Sections without Cover Texts, or some other combination oI the three, merge
those two alternatives to suit the situation.
II your document contains nontrivial examples oI program code, we recommend releasing these
examples in parallel under your choice oI Iree soItware license, such as the GNU General Public
License, to permit their use in Iree soItware.
This action might not be possible to undo. Are you sure you want to continue?
We've moved you to where you read on your other device.
Get the full title to continue reading from where you left off, or restart the preview.