You are on page 1of 76

Distributed Peer Review

M87 Event Horizon Telescope Results

The PHANGS Surveys
Total Solar Eclipse Over La Silla
The Messenger
No. 177 – Quarter 3 | 2019
ESO, the European Southern Observa- Contents
tory, is the foremost intergovernmental
astronomy organisation in Europe. It is Telescopes and Instrumentation
supported by 16 Member States: Austria, Patat F. et al. – The Distributed Peer Review Experiment 3
­Belgium, the Czech Republic, Denmark, Coccato L. et al. – On the Telluric Correction of KMOS Spectra 14
France, Finland, Germany, Ireland, Italy, Gonté F. et al. – Bringing the New Adaptive Optics Module
the Netherlands, Poland, Portugal, Spain, for Interferometry (NAOMI) into Operation 19
Sweden, Switzerland and the United
Kingdom, along with the host country of Astronomical Science
Chile and with Australia as a Strategic Goddi C. et al. – First M87 Event Horizon Telescope Results
Partner. ESO’s programme is focussed and the Role of ALMA 25
on the design, construction and opera- Schinnerer E. et al. – The Physics at High Angular resolution in
tion of powerful ground-based observing Nearby GalaxieS (PHANGS) Surveys 36
­facilities. ESO operates three observato-
ries in Chile: at La Silla, at P
­ aranal, site of Astronomical News
the Very Large Telescope, and at Llano Ventura L. et al. – Total Solar Eclipse Over La Silla 43
de Chajnantor. ESO is the European Christensen L. L. et al. – Science & Outreach at La Silla During the
­partner in the Atacama Large Millimeter/ Total Solar Eclipse 47
submillimeter Array (ALMA). Currently Dennefeld M. et al. – Pointing the NTT at the Sun: Studying the Solar Corona
ESO is engaged in the construction of the During the Total Eclipse 54
Extremely Large ­Telescope. Sani E. et al. – Report on the ESO Workshop “KMOS@5: Star and
Galaxy Formation in 3D — Challenges in KMOS 5th Year” 56
The Messenger is published, in hardcopy Liske J., Mainieri V. – Report on the ESO Workshop “Preparing for 4MOST —
and electronic form, four times a year. A Community Workshop Introducing ESO’s Next-Generation Spectroscopic
ESO produces and distributes a wide Survey Facility 61
variety of media ­connected to its activi- Mroczkowski T. et al. – Report on the ESO Workshop “ALMA Development
ties. For further information, including Workshop” 64
postal subscription to The Messenger, Mérand A., Leibundgut B. – Report on the ESO Workshop “The VLT in 2030” 67
contact the ESO Department of Commu- Yang C. – Fellows at ESO 70
nication at: Jethwa P., Oikonomou F. – External Fellows at ESO 71
Hofstadt D. – Lodewijk Woltjer (1930–2019) 74
ESO Headquarters Personnel Movements 75
Karl-Schwarzschild-Straße 2
85748 Garching bei München, Germany
Phone +498932006-0

The Messenger
Editor: Gaitee A. J. Hussain
Layout, Typesetting, Graphics:
Jutta B
­ oxheimer, Mafalda Martins,
Lorenzo Benassi
Design, P­ roduction: Jutta ­Boxheimer
Proofreading: Peter Grimley,
Caroline Reid

Printed by FIBO Druck- und Verlags GmbH

Fichtenstraße 8, 82061 Neuried, Germany

Unless otherwise indicated, all images in

The Messenger are courtesy of ESO,
except authored contributions which are
courtesy of the respective authors.

© ESO 2019
Front cover: A series of exposures showing the tra-
ISSN 0722-6691 jectory of the Sun over roughly two and a half hours.
The total solar eclipse resulted in almost two minutes
of totality at 20:39 UT. Credit: ESO/P. Horálek

2 The Messenger 177 – Quarter 3 | 2019

Telescopes and Instrumentation DOI: 10.18727/0722-6691/5147

The Distributed Peer Review Experiment

Ferdinando Patat 1 the significant growth of the user com- example, late dropouts during the review
Wolfgang Kerzendorf 2, 3, 4 munity, which has made ESO one of the process can reduce the number of
Dominic Bordelon 1 largest astronomical facilities in the world, pre-meeting reviews per proposal, mak-
Glen Van de Ven 5 the way telescope time applications are ing the triage procedure less robust.
Tyler Pritchard 2 reviewed has remained substantially the While this change was relatively easy to
same since 1993. Barring the necessary implement, experience gained during
increase in the number of reviewers, the Periods 102 and 103 suggests that the
ESO procedure has changed in the details, negative consequences outweigh the
Center for Cosmology and Particle but not in its substance. Following steady benefits. It is clear that further and more
Physics, New York University, USA growth in the numbers of submissions, drastic and structured actions need to
Department of Physics and Astronomy, the current review load is about 70 pro- be taken; these include a move to an
Michigan State University, USA posals per panel member and up to 100 annual cycle and the deployment of a fast
Department of Computational Mathe- for OPC-proper members (the latter serve track channel (FTC; see Patat, 2018a).
matics, Science and Engineering, on a second panel which reviews the
Michigan State University, USA ­recommendations across all science cat- By construction, the FTC requires a short
Department of Astrophysics, University egories). These numbers have reached duty cycle during which referees are
of Vienna, Austria critical levels, requiring a re-evaluation of continuously on duty. The most suitable
the procedures and an examination of the mechanism for reviewing the proposals
effectiveness of peer review. is a Distributed Peer Review (DPR), one
All large, ground- and space-based of the most innovative schemes through
astronomical facilities serving wide The pressure on the peer review process which the load on referees can be allevi-
communities face a similar problem: in has been the subject of a study by the ated (Merrifield & Saari, 2009). This con-
many cases the number of applications ESO OPC Working Group (Brinks et al., cept has been successfully applied to
they receive in response to each call 2012) and the Time Allocation Working the Fast Turnaround channel deployed at
exceeds 1000. This poses a serious Group (TAWG; Patat, 2018a). Both stud- the Gemini Telescope, which has pro-
challenge to running an effective selec- ies identified the excessive number of cessed over 1000 proposals in this way
tion process under the classic peer-­ proposals per referee as the most urgent since 2015. The Gemini Observatory has
review paradigm, in which the propos- problem that ESO needs to tackle. Not published a report (Andersen et al., 2019)
als are assigned to pre-allocated panels only does the workload severely affect and updates are continuously provided
with fixed compositions. Although, in the referees (also increasing the rejection on its webpages 1.
principle, one could increase the size of rate during the recruitment phase), but it
the time allocation committee, this cre- can also have an impact on the quality Depending on the fraction of total tele-
ates logistic and financial problems of the reviews and the feedback provided scope time that is allocated via the FTC,
which place a practical limit on its to the applicants, with potentially serious this channel may also serve to decrease
maxi­mum size, making this solution consequences. The feedback has been the load on the OPC, which would then
unviable beyond a certain volume of repeatedly and consistently identified focus only on proposals with larger time
applications. For this reason, alternative as a major problem by the OPC and the requests. ESO has conducted a system-
solutions must be sought. One of these Users Committee, and via direct commu- atic study aimed at better evaluating the
is the so-called Distributed Peer Review nications from numerous individual users. application of DPR to its programmes.
(DPR) in which, by submitting a pro- Problems with the peer review could In Period 103, in parallel with the regular
posal, the Principal Investigators (PIs) ­ultimately affect the scientific productivity OPC cycle, a DPR experiment was run
agree both to act as reviewers and and impact of the Organisation itself. A involving a subset of submitted propos-
to have their proposal reviewed by their number of recommendations have been als. This article presents a brief descrip-
peers. In this article we report the proposed by the working groups, some tion of the experiment setup and summa-
results of a DPR experiment run by ESO of which are interdependent. rises an analysis of several statistical
in Period 103, in parallel with the regular indicators. More details can be found in
review by the Observing Programmes As a first step, since Period 102 ESO has Kerzendorf et al. (2019).
Committee (OPC). decreased the number of referees (from
six to three) who review a proposal
ahead of the OPC meeting. Triage is then Distributed Peer Review and the DPR
Introduction applied using the three pre-OPC meeting Experiment
grades, with about the lowest 30% of
Following the start of VLT operations in proposals being rejected. At the meeting Different measures to alleviate the load
1998, the number of applications to all non-conflicted panel members are on the reviewers have been and are
use ESO telescopes has been steadily then asked to discuss and grade only the being considered by various facilities.
­growing, exceeding 1100 proposals in surviving proposals. While this measure These include drastic solutions, like the
Period 84. After this peak, the number of has successfully reduced the workload one deployed by the National Science
submissions per semester stabilised at of the panel members, it has become Foundation (NSF, USA) to limit the num-
around 900 (Patat et al., 2017). Despite cumbersome to manage in practice. For ber of applications (Mervis, 2014a). The

The Messenger 177 – Quarter 3 | 2019 3

Telescopes and Instrumentation Patat F. et al., The Distributed Peer Review Experiment

Distributed Peer Review (DPR) concept is participate in the experiment. This implied complete mismatch (orthogonal knowl-
simple; in submitting a proposal the PI that each would review eight proposals edge vectors), while a unit cosine indi-
agrees to review n proposals submitted submitted by peers and have their pro- cates a case of perfect match (parallel
by peers, and to have her/his proposal/s posal refereed by the same number of knowledge vectors). For the purposes of
reviewed by n peers. Also, if s/he submits peers. The participants were given two the statistical analysis, each DT referee
m proposals, s/he accepts to review weeks to complete their reviews and received four proposals with the largest
n × m proposals, hence essentially limit- were informed that the outcome of the similarity, two proposals with median
ing the number of submissions through DPR would have no effect on the fate of similarity, and two proposals with the
a self-regulating mechanism. Following their proposals. By the deadline (22 Octo- lowest similarity.
this idea, the Gemini Observatory ber 2018) 167 (97.1%) had completed their
deployed the DPR for its Fast Turnaround task. In a real implementation the five PIs The participants were not aware of the
channel (Andersen et al., 2019), which is who did not meet the deadline would distribution mechanism just described.
capped to 10% of the total time. The have had their proposals automatically They were just provided with a simple
NSF also explored this possibility with a rejected. In this experiment however, their web-based interface giving them access
pilot study in 2013, in which each PI was proposals were kept in the sample, but to the eight assigned proposals and
asked to review seven proposals sub­ the PIs did not receive the final feedback. allowing them to review, grade and com-
mitted by peers (Ardabili & Liu, 2013; Additionally, the parti­cipating PIs were ment on the applications. Before access-
Mervis, 2014b). The NSF pilot was based asked to fill in a web-based questionnaire ing the proposals, the referees were
on 131 applications submitted by volun- covering various aspects of the experi- asked to sign a non-disclosure agree-
teers within the Civil, Mechanical and ment. A total of 140 (83.8% of the DPR ment, very similar to that signed by the
Manufacturing Innovation Division, but sample, 19% of the total PI sample of OPC and Panel members.
the outcome is unknown as no report on P103) returned the completed form.
the study was published. Interestingly, During the review phase, the participants
a similar pilot experiment was carried out The proposal distribution was performed were also asked to declare any scientific/
in 2016 by the National Institute of Food using two channels, which we will call personal conflicts, while institutional
and Agriculture 2; in this case too the OPC Emulate (OE) and DeepThought ­conflicts were automatically taken into
results were not published. Despite the (DT). In both cases the reviewers were account by the distribution software,
general acceptance that followed the assigned eight proposals each. For the based on the affiliations recorded in the
deployment of this channel at the Gemini OE channel, 60 volunteers were selected User Portal database. For each proposal,
Observatory, to the best of our knowl- at random and assigned, on the basis the referees had to fill in a comment (with
edge the Fast Turnaround channel is the of the category of the proposal each sub- a minimum length of 80 characters), and
only example of DPR being employed by mitted, to the four scientific categories: also provide a self-evaluation of their
a large-scale astronomical facility. A (Cosmology), B (Galaxy Structure and expertise level (high/medium/low) for
Evolution), C (Planets, Star ­Formation and each proposal assigned to them.
In the specific case of ESO, the TAWG Interstellar Medium) and D (Stellar Evolu-
tasked to address these issues has pro- tion). The underlying (and reasonable) Once the review process was completed,
duced a set of recommendations. The assumption is that a scientist submitting the grades of the various referees were
core aim is to reduce the number of a proposal for a given category is an combined using a simple average (similar
applications per reviewer, which has expert in that same area. This emulates to the regular OPC process), and a final
been identified as an urgent action that the case of the real OPC, in which a per- ranking list was compiled. The PIs were
ESO needs to take (Patat, 2018a). The son only receives proposals within her/his then provided with the quartile rank and
deployment of DPR falls within the rec- area of expertise. the individual, unedited anonymous
ommendations. As a first step, and after ­comments. Finally, they were asked to
consulting the advisory bodies, ESO For the remaining 112 volunteers selected provide feedback on the experiment via
decided to run a test during the ESO for the DT channel, the process was as a web-based form; this included a request
Period 103 in parallel to the regular OPC follows. For each scientist, a knowledge to express the usefulness of each com-
review. The experiment was designed in vector was built based on their publica- ment they received on their proposal.
line with the implementation at Gemini, tions, which were downloaded from the
enhancing the process by means of public SAO/NASA Astrophysics Data Sys-
­Natural Language Processing (NLP) and tem database (ADS) and processed by General statistics and demographics
Machine Learning (a different method a machine learning algorithm ­(Kerzendorf,
of using NLP for proposal reviews can be 2017). The same approach was used for Although, in principle, each proposal
found in Strolger et al., 2017). the proposals and applied to their scien- should have been reviewed by eight sci-
tific rationale. The match between the entists and each scientist should have
The DPR experiment was announced in ­referee expertise and the area covered by reviewed eight proposals, because of the
the Call for Proposals for Period 103, the proposal was then quantified through scientific/personal conflicts declared
released on 30 August 2018. A total of the “cosine distance”, which is directly ­during the refereeing process (and to a
172 PIs — representing 23% of all distinct related to the angle formed by the two much smaller extent because five partici-
PIs in that semester — volunteered to hyper-vectors; a null cosine signals a pants did not complete the process),

4 The Messenger 177 – Quarter 3 | 2019

both these numbers were on average Figure 1. Scientific
smaller than eight. The number of review- Seniority (this work) seniority distribution of
the DPR sample (blue)
ers Nr ranged from 4 to 8, with an aver- Seniority (Patat 16)
0.4 and the OPC sample
age of 7.3; in 95% of the cases the num- (orange). From Patat
ber was Nr ≥ 6. The number of proposals (2016).
Np varied from 5 to 8, with an average of 0.3
7.6, and Nr ≥ 6 in 98% of cases. The DPR
produced a total of 4055 distinct grade
pairs, to be compared with the maximum 0.2
number of pairs 172 × 8 × 7/2 = 4816
(see below for more details) one would
obtain in the case of no conflicts and no 0.1

The F/M gender distribution of the DPR 0.0

yet ye a
rs ars ars
participants (32/68) and the scientific hD 4 12 ye 12 ye
N oP tha n n4 – a n
seniority distribution derived from the Le s
s we e e th
B et Mor
DPR questionnaire (see Figure 1) reflect
the underlying PI population of ESO users
(Patat, 2016). Since participation in Figure 2. Distribution of
the number of proposals
the experiment was on a completely vol- 0.5 submitted to ESO by the
untary basis, we cannot exclude the DPR participants.
presence of self-selection biases. For
instance, one could argue that research- 0.4
ers who already had a positive opinion
of the DPR concept would be more will-

ing to participate than opponents, hence 0.3

introducing systematics into the final
analysis. On the other hand, if the com-
munity were strongly against the para- 0.2
digm, one would expect a similar effect.
In general, although we cannot guarantee
that there are no specific attributes that 0.1
lead the participants to self-selection, the
demographics indicate that, if they exist,
they are well hidden. Fewer than Between 3 and More than
3 proposals 10 proposals 10 proposals
An important aspect regarding the
­demographics of the experiment con-
cerns the fraction of junior scientists. there are published studies that indicate one single proposal sub-category (within
Since, as a rule, the regular panel mem- reviewers who self-report higher levels a given scientific category), the panel
bers serving on the OPC are required to of expertise tend to be less generous in members are requested to identify three
have a minimum seniority level (typically assigning the top grades (Gallo et al., sub-categories, ranking them in order
starting with scientists at their second 2016), the differences seen between the of expertise. This information is then used
postdoc onward), this establishes a sig- grade distributions of senior and junior to compose review panels in such a way
nificant difference between the two pools DPR participants are not statistically that the expertise coverage within each
of reviewers. In the case of the OPC, the significant. of them is as broad as possible. This is
distribution is heavily skewed towards required by any schema in which physical
senior members (88%), with a small frac- panels exist, which is in turn a constraint
tion of postdocs (12%) and no students Referee-Proposal matching stemming from the fact that the panels
(Patat, 2016), while the postdoc and stu- have to meet face-to-face and discuss
dent reviewers reach about 18% in the In the regular OPC process, the panel the same set of proposals. This intro-
case of the DPR sample (Figure 1). members are recruited to cover the widest duces a certain rigidity, which is also
possible range of astrophysical areas. related to the relatively small number of
Most DPR participants were relatively Each of the selected reviewers is asked available reviewers.
experienced in submitting proposals (Fig- to declare her/his expertise by providing
ure 2), although almost 60% of them sub-categories from the same list used Since DPR has the advantage of involving
had never served on a time allocation by the applicants to categorise their pro- a much larger number of reviewers, it
commitee before (Figure 3). Although posal. While the PI is allowed to indicate allows a significantly more flexible and

The Messenger 177 – Quarter 3 | 2019 5

Telescopes and Instrumentation Patat F. et al., The Distributed Peer Review Experiment

more objective approach in which, for 0.6 Figure 3. Distribution

of expertise in serving
each proposal, an ad hoc, optimised
on Time Allocation
panel can be formed. A key ingredient in Committees (TAC) for
this approach is the proposal-referee the DPR participants.
matching, which should work without the
need for human supervision, especially 0.4
when the turnaround has to be fast.

For this purpose, the DT algorithm used 0.3

in the DPR experiment was designed to
predict what we call domain expertise, 0.2
which in this context can be considered
to be the objective ability of a given sci-
entist to review a given proposal. Before 0.1
we discuss its reliability, we examine how
referees assessed their own ability to 0.0
review each proposal assigned to them. Never served Served once Served multiple times
As anticipated in the introduction, during on TAC on TAC on TAC
the refereeing process each participant
was asked to express their self-perceived Figure 4. Distribution of
Negative, no PhD yet self-reported domain
expertise level for each of the assigned
knowledge for the differ-
proposals, resulting in about 1200 eva­ Less than 4 years ent scientific seniority of
luations. The distribution of participants’ 0.4
Between 4 and 12 years the DPR participants.
self-evaluated ability to review the assigned
More than 12 years
proposals is presented in Fig­ure 4, where
we have used different ­colours for the 0.3

­different classes of scientific seniority. As

expected, junior scientists tend to perceive
themselves as experts less often than 0.2
senior scientists do. Also, they often indi-
cate that they have limited knowledge of
a given field. We take this is an indication
that the self-evaluated ability of a referee 0.1
to review the assigned proposals is a
useful proxy of the more objective (albeit
more abstract) concept of domain 0.0
Expert General knowledge No knowledge

The data collected in the DPR experiment expertise, which can be considered as a perceive them to be the top and interme-
enable an additional analysis of a possi- reasonable first approximation to the diate classes. As shown in Figure 5, the
ble gender dependence on the above underlying domain knowledge. From a correlation in the intermediate cases
self-evaluation. This has been reported, statistical point of view, this is equivalent becomes fuzzier. With the available data
for instance, by Huang (2013), who con- to computing the Bayesian conditional it is impossible to tell which of the two
cluded that females tend to under-predict probability P (self-reported | DT) of having estimators is responsible for the observed
their performance in certain STEM fields. a certain self-­reported expertise level, noise. If on the one hand we can argue
Our data suggest that, at least for post- given the DT-­inferred level. In simpler that the DT approach has obvious limita-
graduates in the domain of astrophysics, words, one checks how the self-reported tions (which is certainly true), on the other
there is no statistically significant gender and DT-­inferred levels correlate. The hand the self-reported levels are affected
difference. result is p
­ resented in Figure 5, which by a significant level of uncertainty, as
shows an encouragingly high correlation. they are related to subjective perceptions
Since the DT is designed to predict the For instance, the probability that the DT rather than to objective criteria.
expertise of a referee with respect to a considers a match as the worst which
given proposal, the first question one the referee believes is the best, is less Another aspect is the importance of
should ask is how reliable the algorithm then 1%. At the other extreme, it is very proper proposal-referee matching. Our
is. Obviously, there is no absolute refer- likely (78%) that if the DT estimates the direct experience, accumulated over
ence; the DT is one possible objective match is poor, the referee is of the same many years of managing the review pro-
estimate of this quality. Therefore, as a opinion. The agreement on the best cess at ESO, shows that, in addition to
first exploratory test, one can check the matches is at the level of 50%, while for the obvious problem related to exces-
DT results against the self-evaluation of 81% of the best DT matches, the referees sively large numbers of proposals, panel

6 The Messenger 177 – Quarter 3 | 2019

members report a general uneasiness P (helpful comment | DeepThought) Figure 5. Conditional
probability for the
when dealing with proposals in areas in

DeepThought inferred knowledge

­various combinations of
which they feel they are not experts. For self-reported and
a more quantitative assessment, DPR 0.14 0.24 0.39 0.24 DT-­inferred knowledge

participants were asked to express their

level of confidence, using a four-point
scale, when asked to evaluate those
cases; the corresponding distribution is 0.29 0.28 0.27 0.16

presented in Figure 6. In about 60% of the
cases, the reviewers were not comfortable

with this situation. This implies that better

matching of expertise gives the reviewers
a better experience, an aspect which 0.28 0.31 0.27 0.13

should not be underestimated.


1 2 3 4
Not helpful Very helpful
Feedback quality Review evaluation
In the classical review concept, the feed-
back provided by the panel to the PI is
supposed to reflect the consensus opin- 0.5
Figure 6. Distribution of
the answers to the
ion. This paradigm has at least two obvi-
question: “How satisfac-
ous limitations: (a) proposals that are tri- torily were you able to
aged out (i.e., the bottom ~ 30%) are not 0.4 evaluate the proposals
discussed, and the feedback is based for which you were not
an expert?”.

on the opinion of the primary referee; (b) 0.3

for proposals that are discussed during
the face-to-face meeting the primary 0.2
­referee tries to capture the main points of
the discussion and produces a single
comment. There is simply not enough 0.1
time for the panel members to review all
the feedback and to make sure it reflects 0.0
provided an unfair evaluation

all the aspects of the discussion. In the

might not always have been

Mostly; I sometimes missed

Somewhat; I struggled and

Fully; I could evaluate well

and fairly as a non-expert

current implementation at ESO, the com-

the expertise but was still
Not satisfactory; I might

ments are formally supervised by panel

chairs, who are responsible for the integ-
have unintentionally

rity of the feedback (particularly as it

able to evaluate

able to evaluate

relates to the language used). The net

effect, possibly coupled with a sub-­
optimal matching between proposal and
referee, is a high level of dissatisfaction
in the community, which is consistently
reported by the Users Committee; the
dissatisfaction reported is about 30% for The participants were asked to rate each (99% of the sub-sample that responded).
all of ESO and exceeds 50% for ALMA 3. of the comments they received for their In about 40% of the cases the DPR was
proposal, based on its helpfulness. It is reported to have provided better com-
Since the TAWG recommended the use important to stress that they were not ments, while the fraction of comments
of DPR for a FTC, no attempt was made asked whether the comments were good with quality similar to, or better than the
to produce consensus feedback and/or or bad, or whether they liked them or not, OPC reaches about 85%.
to edit/check individual comments, which but whether they were useful for improv-
were distributed to the PIs in their original ing the quality of their proposal. The gen- The analysis of comment helpfulness as
form. The purpose of this implementation eral response was very satisfactory, as a function of the reviewer’s expertise
was two-fold: (a) to get feedback on the shown in Figure 7, with more than 60% (either self-reported or DT-inferred) shows
concept itself, and (b) to detect possible of the comments judged as being useful, that the dependence is mild in the central
problems (for example, inappropriate lan- and about 5% not useful. One of the regions; the experts very rarely gave
guage) generated by the unedited/unfil- questions also concerned the compari- unhelpful comments and, conversely,
tered text. son with the edited OPC comments non-experts rarely gave very helpful com-
received by the PIs in previous semesters ments. A similar analysis as a function of

The Messenger 177 – Quarter 3 | 2019 7

Telescopes and Instrumentation Patat F. et al., The Distributed Peer Review Experiment

the reviewer’s scientific seniority reveals a Figure 7. Distribution

of the “helpfulness”
flat distribution (within the noise), with one
0.4 ­ratings of the referee
remarkable exception: graduate students comments for the entire
seem to be unable to provide very useful DPR sample.
comments. This may signal a training 0.3
issue, which can probably be addressed Fraction
by exposing the students to schemes like 0.2
the DPR. Finally, no statistically significant
difference is seen between the helpful- 0.1
ness of comments written by female and
male referees. 0.0

Somewhat; some comments

might help me to strengthen

Fully; overall the comments

will not help me to improve

Mostly; several comments

Not useful; the comments

will help me to strengthen

will allow me to improve

A brief primer on subjectivity

my proposed project

my proposed project

my proposed project
my proposed project

Figure 8 (below). Pre-

Before we proceed with the comparison meeting OPC referee–­
between the final OPC and DPR out- referee correlation. In
comes, a digression on the subjectivity this density diagram
each point represents a
inherent in the process is necessary. pair of grades attributed
Although it is common knowledge that to the same proposal
two different panels reviewing the same by two distinct referees.
set of proposals would provide different The data are from the
P18 sample.
rankings (and this is often used to compare
time allocation committees to roulette), 4.0
quantitative statements are very rare. This
matter is addressed in great detail in an N(data) = 196153 480
extensive study based on about 15 000
ESO proposals (Patat, 2018b; hereafter 3.5
P18). The interested reader is referred to 420
the paper for a thorough discussion,
while here we will focus only on the con-
cepts relevant to the present discussion. 3.0 360
Referee grade #2

One way of quantitatively describing the

reproducibility of a review process is 300
the correlation between the grades attrib-
uted to the same set of applications by
two distinct bodies. These bodies can be 240
composed of a single individual or of sev-
eral members. We will be talking about
referee–referee (r–r) and panel–panel 2.0 180
(p–p) correlations. In the first instance,
one simply considers all the distinct
grade pairs attributed by referee #1 and
referee #2 to the same set of proposals, 1.5
placing them in a diagram in which the 60
grades are used as coordinates, so that
each single grade pair is represented Corr. coeff. = 0.21
by a point. One can then repeat the pro- 1.0
cess for all possible referee pairs, plotting 1.0 1.5 2.0 2.5 3.0 3.5 4.0
all the corresponding points on the r–r Referee grade #1
plane. Since the same proposal is graded
by many reviewers, each single proposal proposal is np = Nr (Nr –1)/2. For instance, with Np = 172 clouds of points. In the
is represented on the r–r plane by a cloud in the case of the DPR experiment, with case of the DPR experiment, this would
of points. typically Nr = 7, the above combinatorics yield 172 × 21 = 3612 points. In an ideal
formula yields 21 distinct pairs per pro- situation, all the clouds would be very
In the simplifying assumption that each posal. Of course, the same operation can small in size (meaning that all referees
proposal is seen by Nr referees, the num- be repeated for all Np proposals in the would provide very similar grades for the
ber of distinct grade pairs np for each sample, which will populate the diagram given proposal), and so the points would

8 The Messenger 177 – Quarter 3 | 2019

be distributed very close to the straight- is 1, while a null value would signal a Comparing the OPC and DPR
line y = x on the r–r plane. complete disagreement. The average outcomes
agreement is expected to be 0.25 in case
To illustrate what one is to expect in real a fully stochastic process, i.e., when there The first test we apply to the DPR data
life, we have constructed the r–r plane for is no correlation between the two bodies. concerns the subjectivity level character-
the pre-meeting OPC P18 sample, from The concept can be extended to all ising the typical participant. For this pur-
which we derived almost 200 000 grade ­quartiles, including cross-quartile values, pose, we have computed the average
pairs accumulated over 16 ESO cycles. and the quartile agreement matrix (QAM) r–r QAM that we introduced in the previ-
The resulting diagram is presented in Fig- can be constructed. In statistical terms, ous section. Because of the DPR setup,
ure 8. It is important to note that for a the generic element Mij of the QAM is the the ranking list for each referee includes
perfectly stochastic process, the points conditional probability that a proposal at most eight proposals, so each quartile
would be distributed within a circular area, ranked in the i-th quartile by referee #1 is contains no more than two proposals.
with some radial, typically Gaussian, dis- ranked in the j-th quartile by referee #2. Also, at variance with the classical panel
tribution. The fact that the real d­ istribution scheme, the number of proposals in
is elongated along the diagonal direction The application of this concept to the common between two reviewers is typi-
signals that the process is not aleatory. P18 pre-meeting sample shows that, on cally very small. As a direct comparison
This qualitative conclusion can be made average, the ranking lists produced by between ranks is not possible, we use
more quantitative by computing the Pear- two distinct referees have about 33% of a bootstrap approach. Very briefly, for
son linear correlation coefficient, which the proposals in common in their first and each of the 172 proposals we randomly
ranges from –1 (complete anti-correla- last quartiles. In the central quartiles the extract one grade pair and form two
tion) to 1 (complete correlation) and is null intersection is compatible with a purely ranking lists, which are used to compute
for complete uncor­relation. The value random selection (25%). This extends to the quartile agreement fractions. The
derived for the sample is 0.21. Given the the mixed cases (i ≠ j ), with the exception ­process is repeated a large number of
very large number of points, this is a very of the extreme quartiles; the fraction times and the average values and stand-
robust estimate which can be reliably of proposals ranked in the first quartile by ard deviations are derived for each of the
taken as a low correlation. For the same referee #1 and in the fourth quartile QAM elements. The result is presented
reason, however, this value reveals that by referee #2 is ∼ 17%, which deviates in in Table 1. A direct comparison with
there is a statistically significant signal a statistically significant way from the the values derived from the P18 sample
indicating that the process is not com- ­random value. As in the case of the r–r reveals that the two results are statisti-
pletely aleatory. If on the one hand this correlation introduced above, the r–r cally indistinguishable. No meaningful
may sound discouraging, it helps to put agreement fraction gives a quantitative ­difference is seen in the QAMs computed
things in the correct context, as it char- estimate of the high level of subjectivity for the OE and DT sub-samples.
acterises the subjectivity of the process that characterises the process, provid­ing
in a more quantitative and objective way, a precise indication of what one should In a further test, we have investigated the
as opposed to the common statements expect. possible dependence on the scientific
which are normally based on pure anec- seniority level introduced above. Of the
dotal evidence. The reason why the applications are usu- 167 reviewers, 136 provided this informa-
ally evaluated by more than one reviewer tion, which we used to sub-divide the
A different way of measuring the repeata- is to reduce the inherent “noise” which, reviewers into two classes: junior (groups
bility of the process, which we will use as we have just seen, is quite substantial. 0 and 1) and senior (groups 2 and 3).
extensively in the next section, is the For this purpose, the grades attributed These classes roughly correspond to
quartile agreement fraction (P18). The by different referees to the same proposal PhD students plus junior postdocs (37),
concept is as follows. When the same set (typically grouped in panels) are aggre- and advanced postdocs plus senior sci-
of proposals is reviewed by two different gated to form one single figure of merit. In entists (99), respectively. We then com-
bodies #1 and #2, one can compile the ESO implementation (and this is a puted the r–r QAM for the two classes;
the rankings for the two distinct reviews common recipe), this is achieved simply the first quartile terms are 0.22 and 0.32,
based on their distinct grades. The taking the average, with no weights and/ respectively. At face value this indicates
­rankings are then used to derive a merit or rejection. The effect of increasing the a larger agreement between senior
classification within the classical quartile number of reviews is diffusely discussed reviewers. However, the small size of the
scheme. For instance, the top 25% of in P18; here it suffices to say that for
proposals are ranked in the first quartile Nr = 3 the first quartile agreement fraction
of the distribution of grades. grows to 43% and 30% in the first and Table 1. Bootstrapped r–r Quartile Agreement Matrix
for the DPR experiment.
second quartiles, respectively.
Once this is done, one can compute the Referee #1 Referee #2 quartile
fraction of applications ranked in the Armed with these terms of reference we quartile 1 2 3 4
first quartile by review #1 which are also can now discuss the results of the DPR 1 0.33 0.26 0.24 0.18
graded in the same quartile by review #2. experiment. 2 0.26 0.26 0.25 0.23
For a complete agreement the fraction 3 0.24 0.25 0.25 0.26
4 0.18 0.23 0.26 0.34

The Messenger 177 – Quarter 3 | 2019 9

Telescopes and Instrumentation Patat F. et al., The Distributed Peer Review Experiment

junior class produces a significant scatter, Table 2. Average DPR–OPC (pre-meeting) Table 3. DPR–OPC (pre-meeting)
r–r Quartile Agreement Matrix. p–p Quartile Agreement Matrix.
so the difference may not be significant.
DPR referee OPC referee quartile DPR OPC (pre-meeting) quartile
One can extend the above bootstrapping quartile 1 2 3 4 quartile 1 2 3 4
procedure to subsets with a number of 1 0.31 0.26 0.24 0.18 1 0.37 0.26 0.28 0.09
referees Nr > 1. The case of Nr = 3 is par- 2 0.24 0.27 0.25 0.24 2 0.28 0.16 0.28 0.28
ticularly interesting as this is directly 3 0.24 0.23 0.26 0.26 3 0.16 0.40 0.19 0.26
­comparable to the results presented in 4 0.20 0.23 0.25 0.31 4 0.19 0.19 0.26 0.37
P18. The procedure is as follows: we
first make a selection of the proposals This matrix is very similar to that derived of the pre-meeting OPC process (P18).
having at least 6 reviews (164); for each of within the DPR reviews (see Table 1), pos- Note that, given the large noise inherent
these we randomly select two distinct sibly indicating a DPR–OPC r–r agree- in the process, a much larger data set
(i.e., non-intersecting) subsets of Nr = 3 ment slightly lower than the correspond- (or more realisations of the experiment)
grades each, from which two average ing DPR–DPR. A check performed on would be required to reach a sufficiently
grades are derived; the subsequent steps the two sub-samples for the junior and high statistical significance and to make
are identical to the r–r procedure, and senior DPR reviewers (according to the robust claims about possible systematic
lead to what we will call the p–p QAM. classification described above) has given deviations.
statistically indistinguishable results.
The first-quartile agreement turns out to The fact that in the real OPC process
be 41%, while for the second and third As explained in the introduction, the pro- there is a face-to-face meeting consti-
quartiles this is 30%. The top-bottom posals were reviewed by Nr = 3 OPC tutes the most pronounced difference
quartile agreement is 10%. These values ­referees in the pre-meeting phase. This between the two review schemes. In the
are very similar to those presented in constitutes a significant difference, in meeting, the opinions of single reviewers
P18 for the OPC process for Nr = 3 sub-­ that the DPR ranking is typically based are changed during the discussion, so
panels. As for the r–r case, the OE and on ~ 7 grades, whereas the pre-meeting that grades assigned by individual refer-
DT sub-samples yield statistically indistin- OPC ranking rests on 3 grades only. ees are not completely independent ­
guishable values. The conclusion is that, With this caveat in mind, one can never- from each other (as opposed to in the
in terms of self-consistency, the DPR theless compute the QAM for the two pre-meeting phase, in which any signifi-
review behaves in the same way as the overall ranking lists. The result is pre- cant correlation should depend only on
pre-meeting OPC process. sented in Table 3. At face value, about the intrinsic merits of the proposal). The
37% of the proposals ranked in the 1st effects of the meeting can be quantified
We now come to what is perhaps one of quartile by the DPR were ranked in the in terms of the quartile agreement frac-
the most interesting aspects. As antici- same quartile by the OPC, with a similar tions between the pre- and post-meeting
pated, the proposals used in the DPR fraction for the bottom quartile. When outcomes, as outlined in Patat (in prepa-
experiment were also subject to the regu- looking at these values, one needs to ration; hereafter called P19). Based on
lar OPC review. This enables the com­ consider that this is only one realisation, the P18 sample, P19 concludes that the
parison between the outcomes of the two which is affected by large scatter, as change is significant; on average, only
selections, with the caveats outlined can be deduced from the comparatively 75% of the proposals ranked in the top
above about their inherent differences. large fluctuations in the QAM. These are quartile before the meeting remain in the
evident when compared to, for instance, top quartile after the discussion (about
For a first test we used a bootstrap the average values obtained from the 20% are demoted to the second quartile,
procedure in which, for each proposal bootstrapping procedures described and 5% to the third quartile). P19 charac-
included in the DPR, we randomly above. The numerical simulations show terises this effect by introducing the
extracted one evaluation from the DPR that the standard deviation of a single Quartile Migration Matrix (QMM). For the
(typically one out of 7) and one from realisation is ~ 0.1. specific case of Period 103, the QMM
the OPC (one out of 3), forming two is reported in Table 4 for the subset of the
ranking lists from which a r–r QAM was Using the model presented in P18, one DPR experiment. Of the initial 172 pro-
­computed. The operation was repeated can predict that, on average, the top and posals included in the DPR sample, 36
a large number of times and the average bottom quartile agreement between the were triaged out in the OPC process and
and standard deviation matrices were DPR and the pre-meeting OPC should be are therefore not considered.
constructed. This approach provides a around 0.5 (see Kerzendorf, 2019 for
direct indication of the DPR-OPC agree- more detail). The observed value (0.37) As anticipated, the effect is very marked;
ment at the r–r level and overcomes differs at the 1.3-s level from the average the meeting does have a strong effect
the problem that the two reviews have value. For the central quartiles the differ- on the final outcome. In light of these
a ­different number of evaluations per ence is at the ~ 1.5-s level. Therefore, facts, we can finally inspect the QAM
­proposal (see below). The result is pre- although lower than expected on aver- between the DPR and the final outcome
sented in Table 2. The typical standard age, the observed DPR–OPC agreement of the OPC process. This is presented
deviation of single realisations from the is statistically consistent with that in Table 5. With the only possible excep-
average is 0.06. expected from the statistical description tion of M4, 4, which indicates a relatively

10 The Messenger 177 – Quarter 3 | 2019

Table 4. OPC Quartile Migration Matrix for the Table 5. DPR–OPC (post-meeting) processing. The next logical step is to
DPR sub-sample (N = 136). Quartile Agreement Fraction.
expand this experiment and distribute a
OPC pre-meeting OPC post-meeting quartile DPR OPC post-meeting quartile fraction of observing time using DPR at
quartile 1 2 3 4 quartile 1 2 3 4 more facilities. More than 95% of the
1 0.56 0.32 0.12 0.00 1 0.26 0.38 0.24 0.12 ­participants suggest an implementation
2 0.32 0.32 0.29 0.06 2 0.24 0.35 0.24 0.18 of such a scheme for some part of the
3 0.12 0.26 0.38 0.24 3 0.32 0.12 0.29 0.26 ESO proposal types, with 75% support
4 0.00 0.09 0.21 0.71 4 0.19 0.15 0.24 0.44 for the short programmes (time requests
< 20 hours). Fewer than 5% of the
marked agreement for the proposals in weakest aspect of the DPR. However, responses were against implementing
the bottom quartile, the two reviews it remains unclear whether panel discus- DPR for any of the programme types. In
appear to be almost completely uncorre- sions lead to the selection of better particular, about 70% of the responses
lated. By means of simple Monte-Carlo ­science. In this respect, it is important to are in favour of deploying DPR for the
calculations one can show that for two note that several studies have shown that Fast Track Channel, while only about 15%
fully aleatory panels, the standard devia- panel meetings can increase the differ- are against it (the remaining 15% is indif-
tion of a single realisation around the ences between two panels with respect ferent). We take this as a clear indication
average value (0.25) is 0.10. We conclude to the pre-meeting agreement. In other of support.
the majority of the Mi,j elements in Table 5 words, while the meeting increases the
are consistent with a stochastic process internal consensus by polarising different One of the objections that is typically
at the 1-s level. opinions within the panels, it does not made to the DPR concept is that, by dis-
lead to a better panel-panel agreement tributing the proposals to a larger number
The main conclusion of this analysis is (see Obrecht et al., 2007 and references of unselected scientists, it increases the
that, while the pre-meeting agreement therein). One would expect the discus- chances of information leakage and pla-
is consistent, with the DPR and OPC sions to bring judgment closer to identify- giarism. In the specific case of the DPR
reviewers behaving in a very similar way ing the best science; however, these experiment, the proposals were distrib-
(in terms of r–r and p–p agreements), studies indicate that a face-to-face meet- uted to 172 reviewers, while in the OPC
the face-to-face meeting has the effect of ing does not necessarily make the pro- process the applications were seen by 78
significantly increasing the discrepancy cess better. individuals. However, while in the OPC
between the two processes. However, implementation each reviewer has access
we caution that the sample is relatively to all proposals assigned within her/his
small, and therefore the results are signifi- Conclusions and outlook panel (typically 70–80), the DPR reviewer
cantly affected by noise. sees a factor of ~ 10 fewer proposals.
Gemini has already implemented a Therefore, under the reasonable hypo­
That the DPR–OPC agreement is smaller ­variant of this mechanism successfully thesis that the fraction of “malevolent”
than the internal DPR–DPR agreement over the past few years for their Fast ­scientists is the same in both review bod-
is not unexpected, as there are intrinsic Turnaround (Andersen et al., 2019). The ies (which are selected from the same
differences between the two setups, the approach presented here enhances this community), one would actually expect
largest one being the absence of a face- process, using better review-­proposal that the DPR is less prone to confidential-
to-face meeting, which is potentially the matching based on natural language ity issues on average. To get a direct
opinion from DPR participants, the ques-
Figure 9. Distribution of tionnaire contained an explicit question
the answers to the
about this aspect. The distribution of the
question: “For which
types of proposals responses is shown in Figure 10. Exclud-
0.25 do you think distributed ing the “no strong opinion” cases, 66%
peer review would be of the users declared themselves to be
beneficial?” in the DPR
0.20 equally or more confident in the DPR­
­process, resulting in about a third of the

users placing more trust in the classical


0.10 Another concern that is often heard when

discussing DPR is the possible presence
of biases. Again, the specific question
put to the participants regarding this
point does not support this concern; 74%
0.0 of the respondents believe DPR is equally
Short Regular Large Short, Regular, Short, None
or more robust against biases (Figure 11).
regular large regular,

The Messenger 177 – Quarter 3 | 2019 11

Telescopes and Instrumentation Patat F. et al., The Distributed Peer Review Experiment

Figure 10. Distribution gives an objective criterion to assign a

of answers to a question
particular expertise, eliminating biases in
0.4 about how secure the
participants felt about self-reporting. DPR implicitly removes
the concept of panel, which adds rigidity

confidentiality issues.
to the process. For instance, it maximises
0.2 the overlap in evaluations, which is a
­t ypical issue in pre-allocated panels. The
lack of a face-to-face meeting prevents
0.0 strong personal opinions from having a
pivotal influence on the process. Also,
concerned about confidentiality

DPR involves a larger part of the commu-

I am more concerned about
I am less concerned about

confidentiality issues in the

confidentiality issues in the

issues in the DPR process

nity, increasing its democratic breadth

I am neither more nor less

than in the OPC process

DPR process than in the

DPR process than in the

I have no strong opinion

and exposing all applicants to the typical

quality of the proposals. This allows them
to better understand if their request is
not allocated time by placing it in a wider
OPC process
OPC process

on this point

context, which will help to improve their

proposal-writing skills, training the mem-
bers of the community without additional

We acknowledge that the lack of a meet-

Figure 11. Distribution of ing does not allow the exchange of
answers to a question ­opinions and the possibility of asking and
about the robustness of
the process against answering questions to/from the peers.
0.2 Despite the fact that its effectiveness

remains to be demonstrated and quanti-
fied (see above), it is clear that the social,
0.1 educational and networking aspects
of the face-to-face meeting should not
0.0 be undervalued. In this respect, we note
that the resources freed by the DPR
I have no strong opinion
against biases than the

against biases than the

process is more robust
I think that DPR review

I think that DPR review

I think that DPR review

process is less robust

approach can be used by the organisa-

against biases as the
OPC review process

OPC review process

OPC review process

process is as robust

tions for education and community

­networking (training on proposal writing,
fostering collaborations, etc.).
on this point

In April and May 2019, results of the

DPR experiment were presented to the
ESO governing bodies most closely
­concerned with the Peer Review process
The main conclusions drawn from the To these aspects, which come directly (i.e., the Scientific Technical Committee,
DPR experiment can be summarised as from the data, other positive facts can the Users Committee and the Observing
follows: be added. DPR allows a much larger sta- Programmes Committee). The ensuing
– The DeepThought-enhanced DPR tistical basis enabling robust outlier rejec- discussions have resulted in a wealth of
experiment was very well received by tion (the number of proposals per referee useful feedback that is being discussed
the participants. can be easily brought to 10–12) and it internally. We would like to conclude
– The mechanism allows an optimal removes possible biases generated by by pointing out that these kinds of stud-
referee-proposal matching. panel member nominations. The larger ies are crucial if we are to progress from
– The DPR process is as subjective as pool of scientists allows much better a situation in which the classical peer
the OPC process. ­coverage in terms of proposal expertise review process is adopted notwithstand-
– The participants do not see the confi- matching, and the smaller number of ing its limitations simply due to the lack of
dentiality and bias issues as being more ­proposals per reviewer allows more care- better alternatives. As scientists, we firmly
severe than in the classical scheme. ful work and more useful feedback. believe in experiments, including those
– ESO should consider deploying DPR for that address the selection of the experi-
regular proposals below a certain Another aspect of the DeepThought ments themselves.
time request, while leaving the classical approach to proposal-referee matching is
review for larger time requests. that it can be semi-automated; it also

12 The Messenger 177 – Quarter 3 | 2019

Acknowledgements  istributed Peer Review Pilot in Foundational
D Kerzendorf, W. E. et al. 2019, submitted to Nature
­Program: Astronomy
The authors wish to express their gratitude to the distributed-­p eer-review-pilot-foundational-program Merrifield, M. R. & Saari, D. G. 2009, Astronomy and
167 volunteers who participated in the DPR experi- Report from ESO Users Committee No. 42 (2018): Geophysics, 50, 4.16
ment, for their work and enthusiasm. The authors are Mervis, J. 2014a, Science, 344, 1328
also grateful to Markus Kissler-Patig for passionately tees/uc/uc-42nd/UCreport2018.pdf Mervis, J. 2014b, Science, 345, 248
promoting the DPR experiment following his experi- Obrecht, M., Tibelius, K. & D’Aloisio, G. 2007,
ence at Gemini; to ESO’s Director General Xavier Research Evaluation, 16 (2), 79
Barcons and ESO’s Director for Science Rob Ivison References Patat, F. 2016, The Messenger, 165, 2
for their support; and to Hinrich Schütze for several Patat, F. et al. 2017, The Messenger, 169, 5
suggestions on the NLP process. Andersen, M. et al. 2019, AAS, 233, 455.03 Patat, F. 2018a, The Messenger, 173, 7
Ardabili, P. N. & Liu, M. 2013, CoRR, arxiv:1307.6528 Patat, F. 2018b, PASP, 130, 084501
Brinks, E. et al. 2012, The Messenger, 150, 20 Strolger, L.-G. et al. 2017, AJ, 153, 181
Links Gallo, S. A., Sullivan, J. H. & Glisson, S. R. 2016,
PLoS ONE, 11, e0165147
 emini Observatory Fast Turnaround Observing
G Huang, C. 2013, European Journal of Psychology of
Mode webpage: Education, 28, 1
observing-gemini/proposal-routes-and-observing- Kerzendorf, W. E. 2017, Journal of Astrophysics and
modes/fast-turnaround Astronomy, arxiv:1705.05840
ESO/G. Hüdepohl (

Snowfall at Paranal is
a rare phenomenon that
serves to utterly trans-
form the surroundings of
the VLT/I into an other-
worldly landscape.

The Messenger 177 – Quarter 3 | 2019 13

Telescopes and Instrumentation DOI: 10.18727/0722-6691/5148

On the Telluric Correction of KMOS Spectra

Lodovico Coccato 1 rate atmospheric and instrumental (MIPAS) atmospheric profiles for temper-
Wolfram Freudling 1 effects, (for example, the instrument ature, humidity, water vapour and other
Alain Smette 1 response) if a large wavelength range molecules, and (d) analytic functions
Eleonora Sani 1 of stellar continuum is absorbed by or user-provided files for the instrumental
Jose A. Escartin 1, 2 blended absorption lines. Last but not spectral resolution. The fit to the telluric
Yves Jung1 least, the noise and imperfections in absorption lines in the observed spectra
Gurvan Bazin1 the data reduction of these stars are inev- provides the integrated column density
itably propagated to scientific spectra. of individual molecules. Future versions
will further improve the quality of the
ESO Alternatively, one can model the atmos- model by including real-time measure-
 ax-Planck-Institut für extraterrestrische
M phere, generate its transmission spec- ment of precipitable water vapour and
Physik, Garching, Germany trum and apply it to observations. The other molecules along the line of sight of
model itself can be obtained by fitting the exposures.
well-defined telluric lines to the spectrum
The presence of strong absorption of either a standard star or a sufficiently In the following, we describe the improve-
lines in the atmospheric transmission bright science target. In general, a model ments in the quality of KMOS (Sharples
spectrum affects spectroscopic obser- depends on four components: (a) a radia- et al., 2013) spectra obtained with
vations, in particular those in the near- tive transfer model; (b) a set of parame- the model approach using molecfit with
and mid-infrared. Therefore, there is the ters that determines the absorption and respect to the empirical method. Data
need to correct scientific observations transmission properties of individual were reduced using the KMOS pipeline
for this effect, a process known as tel- ­molecules; (c) atmospheric profiles of (Davies et al., 2013). In the model
luric correction. The use of a detailed temperature, humidity, and volume mix- approach, the atmospheric model was
model of the atmospheric transmission ing ratio for the molecules involved; and obtained by fitting a number of pre-­
spectrum brings several advantages (d) instrumental parameters such as defined telluric lines on a standard star
over the method of empirically deriving spectral resolution. This model-depend- spectrum observed close in time to
corrections using observations of a ent approach has several advantages the scientific data (i.e., the same standard
­telluric standard star. In this paper, we over the empirical method. First, no addi- star that was used in the empirical
discuss and compare the two methods tional noise or sources of error coming method). The telluric correction over
applied to K-band Multi-Object Spec- from the standard star observations and the full wavelength range was then com-
trograph (KMOS) observations and reduction are propagated to the science puted accounting for the differences in
show the improvements in the quality of spectra. Second, it allows additional airmass and spectral resolution between
the final products obtained by imple- components to be taken into account, the s­ cientific spectrum to correct and
menting the modelling technique such as the amount of precipitable water the standard star. As a test-bench for
offered by the ESO molecfit sky tool. vapour from external sources and inac- comparison, we processed one month of
curate wavelength calibrations, and dif- KMOS data and compared the results
ferences between the observations of the obtained with these two different telluric
Correction for atmospheric transmission standard star and the science target (for correction strategies.
in spectroscopic data example, airmass and spectral resolu-
tion). On the other hand, using a model of
Ground-based spectroscopic observa- the atmosphere for the telluric correction Benefits of the molecfit strategy for
tions are strongly affected by the Earth’s risks the introduction of systematics KMOS observations
atmosphere. In particular, spectra of because of limitations in the modelling. In
objects taken in the near- and mid-infra- practice, the artefacts caused by such As described previously, because the
red wavelength ranges are characterised systematics are outweighed by the molecfit correction is based on a model,
by a forest of absorption lines, called improvements made in the corrections. it does not add noise to the final products
­telluric absorptions. These features are or defects such as uncorrected cosmic
caused by (mainly water and OH) mole- The model approach has been devel- rays that are embedded in the standard
cules present in the atmosphere that oped in a software package named star spectrum. Figure 1 shows a compari-
absorb the light from astrophysical molecfit (Kausch et al., 2013; Smette et son between the mean signal-to-noise
sources. The standard way to correct for al., 2015). Molecfit uses (a) the Line-by- per pixel of the datacubes obtained by
this effect is to acquire a spectrum of line Radiative Transfer Model 1 (LBLRTM) correcting the telluric absorption directly
a bright and featureless star close in time algorithm (Clough, Iacono & Moncet, with a standard star (i.e., the empirical
and airmass to the scientific target, and 2005) to compute the radiative transfer method) and by modelling the atmos-
compare it either with its model or, if model, (b) the high-resolution transmis- pheric absorptions with molecfit. The
available, with a spectrum taken from sion molecular absorption (HITRAN) ­signal-to-noise is measured in a wave-
space. This empirical strategy, however, database 2 for the molecular parameters, length region that is free of sky or telluric
has some drawbacks. First, it requires (c) Global Data Assimilation System 3 lines, and therefore is an indication of
additional (expensive) telescope time. (GDAS) and ESA Michelson Interferome- the noise added by the telluric correction.
Second it can be complicated to sepa- ter for Passive Atmospheric Sounding 4 As expected, the data corrected with

14 The Messenger 177 – Quarter 3 | 2019


2-1 23#  

2-1 23#


 2-1 23#


2-1 23 1  


molecfit contain less noise; the effect is ments for the observation of the telluric Figure 1. Left panel: Comparison between the
signal-­to-noise ratio (per pixel) of the final KMOS
much more visible for bright objects standard stars are relaxed. The conse-
datacubes obtained with the empirical corrections
which have signal-to-noise ratios close quence is that some telluric lines are and the method with molecfit on-standard model
to those of the telluric standard stars. over- or under-corrected by up to or over for the telluric correction. The colour of the symbols
No major improvement is expected for 10%, because the column density of the is proportional to the signal-to-noise of the telluric
star used in the data reduction; predicted trends for
objects with signal-to-noise < 50, because molecules, and therefore the atmos­pheric
several values of the signal-to-noise of the telluric
the noise in the telluric standard is negli- transmission, is linked to the airmass. standard are shown. The dashed black line shows
gible with respect to the total noise in The empirical approach offered by the the 1:1 relation. Right panel: Example of spectra
the data. Even with the model approach KMOS pipeline does not account for ­c orrected with the empirical corrections (in red) and
with the model approach with molecfit (in black)
there are systematic artefacts in the such airmass differences, whereas the
for the dataset KMOS.2018-10-23T07:35:07.185.
reduced spectrum that are due to residu- model approach does. Figure 2 shows
als in the sky subtraction or cosmic the change in intensity in the telluric line
ray cleaning. However, with the empirical at 1.27 µm for a difference in airmass
approach the number of artefacts is of Dz = 0.34, and its effects on the cor- with another arm. This issue can be taken
higher; these additional artefacts are not rected science spectrum; accounting into account in the model approach, by
due to the sky subtraction in the science for this difference overestimates the including a set of static calibrations that
spectra (in fact, the same procedure as absorption feature at 1.27 µm by ~ 10%. reproduce the wavelength dependency of
in the model approach is used). Some of The bottom panel of Figure 2 shows a the instrumental spectral resolution for
these are inherited from imperfections systematic artefact at 1.27 µm, which each arm and instrument configuration.
in the reduction of the standard star while is due to there being no correction for the These calibrations are included in the
others are due to the limitations of the difference in airmass between science KMOS pipeline distribution and allow one
empirical method in dealing with differ- and standard star observations. to compute the telluric correction for
ences between the observations of the each arm with the exact shape of the
science and the telluric star (see Figures 2 Another limitation of the empirical method absorption lines, regardless of which arm
and 3). is that taking a telluric calibration in each the telluric standard was observed with.
of the 24 arms is time consuming. There- Figure 3 compares the effects of taking
One of the limitations of the empirical tel- fore, the large majority of programmes and not taking into account the spectral
luric correction method is that the telluric observe a telluric standard star in only 3 resolution during the modelling in the
star and the science target are observed out of the 24 arms available in KMOS, i.e., final KMOS products. The shape of the
at different airmasses. The KMOS night one per detector. However, the spectral absorption features can differ by 10% or
calibration plan is designed to minimise resolutions of the various arms are differ- more, leading to artificial features in the
such differences; for example, a telluric ent, therefore the absorption features in final corrected spectrum.
standard is observed every two hours atmospheric transmission will have differ-
at airmasses close to the scientific tar- ent shapes in different arms. This means
gets observed during the night. However, that the observed shape of the absorp- Molecfit implementation in the KMOS
­differences up to ~ 0.4 in airmass are tion lines in telluric correction determined instrument pipeline
unavoidable owing to observational con- for a star observed with one arm does
straints, in particular during observations not fully match the shape of the telluric The original molecfit software interface
in visitor mode during which the require- lines of a scientific spectrum obtained does not support the use of KMOS data

The Messenger 177 – Quarter 3 | 2019 15

Telescopes and Instrumentation Coccato L. et al., On the Telluric Correction of KMOS Spectra

Atmospheric transmission Atmospheric transmission

Ratio Ratio


0.9 0.8

0.8 0.6

Not considering airmass difference 0.2 Not using appropriate spectral resolution
0.6 Considering airmass difference Using appropriate spectral resolution

1.25 1.26 1.27 1.28 2.36 2.38 2.40 2.42 2.44

Wavelength (µm) Wavelength (µm)

Science spectrum Science spectrum

1.2 1.2

1.1 1.1
Normalised flux
Normalised flux

1.0 1.0

0.9 0.9
Not considering airmass difference Not using appropriate spectral resolution
Considering airmass difference Using appropriate spectral resolution

1.25 1.26 1.27 1.28 2.36 2.38 2.40 2.42 2.44

Wavelength (µm) Wavelength (µm)

Figure 2. Comparison between the outcome of dif- block (we define this as the molecfit ­ Figure 3. Comparison between the outcome of dif-
ferent molecfit models that account for (black) ferent molecfit models that account for (black) and
on-science approach).
and ignore (red) the airmass difference between the ignore (red) the differences in spectral resolution
standard star and the science spectra. The top panel between the arm used to observe the standard star
shows the atmospheric transmissions and their The empirical method is much faster, but and the arm used to observe the scientific target.
ratio (in green). The airmass difference between the generally does not return the best results. The top panel shows the atmospheric transmissions
two models is D z = 0.34. The bottom panel com- and their ratios (in green). The bottom panel com-
Nevertheless, it is useful for a quick look
pares the science spectra corrected with these dif- pares the telluric-corrected science spectra (arbitrar-
ferent transmissions. The dataset used here is at the data or in those cases where the ily shifted). The file used here is KMOS.2017-02-
KMOS.2016-12-21T03:18:57.095. atmospheric fit does not converge. The 13T05:45:03.492.
molecfit methods are computationally
slower but return the best results in the
directly, because of the complicated multi- vast majority of cases. Both the molecfit approaches. For each spectrum, the
extension structure that requires special on-standard and on-science approaches user can select a reference scientific
treatment. In order to provide a con­ model the atmosphere by fitting a num- exposure from which to obtain the
venient interface, the molecfit algorithms ber of telluric absorption f­eatures in a atmospheric transmission. It can be the
have been integrated into the KMOS ­reference spectrum. In the on-standard same target or another; for example, a
pipeline, which now offers three strate- approach, the default wavelength regions bright target can be used to compute the
gies to correct for telluric absorptions in of the recipe can safely be used, whereas correction for all other observations in
the observations: for the on-science approach it is advisa- the same observing block. The loop is
1. Use the telluric standard star spectrum ble to carefully adjust the fitting regions rerun for each input exposure accounting
directly to correct the science data, avoiding intrinsic features of the science for changes in airmass and spectral
i.e., the empirical method. spectrum. Then, the full telluric correction resolution.
2. U se the standard star spectrum as is obtained for the entire wavelength range
­reference to model the atmosphere and it accounts for the difference in air- All three correction strategies have been
and derive its transmission to correct mass between the reference and science integrated into the data reduction work-
science data on the same night spectra. A set of static calibration files flow that can be executed within the
(we call this the molecfit on-standard provide the recipes with tables giving EsoReflex data reduction environment
approach). the wavelength-dependent instrument (Freudling et al., 2013). The KMOS work-
3. Use one science spectrum as a refer- spectral resolution and instrument flow includes automatic organisation
ence to model the atmosphere and response for each integral field unit (IFU), of the data and interactive tools to visual-
derive its transmission to correct the grating, and instrument rotator angle. Fig- ise and control the telluric correction as
science spectrum itself or other sci- ure 4 illustrates the data reduction cas- well as other reduction steps (Figure 5).
ence data in the same observing cades due to the two different molecfit In particular, it includes a tool to select

16 The Messenger 177 – Quarter 3 | 2019

Figure 4. Schematic representation of
the two molecfit strategies imple-
mented in the KMOS EsoReflex work-
1DCTBSHNM flow. The default molecfit on-standard
NERS@MC@QC 1DRONMRD 1DCTBSHNM approach (on the left) obtains atmos-
RS@QNARDQU@SHNMR BTQUD NERBHDMBDC@S@ pheric parameters from a telluric
standard observation before comput-
ing the atmospheric transmission at
C@S@BTADR the airmass and spectral resolution of
$WSQ@BSRODBSQ@ the scientific exposure. In the molecfit
RODBSQ@ on-science approach (on the right),
the process is applied directly to the
science exposures.


OOKX Figure 5 (below). The EsoReflex
OOKXBNQQDBSHNM BNQQDBSHNM KMOS workflow. Each green box rep-
+NNONM resents a step in the data reduction
C@S@BTADR chain. Orange boxes identify interac-
tive components in which the user has
the opportunity to inspect the prod-
ucts of that specific step and re-run
3DKKTQHB BNQQDBSDCC@S@BTADR the c­ orresponding recipe with different
parameters. Interactive modules
that are specific to telluric correction
are marked by the red ellipse. User-
defined scripts can be plugged into
the workflow as well.









k@SR M











The Messenger 177 – Quarter 3 | 2019 17

Telescopes and Instrumentation Coccato L. et al., On the Telluric Correction of KMOS Spectra

wavelength regions and the reference by looking at different locations on Sharples, R. et al. 2013, The Messenger, 151, 21
Smette, A. et al. 2015, A&A, 576, 77
spectra to fit the atmosphere, which are the sky (despite closeness in time or air-
fundamental steps in the on-science mass). The on-science method, however,
approach. An EsoReflex tutorial5 (Coccato requires interactive selection of bright Links
et al., 2019) that guides the user through ­science spectra to use as references,
Atmospheric & Environmental Research (AER)
the data reduction and an updated and a careful selection of the wavelength
Radiative Transfer Working Group Website:
­pipeline manual are available at the ESO ranges to fit. For those reasons it is not
instrument pipelines webpage 6. used for the archive products. In future, 2
HITRAN database:
the molecfit tools will be integrated into all 3
National Centers for Environmental Information
Global Data Assimilation System (GDAS):
The on-standard molecfit approach is the near- and mid-infrared instrument­
used for the in-house ESO reduction pipelines and workflows to grant the user data/model-datasets/global-data-assimilation-
of KMOS observations for the ESO flexibility to perform telluric correction in system-gdas
archive; these reduced data products will the most efficient way. 4
ESA Michelson Interferometer for Passive Atmos-
pheric Sounding:
be available to the astronomical commu-
nity soon through the ESO archive sci- instruments/mipas
ence portal 7. Improvements with respect References 5
T he Reflex KMOS tutorial (Coccato et al. 2019) can
to this method can be obtained case Clough, S. A., Iacono, M. J. & Moncet, J.-L. 2005,
be downloaded from the following link: http://www.
by case with the on-science molecfit J. Geophys. Res., 97, 1576 6
V LT instrument pipelines:
approach. Indeed, this approach limits Coccato, L. et al. 2019, Reflex KMOS tutorial, software/pipelines/
the differences in the molecule column issue 6.0 7
ESO Science Archive Portal:
Davies, R. et al. 2013, A&A, 558, 56
densities between the scientific spectrum Freudling, W. et al. 2013, A&A, 559, 96
and the target spectrum that arise simply Kausch, W. et al. 2015, A&A, 576, 78
M. Cabral/ESO

The VLT/I at sunset.

18 The Messenger 177 – Quarter 3 | 2019

Telescopes and Instrumentation DOI: 10.18727/0722-6691/5149

Bringing the New Adaptive Optics Module for

Interferometry (NAOMI) into Operation

Frédéric Gonté 1 Enrico Marchetti 1 NAOMI was developed by a consortium

Jose Antonio Abad 1 Stewart Mclay 1 composed of IPAG and ESO. Its Provi-
Roberto Abuter 1 Anthony Meilland 6 sional Acceptance Chile review was
Emmanuel Aller Carpentier 1 Alexander Meister 1 held in April 2019. The NAOMI systems
Jaime Alonso 1 Antoine Mérand 1 that have been installed on the Auxiliary
Luigi Andofalto 1 Thibaut Moulin 2 Telescopes make the Very Large Tele-
Pablo Barriga 1 Luca Pasquini 1 scope Interferometer (VLTI) and its
Jean-Philippe Berger 2 Jérôme Paufique 1 instruments much less dependent on
Jean-Luc Beuzit 2 Isabelle Percheron 1 the atmospheric and dome ­seeing con-
Israel Blanchard 1 Lorenzo Pettazzi 1 ditions. NAOMI increases the interfer-
Henri Bonnet 1 Oliver Pfuhl 5 ometer’s operability and improves the
Guillaume Bourdarot 2 Duc Phan 1 performance of its instruments and,
Pierre Bourget 1 Andres Pino 1 very early on, was identified as being
Roland Brast 1 Werther Pirani 1 critical to the VLTI. In this a
­ rticle, we
Paul Bristow 1 Jutta Quentin 1 review the project, describe its princi-
Luis Caniguante 1 Andrew Rakich 1 ples and architecture, and offer a pre-
Susana Cerda 1 Andrés Ramirez 1 view of the improvements it brings to
Claudia Cid 1 Robert Ridings 1 VLTI instruments.
Alex Correa 1 Mario Riedel 1
Eric Cottalorda 2 Javier Reyes 1
Benjamin Courtney-Barrer 1 Sylvain Rochat 2 Context
Pascaline Darré 1 Juan Sanchez 1
Bernard Delabre 1 Gonsalo Santos Tomás 1 Adaptive optics were considered for
Alain Delboulbé 2 Christian Schmid 1 interferometric instruments even before
Roderick Dembet 1 Pavel Shchekaturov 4 non-interferometric instruments, as
Ronald Donaldson 1 Nicolas Schuhler 1 the measurement of high-quality interfer-
Reinhold Dorn 1 Matthias Seidel 1 ometric observables depends strongly
Jorge Dupeyron 1 Christian Soenke 1 on the wavefront quality. Therefore, the
Christophe Dupuy 1 Eric Stadler 2 implementation of the adaptive optics
Sebastian Egner 1 Christian Stephan 1 (AO) systems on the Auxiliary and Unit
Frank Eisenhauer 5 Marcos Suárez 1 Telescopes (ATs and UTs) of the VLTI
Lorena Faundez 1 Mirko Todorović 1 has been recommended ever since the
Enrico Fedrigo 1 Guillermo Valdes 1 launch of the VLTI project (Beckers,
Gerhard Fischer 1 Cristophe Verinaud 1 1990). Consequently, between 2003 and
Christoph Frank 1 Julien Woillez 1 2005 the UTs were equipped with the
Eloy Fuenteseca 1 Gérard Zins 1 ­visible Multi-Application Curvature Adap-
Philippe Gitton 1 Sebastian Zúñiga-Fernández 1, 7, 8 tive Optics systems (MACAO; Arsenault,
Thibaut Guerlet 1 2003); later, in 2016, the UT AO coverage
Sylvain Guieu 2 was extended into the infrared with the
Pablo Gutierrez 1 1
ESO Coudé Infrared Adaptive Optics system
Pierre Haguenauer 1 2
Institut de Planétologie et d’Astrophy- (CIAO; Scheithauer, 2016).
Andreas Haimerl 1 sique de Grenoble (IPAG), Université
Xavier Haubois 1 Grenoble Alpes, CNRS, France
Cédric Heritier 1 3
KRP Mechatec GmbH, Garching, The Auxiliary Telescopes
Stefan Huber 1 Germany
Norbert Hubin 1 4
Pactum LTD, London, UK Although the design concept for the ATs
Paul Jolley 1 5
Max Planck Institute for Extraterrestrial developed by the Institut de Radioastro­
Laurent Jocou 2 Physics, Garching, Germany nomie Millimétrique (IRAM; Von Der Lühe,
Jean-Paul Kirchbauer 1 6
L aboratoire Lagrange, Université Côte 1997) included an AO system, this was
Johann Kolb 1 d’Azur, Observatoire de la Côte d’Azur, omitted when construction started owing
Johan Kosmalski 1 CNRS, France to a lack of resources, both in terms of
Peter Krempl 3 7
Universidad de Valparaíso, Instituto de personnel and funds. Instead, a tip/tilt
Carlos La Fuente 1 Física y Astronomía (IFA), Chile corrector based on the System for Tip-
Jean-Baptiste Le Bouquin 2 8
Núcleo Milenio de Formación Planetaria tilt Removal with Avalanche Photodiodes
Miska Le Louarn 1 (NPF), Valparaíso, Chile (STRAP; Bonaccini, 1997) and a fast
Paul Lilley 1 steering mirror was studied and imple-
Bruno Lopez 6 mented (Koehler, 2002, 2004). The ATs,
Marcelo Lopez 1 depicted in Figure 1, each have a primary
Yves Magnard 2 mirror of 1.8-metre diameter. They are

The Messenger 177 – Quarter 3 | 2019 19

Telescopes and Instrumentation Gonté F. et al., Bringing NAOMI into Operation

used by the VLTI for around 75% of the

Peter Krempl
available nights.


NAOMI has completely replaced STRAP

and is now the only wavefront correction
system available on the ATs. It has three Corrective optics
and M6 cabinet
main components that are fully embed-
ded in the telescopes: a wavefront sen-
sor; a deformable mirror; and a real-time
computer. The wavefront sensor was
developed at ESO headquarters; it is a
Shack–Hartmann Sensor sensitive to visi-
ble wavelengths with a 4 × 4 lenslet array,
of which 12 sub-apertures are used. The
low number of sub-pupils was driven by
the requirement to maintain a high sensi-
tivity for the VLTI-AT array. The detector Signal cabinet
is an off the shelf EM-CCD camera iXon in the back
ultra 897 developed by ANDOR (UK). The
wavefront sensor is also integrated on a
field-tracking stage to pick up the correct
field. A neutral density filter wheel has
been integrated just above the wavefront
sensor to adjust the illumination in case
of high flux, and a notch filter has been
integrated into the wavefront sensor
optics to reject any contamination from
the lasers of the Adaptive Optics Facility Wavefront sensor
and cold box
(AOF; Arsenault, 2017) implemented on
UT4. The full wavefront sensor and its fil-
ter wheel are installed in the lower part
of the Relay Optics Structure. It receives
the visible part of the light after the M9
dichroic mirror which redirects the infra-
red component to the VLTI instruments.
a temperature-controlled calibration Figure 1. 3D cross-sectional view of an AT showing
the locations of the NAOMI components. The correc-
The corrective optics (Le Bouquin, 2018), bench was necessary at Paranal. IPAG
tive optics and its electronics are implemented at
developed by IPAG, consist of a deforma- finalised this bench, which is now used the level of the azimuth; the wavefront sensor and its
ble mirror (DM241), which has 17 actua- as a maintenance tool on Paranal. We electronics are implemented in the lower part of the
tors across its clear aperture, produced discovered on Paranal that the deforma- Relay Optics Structure. The real-time computer and
the remaining control electronics are implemented in
by ALPAO France, and which is integrated ble mirror had a more complex thermal
the signal cabinet.
in a motorised gimbal mount. It is installed dependence than previously understood
at the M6 pupil location in the coudé train. and that the mirror displayed unpredicta-
In addition to compensating for atmos- ble behaviour. It is therefore necessary project. It uses the same platform as for
pheric turbulence, it provides the chop- to calibrate the deformable mirror on the CIAO of the UTs, which facilitated its
ping capability required by MATISSE. The sky at each target acquisition. The result development in Garching and simplifies
DM241 has a clear aperture of 37.5 mm, of the characterisation of the mirrors the operation on Paranal. The hardware is
but NAOMI uses only the central part was presented at the Adaptive Optics implemented in the electronics rack of
over a 28-mm (11-actuator) diameter pupil for Extremely Large Telescopes (AO4ELT) the signal cabinet (see Figure 1). This unit
as seen in Figure 2 (right). conference in June 2019 1 (see makes the acquisition on the Shack–­
Haguenauer, 2019). Hartmann Sensor, analyses the wavefront
The deformable mirrors were character- aberration and the pupil derotation to
ised at room temperature at IPAG and The real-time computing unit is based on be corrected, calculates the correction
then in a thermal chamber between 5 and the ESO-standard SPARTA Light platform to be applied to the deformable mirror
20 degrees C at ESO headquarters. The (Suárez Valles, 2012) and has been cus- and then sends the command to the mir-
outcome convinced the project team that tomised by PACTUM Ltd for the NAOMI ror. Since the deformable mirror and the

20 The Messenger 177 – Quarter 3 | 2019

Figure 2. Real-time displays from the real-time com-
puter. The left picture shows the Shack–Hartmann
image. The four unused spots at the corners are
vignetted at more than 50% by the pupil (represented
by a white circle). The sub-apertures are represented
in yellow. The picture on the right shows the voltage
applied to the actuators of the deformable mirror, the
white circle indicating the pupil footprint.

the project team had the required skills.

The real-time computer software followed
the same approach with model-based
testing before its integration onto the test
bench and was delivered by PACTUM Ltd.

A prototype of the wavefront sensor was

first developed to validate the opto-­
mechanical concept as well as the inte-
ANDOR camera are commercial prod- of actuators. IPAG provided phase maps gration and alignment procedure. The
ucts, interfaces between SPARTA light for different levels of turbulence which required mechanical improvement to the
and these components had to be devel- were run in parallel to the applied correc- design was then made by KRP Mechatec
oped. The real-time computer measures tion by the AO loop. This approach was GmbH who also provided the tooling
24 gradients on the wavefront sensor at much more flexible and more useful than required for the MAIT and the AIV. The
a frequency up to 500 Hz and corrects using a turbulence generator with phase electronics development and integration
up to 14 modes on the deformable mirror plates. Finally, a derotator integrated were supported by the electronics groups
depending of the brightness of the star. into the Relay Optics Structure was used at ESO in order to account for the evolu-
to create the pupil rotation on the wave- tion of the ATs and to conduct a thorough
front sensor to be fully representative of analysis of the cabling and interface.
Manufacturing, Assembly, Integration the Paranal conditions. An infrared cam-
and Testing (MAIT) era was integrated into the bench to The system test began in November
record the resulting star-like image. This 2017, led by the system engineer and
The MAIT phase began in January 2017 same camera was later re-used during the AO engineer while other team mem-
with several activities in parallel. IPAG the Assembly, Integration and Verification bers produced and tested the final ver-
worked on the deformable mirror charac- (AIV) phase for the stand-alone charac- sions of the electronics, wavefront sen-
terisation, the corrective optics integra- terisation of NAOMI on sky. sors and corrective optics. In addition to
tion and its dedicated electronics while the NAOMI test bench, the project team
ESO worked on the wavefront sensor, The NAOMI software is fully embedded developed dedicated test benches for
the electronics, the real-time computer, in the AT software. It was developed in the electronics, the wavefront sensor and
the software and the test bench. The the control model implemented at ESO the corrective optics. This allowed the
integration and alignment of the NAOMI headquarters before being tested on
test bench were carried out by members the NAOMI test bench. The ESO software
of the ESO Mechanical Engineering department obtained two years of exter- Figure 3. NAOMI test bench in integration.
and Optical Engineering Departments. nal support from CNRS in order to ensure The corrective optics are seen at the top right.
The goal was to replicate as accurately
as possible the behaviour of an on-sky
Emmanuel Aller Carpentier

target in the laboratory as it would be

observed by an AT at Paranal.

The bench consisted of an optical

table and a replica of the Relay Optics
­Structure used by the AT. The optical
table facilitated simulation of sky and
AT behaviour. An artificial on-sky target
was created via a monomode fibre
connected to a halogen lamp. A pupil
mask was used to define the spider of
the AT (the spider is the strut that sup-
ports the mirror) while the deformable
mirror itself simulated the atmospheric
turbulence thanks to its high density

The Messenger 177 – Quarter 3 | 2019 21

Telescopes and Instrumentation Gonté F. et al., Bringing NAOMI into Operation

Figure 4. Some of the members of the NAOMI pro-

Thibaut Moulin

ject team during AIV; first row, from left to right:

­T hibaut Guerlet, Christian Soenke, Sylvain Rochat,
Luis Caniguante; second row, left to right: Pascaline
Darré, Pabio Gutierrez, Emmanuel Aller Carpentier,
Frederic Gonté, Roderick Dembet, Stefan Huber,
Alain Delboulbé, Peter Krempl, Laurent Jocou, Pierre
Bourget, Sylvain Guieu, Thibaut Moulin.

­ ifferent aspects of the implementation.

The first team took care of the cooling
and cabling upgrade of the Relay Optics
Structure and the implementation of the
NAOMI components. The second worked
on the replacement and upgrade of
the electronics cabinets while the third
focussed on the internal re-alignment
of the ATs with the corrective optics (Fig-
ure 5). With so many activities in parallel,
verification of all of the components to be required for AIV. The NAOMI AIV was regular reorganisation and reassignment
delivered to Paranal independently of the first time that a system was imple- of tasks was necessary, requiring the
the main NAOMI test bench, which was mented simultaneously on four tele- staff to be flexible.
fully dedicated to the NAOMI functional scopes at Paranal. To mitigate the risks,
and system test. the main procedures were fully rehearsed The assembly and integration part of the
before the start of AIV. The plan also AIV proceeded faster than planned,
In parallel with the systems test, the AIV included adequate contingency in the and on 18 September the first AT was
team and several engineers from Paranal schedule, and redundancies in the staff ready for the first stand-alone verification
were trained on the test bench every time competences. on sky, during which the functionalities
a corrective optics or a wavefront sensor and AO performance could be evaluated
needed to be exchanged to ensure a high AIV began on 6 September 2018, follow- (see Figure 6). Soon after, two telescopes
level of confidence in the procedures. ing a few days of preparation during were available for verification working
Provisional Acceptance Europe, held in which the four Relay Optics Structures together with the VLTI. It was impressive
July 2018, concluded these system tests. were removed and transported to the to see the NAOMI team simultaneously
New Integration Hall and the first AT in the New Integration Hall, in the AT Ser-
was moved to the service station to allow vice station, in the VLTI building, and in
Assembly, Integration and Verification the removal of the fast steering mirror the Control Room working hard toward
and the implementation of the corrective getting the first closed-loop with NAOMI.
AIV was prepared well before the arrival optics. The AIV included 32 missions from
of the NAOMI equipment on Paranal. Europe, with an average of nine project
Several missions to Chile were necessary Many activities were then carried out in members working at Paranal over a two-
to verify the ATs, check the interfaces, parallel, with three teams focusing on month period.
and ensure the feasibility of the AIV
plan via discussions with Paranal staff.
Julien Woillez

The plan accounted for various factors,

including the availability of tools such as
the crane, the staff, the locations where
various activities had to be performed,
and the logistics of accommodating the
14-member project team at the

It became apparent that science opera-

tions of the VLTI with ATs would have
to be suspended during AIV of NAOMI;
which necessitated minimising the time

Figure 5. Intervention at the beginning of the

night on the corrective optics of AT4 conducted
by ­Pascaline Darré and Sylvain Guieu.

22 The Messenger 177 – Quarter 3 | 2019

the first preset on the selected target
Pierre Haguenauer

can be launched by the instrument. The

sky coordinates are sent to the VLTI
which propagates them to each AT.

NAOMI has a specific acquisition proce-

dure called discoball to detect any object
within 22.4 arcseconds. This permits pre-
cise centring of the AT in the 6-arc­second
acquisition field of view of NAOMI and
subsequent closing of the loop. The field
of view in the closed loop is 2.25 arc­
seconds. The full process between the
launch of the preset and the closed loop
is automatic, requiring approximately
140 seconds. A manual procedure can
be used for very specific targets, such
Figure 6. Acquisition of a star with the infrared Strehl of NAOMI closely during these opera- as a close binary. At the end of the
camera (H-band) in open loop (left) and in closed
tions. The team made extensive use of night, the NAOMI system is automatically
loop (right) for quasi-instantaneous Strehl measure-
ment (50 ms integration time) during the stand-alone the Garching Remote Access Facility re-checked while the ATs are being
verification of AT1. before the second commissioning, in closed and set in standby mode for the
order to solve issues discovered by the day. The day team then checks the
Paranal operation team. ­calibration data and applies any correc-
Commissioning tions that may be needed before the
The second commissioning period, which next VLTI start up.
The NAOMI project had two commis­ began on 25 February 2019, focused
sioning periods led by IPAG and sup- on the VLTI instrument performance All the telescopes used by the VLTI are
ported by ESO; these involved 14 mis- gains resulting from NAOMI, in particular now in operation with adaptive optics
sions. During this time NAOMI was also the improvement for faint stars, with systems as recommended by Pierre
commissioned on the current instrument lower than average seeing and the AT Léna (Léna, 1987) and Jacques Maurice
suite at VLTI: GRAVITY (GRAVITY Collab- dome seeing effect. NAOMI greatly ­Beckers (Beckers, 1990); we remain
oration et al., 2017), the Multi AperTure improves the injection flux in the spatial indebted to them for their vision.
mid-Infrared SpectroScopic Experiment filters and/or the fibres of the interfero-
(MATISSE; Lopez, 2014) and the Preci- metric instruments. It improves the signal
sion Integrated Optics Near Infrared to noise of each instrument and signifi- References
ExpeRiment (PIONIER; Zins, 2011). Com- cantly reduces the loss of fringes on the Arsenault, R. et al. 2003, The Messenger, 112, 7
missioning began with the first light of Gravity Fringe Tracker. NAOMI is much Arsenault, R. et al. 2017, The Messenger, 168, 8
NAOMI using the four ATs and GRAVITY more robust to atmospheric conditions Beckers, J. M. et al. 1990, The Messenger, 60, 1
on 20 October. It was dedicated to verify- than STRAP. It corrects the dome seeing Bonaccini, D. et al. 1997, Proc. SPIE, 3126, 77
Gitton, P. et al. 2004, Proc. SPIE, 5491, 944
ing functionality with the VLTI and ensur- observed with low wind conditions and Gravity Collaboration 2017, The Messenger, 170, 10
ing stable operation. In order to measure corrects the atmospheric seeing well Haguenauer, P. & Guieu, S. 2019, Proc. AO4ELT
how NAOMI improved the performance of above 1.4 arcseconds. The performances conference, in press
the PIONIER and GRAVITY instruments, of VLTI instruments with NAOMI are Koehler, B. et al. 2004, The Messenger, 115, 15
Koehler, B. et al. 2002, The Messenger, 110, 21
several indicators were monitored, includ- detailed in Woillez et al. (2019). Le Bouquin, J. B. et al. 2018, Proc. SPIE, 1070371
ing throughput, flux drop-out estimators Léna, P. 1987, The Messenger, 50, 53
and fringe tracking residuals. All meas- Lopez, B. et al. 2014, The Messenger, 157, 5
urements were compared against tip/tilt- Operations Scheithauer, S. et al. 2016, Proc. SPIE, 9909,
only correction. The chopping capability Suárez Valles, M. 2012, Proc. SPIE, 8447, 84472Q
was functionally tested with MATISSE. NAOMI is now fully integrated in the oper- Woillez, J. et al. 2015, The Messenger, 162, 18
The reliability of the NAOMI system was ation of the VLTI. A typical observing Woillez, J. et al. 2019, A&A, 629, A41
analysed and found slightly non-compliant night always follows the same procedure: Zins, G. et al. 2011, The Messenger, 146, 12
owing to a few issues that had a minor two hours before sunset, the VLTI start-
impact on operation. At the end of the up procedure is launched, the functionali- Links
first commissioning on 17 November the ties of each subsystem are automatically
ATs were transferred back into science verified, and the optical path is validated  daptive Optics for Extremely Large Telescopes
Conference (A04ELT) held in Québec City, Canada
operation. up to the InfraRed Imaging S ­ ensor (IRIS; in June 2019:
Gitton, 2004) in the VLTI Laboratory —
The ESO team with the support of IPAG allowing the verification of the alignment
continued to follow the performance of the pupil and of the field. After twilight,

The Messenger 177 – Quarter 3 | 2019 23

Astronomical Science






EHT Collaboration/M. Kornmesser/ESO

The centre of the giant elliptical galaxy M87 seen

at spatial resolution scales spanning six orders of
magnitude. The detailed structure of the relativistic
jet is revealed by observations at different radio
wavelengths using several interferometric facilities,
zooming into the supermassive black hole imaged
by the EHT collaboration.

Astronomical Science DOI: 10.18727/0722-6691/5150

First M87 Event Horizon Telescope Results and

the Role of ALMA

Ciriaco Goddi 1, 2 5
Onsala Space Observatory, Chalmers lengths that for the first time included
Geoff Crew 3 University of Technology, Sweden the Atacama Large Millimeter/
Violette Impellizzeri 4 6
Department of Astronomy and Astro- submillimeter Array (ALMA). The addi-
Iván Martí-Vidal 5, 6 physics/Astronomical Observatory, tion of ALMA as an anchor station
Lynn D. Matthews 3 ­University of Valencia, Spain has enabled a giant leap forward by
Hugo Messias 4 7
Max-Planck-Institut für Radioastronomie increasing the sensitivity limits of the
Helge Rottmann 7 (MPIfR), Bonn, Germany EHT by an order of magnitude, effec-
Walter Alef 7 8
Center for Astrophysics | Harvard & tively turning it into an imaging array.
Lindy Blackburn 8 Smithsonian, Cambridge, USA The published image demonstrates that
Thomas Bronzwaer 1 9
Steward Observatory and Department it is now possible to directly study
Chi-Kwan Chan 9 of Astronomy, University of Arizona the event horizon shadows of SMBHs
Jordy Davelaar 1 ­Tucson, USA via electromagnetic radiation, thereby
Roger Deane 10 10
Centre for Radio Astronomy Tech- transforming this elusive frontier from
Jason Dexter 11 niques and Technologies, Department a mathematical concept into an astro-
Shep Doeleman 8 of Physics and Electronics, Rhodes physical reality. The expansion of
Heino Falcke 1 University, Grahamstown, South Africa the array over the next few years will
Vincent L. Fish 3 11
Max-Planck-Institut für Extraterres- include new stations on different conti-
Raquel Fraga-Encinas 1 trische Physik, Garching, Germany nents — and eventually satellites in
Christian M. Fromm 12 12
Institut für Theoretische Physik, Goethe space. This will provide progressively
Ruben Herrero-Illana 18 Universität, Frankfurt am Main, sharper and higher-fidelity images of
Sara Issaoun 1 Germany SMBH candidates, and potentially even
David James 8 13
Joint Institute for VLBI ERIC (JIVE), movies of the hot plasma orbiting
Michael Janssen 1 Dwingeloo, the Netherlands around SMBHs. These improvements
Michael Kramer 7 14
Anton Pannekoek Institute for Astron- will shed light on the processes of black
Thomas P. Krichbaum 7 omy, University of Amsterdam, the hole accretion and jet formation on
Mariafelicia De Laurentis 19, 20 Netherlands event-horizon scales, thereby enabling
Elisabetta Liuzzo 21 15
Instituto de Radioastronomía Milimétrica, more precise tests of general relativity
Yosuke Mizuno 12 IRAM, Granada, Spain in the truly strong field regime.
Monika Moscibrodzka 1 16
Mullard Space Science Laboratory,
Iniyan Natarajan 10 University College London, Dorking,
Oliver Porth 14 UK Supermassive black holes and their
Luciano Rezzolla 12 17
Kavli Institute for Astronomy and Astro- shadows: a fundamental prediction of
Kazi Rygl 21 physics, Peking University, Beijing, general relativity
Freek Roelofs 1 China
Eduardo Ros 7 18
ESO Black holes are perhaps the most
Alan L. Roy 7 19
Dipartimento di Fisica “E. Pancini,” ­fundamental and striking prediction of
Lijing Shao 17, 7 ­Universitá di Napoli “Federico II”, ­Einstein’s General Theory of Relativity
Huib Jan van Langevelde 13, 2 Naples, Italy (GR), and are at the heart of fundamental
Ilse van Bemmel 13 20
INFN Sez. di Napoli, Compl. Univ. di questions attempting to unify GR and
Remo Tilanus 1, 2 Monte S. Angelo, Naples, Italy quantum mechanics. Despite their impor-
Pablo Torne 15,7 21
INAF–Istituto di Radioastronomia, tance, they remain one of the least tested
Maciek Wielgus 8 ­Bologna, Italy concepts in GR. Since the 1970s, astron-
Ziri Younsi 16,12 omers have been accumulating indirect
J. Anton Zensus 7 evidence for the existence of black holes
on behalf of the Event Horizon In April 2019, the Event Horizon Tele- by studying the effects of their gravita-
­Telescope collaboration scope (EHT) collaboration revealed the tional interaction with their surrounding
first image of the candidate super- environment. The first such evidence
massive black hole (SMBH) at the cen- came from the prototypical high-mass
 epartment of Astrophysics, Institute
D tre of the giant elliptical galaxy Messier X-ray binary Cygnus X-1, where a star
for Mathematics, Astrophysics and 87 (M87). This event-horizon-scale orbits an unseen compact object of ~ 15
­Particle Physics (IMAPP), Radboud image shows a ring of glowing plasma solar masses, apparently feeding on
­University, Nijmegen, the Netherlands with a dark patch at the centre, which is material from its stellar companion at only
Leiden Observatory—Allegro, Leiden interpreted as the shadow of the black 0.2 au. More evidence has come from
University, Leiden, the Netherlands hole. This breakthrough result, which studies of the Galactic Centre, where
Massachusetts Institute of Technology represents a powerful confirmation of ~ 30 stars have been tracked in tight, fast
Haystack Observatory, Westford, USA Einstein’s theory of gravity, or general orbits (up to 10 000 km s –1) around a
Joint ALMA Observatory, Vitacura, relativity, was made possible by assem- radio point source named ­Sagittarius A*
Santiago de Chile, Chile bling a global network of radio tele- or Sgr A* (Gillessen et al., 2009), ­ practi-
scopes operating at millimetre wave- cally ruling out all mechanisms

The Messenger 177 – Quarter 3 | 2019 25

ESO/L. Benassi/O.Furtak
Astronomical Science Goddi C. et al., First M87 Event Horizon Telescope Results and the Role of ALMA




SPT > 2020

r­ esponsible for their motions, except for a its Schwarzschild radius: Figure 1. Locations of the participating telescopes
of the Event Horizon Telescope (EHT; shown in blue)
black hole with a mass of about four mil- RSch = 2 GMBH/c2 = 2 rg,
and the Global mm-VLBI Array (GMVA; shown in
lion solar masses. where rg is the gravitational radius, MBH is ­yellow) during the 2017 global VLBI campaign. Addi-
the black hole mass, G is the gravitational tional telescopes that will observe in 2020 are
Perhaps the most compelling evidence constant, and c is the speed of light. The shown in light blue; the GLT also joined in the cam-
paign conducted in 2018. Planned telescopes that
came in 2015, with the detection by angular size, subtended by a non-rotating
may join the EHT in the future are shown in green.
the advanced Laser Interferometer BH with diameter 2 RSch is:
­Gravitational-Wave Observatory (LIGO) of qSch = 2 RSch/D ≈ 40 (MBH /106 M☉)(kpc/D)
gravitational waves: ripples in space-time in microarcseconds (µas), where the candidate SMBH in the Universe. With
produced by the merger of two stellar- black-hole mass is expressed in units of a mass of 4.15 million solar masses and
mass black holes (Abbot et al., 2016). one million solar masses and the black at a distance of 26 400 light years or
Despite this breakthrough discovery, hole’s distance (D) is in kiloparsecs. For 8.1 kpc (Gravity collaboration et al., 2019),
there was until very recently no direct evi- stellar-mass black holes (with masses this SMBH is a factor of a million times
dence for the existence of an event of a few to tens of solar masses), qSch larger than any stellar mass black hole in
­horizon, the defining feature of a black lies well below the resolving power of any the Galaxy and at least a thousand times
hole and a one-way causal boundary current telescope. SMBHs, which are closer than any other SMBH in other
in spacetime from which nothing (includ- thought to reside at the centre of most ­galaxies. The second-best candidate is
ing photons) can escape. On 10 April galaxies, are millions to billions of times found in the nucleus of the giant elliptical
2019, the EHT provided the very first the mass of the Sun, but as they are galaxy M87, the largest and most mas-
resolved images of a black hole, demon- located at much greater distances, their sive galaxy within the local supercluster
strating that they are now observable apparent angular sizes are also generally of galaxies in the constellation of Virgo.
astrophysical objects and opening a too small to be resolved using conven- Located 55 million light years from the
new and previously near-unimaginable tional observing techniques. Fortunately, Earth (or 16.8 Mpc), it hosts a black hole
window onto black hole studies. there are two notable exceptions: Sgr A* of 6.5 billion solar masses. Therefore,
and the nucleus of M87. even though M87 is ~ 2000 times as dis-
In order to conduct tests of GR using tant, it is ~ 1500 times as massive as
astrophysical black holes, it is crucial to Sgr A*, yielding a (slightly) smaller but
observationally resolve the gravitational Sgr A* and the nucleus of M87: the comparable angular size of the black hole
sphere of influence of the black hole, “largest” black hole shadows in our sky shadow on the sky. Owing to the combi-
down to scales comparable to its event nation of their masses and proximity,
horizon. For a non-rotating black hole, Sgr A*, at the centre of our own Galaxy, both Sgr A* and the nucleus of M87 sub-
the radius of the event horizon is equal to hosts the closest and best constrained tend the largest angular size on the sky

26 The Messenger 177 – Quarter 3 | 2019

among all known SMBHs (qSch ≈ 20 and

ALMA (ESO/NAOJ/NRAO), J. Pinto & N. Lira

15 μas, respectively). This makes Sgr A* Location Location Location
and M87 the two most suitable sources A B C
for studying the accretion process and jet







formation in SMBHs, even enabling tests



θ θ θ
of GR at horizon-scale resolution.
Radio telescopes

The “shadow” of a black hole

Analog Analog Analog
0 0 0 0 010 0 0 0 0 0 0 0 0 010 0 0 0 0 0 0 0 0 010 0 0 0 0
0 0 0 111 0 1 0 0 0 0 0 0 111 0 1 0 0 0 0 0 0 111 0 1 0 0 0
Digital conversion
Although by definition black holes cannot 0 0 0 11 0 1 0 0 0 1
0 11 0 0 1 0 111 0
0 0 0 11 0 1 0 0 0 1
0 11 0 0 1 0 111 0
0 0 0 11 0 1 0 0 0 1
0 11 0 0 1 0 111 0

be seen, we can detect light which

passes very close to the event horizon Atomic clocks

before escaping, allowing us to see what

is around the black hole.
Hard drives at
radio telescopes
So what would a black hole actually look
like if we could observe it? David Hilbert
began calculating the bending of light
around a Schwarzschild (non-rotating)
black hole in 1917. Bardeen (1973) subse-
quently calculated the geometrical prop-
erties of a rotating black hole’s silhouette
against a bright background (an orbiting
star). Although the likelihood of a black
hole passing in front of a star is very
small, black holes never appear “naked”
in astrophysical environments since their
extreme gravitational fields will pull and
compress matter from their surroundings,
eventually forming a disc of luminous
Data reduction
plasma. Luminet (1979) performed simu-
lations of a black hole surrounded by a
geometrically thin, optically thick, accre-
tion disc. Falcke, Melia and Agol (2000)
demonstrated that an accreting black
hole embedded in a plasma that is opti-
cally thin at millimetre wavelengths (like
the plasma expected to surround Sgr A*) Image

would produce a bright ring of emission

with a dim “shadow” cast by the black
hole event horizon in its interior. They
suggested that such a shadow might be
detectable towards the Galactic Centre in an angular diameter on the sky of ~ 50 Figure 2. A schematic diagram of the VLBI tech-
nique. Radio wave signals collected by individual
using the technique of very long baseline and ~ 40 μas (as viewed from the Earth)
antennas are converted from analogue to digital and
interferometry (VLBI) at millimetre for Sgr A* and M87, respectively. Although recorded onto hard disks together with the time­
wavelengths a. very small, this angular size can now be stamps provided by extremely precise atomic clocks
resolved by the VLBI technique at milli- at each station. In the 2017 campaign, a total of
about 4000 TB of recorded data was obtained. The
The shadow and ring are caused by a metre wavelengths using the EHT.
data were shipped from each station to a central
combination of light bending and photon location where a supercomputer (the correlator)
capture at the event horizon. The size combined the signals between all pairs of antennas
scale of the emission ring is set by Imaging black holes with the (synchronised using the timing information at each
station). The output of the correlator is hundreds of
the photon capture radius Rc. For a non-­ Event Horizon Telescope (EHT)
gigabytes, which is further reduced during data
rotating Schwarzschild black hole, ­c alibration down to tens to hundreds of megabytes.
Rc = √27 rg ~ 2.5 RSch. The factor of ~ 2.5 The VLBI technique at millimetre The end product of the VLBI data processing is an
comes from gravitational lensing, which wavelengths astronomical image.
increases the radius of the photon ring
with respect to the Schwarzschild radius For VLBI to work, a network of radio tele- (see, for example, Figure 1) must observe
as seen from the observer, resulting scopes spread across different continents the same source at exactly the same time

The Messenger 177 – Quarter 3 | 2019 27

Astronomical Science Goddi C. et al., First M87 Event Horizon Telescope Results and the Role of ALMA

achieve an angular resolution as fine as operating in over 20 countries/regions.

EHT Collaboration

M87* 11 April 2017

20 µas, which is sufficient to resolve the The key elements of the road­map
shadow of both Sgr A* and M87. There- towards an imaging array were the addi-
fore, the VLBI technique effectively mimics tion of new sites to better sample the
a virtual telescope with the size of the Fourier plane, and the improvement in
Earth. sensitivity needed to detect weak signals
on short timescales (Doeleman et al.,
While VLBI is well-established at centi­ 2010). Two technological developments
metre wavelengths, its extension to wave- were crucial for the latter: (1) improve-
lengths as short as 1.3 mm only began in ment in the observing bandwidth by
the 1990s (for example, Padin et al., 1990; increasing recording rates — over the last
Krichbaum et al., 1997; ­Doeleman & 10 years, EHT data rates increased from
Krichbaum, 1999). Challenges at shorter 4 to 64 Gb per second (Gbps); and (2)
wavelengths include the reduced aper- development of phased-array systems to
50 µas ture efficiency and small diameter of radio- combine the collecting area of existing
telescopes, increased noise in radio connected-element (sub)millimetre inter-
5 April 6 April 10 April receiver electronics, higher atmospheric ferometers, which has led to the inclusion
opacity, and above all, stronger distortion in the EHT of ALMA and the Submilli­
effects on the wavefronts from water meter Array (SMA) and the future incor-
vapour in the troposphere, which limits the poration of the NOrthern Extended Milli-
phase coher­ence to only a few seconds. meter Array (NOEMA).
By 2003, several active galactic nuclei
0 1 2 3 4 5 6 (AGN) had been detected on interconti-
Brightness temperature (10 9 K) nental baselines between Pico Veleta Phasing ALMA: turning ALMA into a
(Spain) and the Heinrich Hertz Telescope giant single-dish VLBI station
Figure 3. Image of the supermassive black hole
M87* as obtained with the EHT (on four different (HHT – Arizona, USA) at wavelengths of
days) in April 2017. Top panel: EHT image of M87* both 1 and 2 mm (Krichbaum et al., 2004; ALMA is the most sensitive (sub)millimetre-
from observations on 11 April as a representative Doeleman et al., 2005). wavelength telescope ever built. It consists
example of the images collected during the 2017
of 54 12-metre and 12 7-metre antennas
campaign. The angular resolution of the observation
(20 μas) is shown in the lower right corner. The located on the Chajnantor ­plateau in
image is shown in units of brightness temperature. Formation of the EHT project the Atacama desert in Chile, the highest,
North is up and east is to the left. Bottom panels: ­driest (accessible) desert on the Earth,
similar images taken on d ­ ifferent days showing the
Early pathfinder experiments (Krichbaum and it ordinarily operates as a connected-
stability of the ring structure across the observing
week. et al., 1998) detected Sgr A* with the element interferometer. Although the
baseline between Pico Veleta and the implementation of a VLBI mode was
and in the same frequency band. Individ- Plateau de Bure Interferometer, but not part of the baseline project, the desir-
ual antennas record their signals (plus the resolution was insufficient to probe ability of phasing the entire array for VLBI
time stamps from very precise atomic horizon scales. In the mid-2000s, a had been recognised (Wright et al., 2001;
clocks) onto computer hard disks which focused effort to boost sensitivity through Shaver, 2003) and some of the architec-
are then shipped to a central location, increased bandwidth led to the develop- ture needed to sum signals from all
where a supercomputer (called a correla- ment of fully digital VLBI backends, ALMA antennas was built into the ALMA
tor) combines (cross-correlates) the sig- with the goal of intercontinental 1.3-mm correlator (Escoffier et al., 2007).
nals between all pairs of antennas, syn- VLBI of Sgr A*. These systems were
chronising them using the recorded deployed at sites in Arizona, California, Motivated by the prospect of using
timing information from each station. Fig- and Hawai’i, and event-horizon-scale ALMA for horizon-scale observations of
ure 2 is a diagram illustrating the data structures were detected in both Sgr A* supermassive black holes (Doeleman
acquisition and processing path in a VLBI and M87 (­Doeleman et al., 2008, 2012). et al., 2009a, 2010), the case for phasing
experiment. These precursory scientific results moti- ALMA was renewed. An international
vated a strategy aimed at building a global team, led by Doeleman at MIT Haystack
The achievable angular resolution for an 1.3-mm VLBI array capable of imaging Observatory, proposed an ALMA Phasing
interferometer is given by q ~ l /B (in the shadows of the SMBHs in both Sgr A* Project (APP) to the US National Science
­radians), where l is the observed wave- and M87, and spurred on the formation Foundation. The APP was accepted by
length and B is the maximum distance of the EHT project, which was proposed the ALMA Board in 2011 and was com-
between the telescopes (or baseline). during the US 2010 Decadal Survey pleted in 2018 thanks to an international
Hence, higher frequencies (shorter wave- (Doeleman et al., 2009a) b. effort with contributors from the USA,
lengths) and longer baselines provide Europe, and Asia.
the highest resolving power. At 1.3 mm Since then, the EHT collaboration (or
(corresponding to a radio frequency of EHTC) 1 has grown to include over 250 The heart of the APP is a beamformer
230 GHz), Earth-diameter VLBI baselines members representing ~ 60 institutes, system, which electronically combines

28 The Messenger 177 – Quarter 3 | 2019

the collecting area of ALMA by aligning precise hydrogen maser (required to Figure 4. The supermassive black hole at the centre
of M87. Left: The black hole feeds on a swirling disc
the signals from individual ALMA anten- properly tune and synchronise the signals
of glowing plasma (shown in red), driving a powerful
nas in phase to form a coherent sum in the VLBI array). Numerous software relativistic jet across several thousands of light years
­signal. Currently, up to 43 12-metre enhancements were also required, (shown in grey; simulation by Davelaar et al. 2018).
antennas are used for the phased array, including the implementation of an ALMA Bottom right: Approaching the black hole, gravity is
so strong that light is severely bent, creating a bright
but the number may be smaller depend- VLBI Observing Mode (VOM) and a
(almost circular) ring. The north–south asymmetry
ing on the array configuration and the phase solver to adjust the phase of each in the emission ring is produced by relativistic beam-
weather conditions. This effectively turns of the ALMA antennas during observa- ing and Doppler boosting (matter in the bottom
ALMA into a giant virtual single dish tions to allow coherent summation of part of the image is moving toward the observer) and
is mediated by the black hole spin (which is pointing
(hereinafter called “phased ALMA”), and their signals. The initial scope of the APP
away from Earth and rotating clockwise). Gravita-
is equivalent to adding a ~ 70-metre dish effort and implementation can be found tional lensing magni­fi es the apparent size of the
to existing mm-VLBI arrays. In order to in Doeleman et al. (2010); the final working black hole’s event horizon into a larger dark shadow.
phase ALMA successfully and incorpo- implementation is described in ­Matthews The emission between the photon ring and the
event horizon is due to emitting plasma either in the
rate it as a VLBI station, the APP had to et al. (2018) and Goddi et al. (2019).
accretion flow and/or at the footprint of the jet (this
add several new hardware and software emission is generally too dim to be detected by the
components. These included an optical A broad science case for the use of the EHT; see Younsi et al., in preparation for detail). Top
fibre link system to transport the phased- ALMA Phasing System was assembled right: While the EHT can zoom in very close to the
event horizon, down to scales of only 0.01 light years
sum signal from the ALMA Array Opera- by the international community in white
(or 3.7 light days), i.e., a region comparable to the
tions Site (at an altitude of 5100 m) to papers c (Fish et al., 2013; Tilanus et al., size of our Solar System, the relativistic jet (extended
the ALMA Operations Support Facility (at 2014). Starting in 2016, VLBI as an across several thousand light years) can be probed
2900 m), where a set of Mark 6 VLBI observing mode was made available to using ALMA intra-baselines, recorded during the
EHT observations (greyscale image; Goddi et al.
recorders was installed. ALMA’s original the astronomy community through the
2019 and in prep.).
rubidium clock was replaced with a more normal ALMA proposal system, with an

C. Goddi, Z. Younsi, J. Davelaar/M.Kornmesser/ESO

The Messenger 177 – Quarter 3 | 2019 29

Astronomical Science Goddi C. et al., First M87 Event Horizon Telescope Results and the Role of ALMA

M87* K-jet, snapshot responds to SNR > 10 on non-ALMA of imaging its key science targets. The
J. Davelaar, F. Roelofs

baselines and > 100 on ALMA base- global array included eight telescopes in
lines. Therefore, the addition of ALMA six different geographical sites: the South
into the EHT array greatly facilitates Pole Telescope (SPT), the Arizona Radio
detections, especially for weak signals Observatory’s Submillimeter Telescope
(for example, long baseline length or (SMT), the Large Millimeter Telescope
bad weather). Using ALMA as a highly Alfonso Serrano (LMT) in Mexico, the
sensitive reference station has enabled IRAM 30-metre telescope in Spain, the
critical corrections for ionospheric and SMA and the James Clerk Maxwell Tele-
tropospheric distortions at the other EHT scope (JCMT) in Hawai’i, and APEX and
sites (see EHTC et al., 2019c for details). ALMA in Chile. These telescopes pro-
II. ALMA has a central location in the EHT vided baseline lengths up to 10 700 km
array (see Figure 1), and is therefore towards M87, resulting in an array with a
essential for the baseline coverage and resolution of ~ 20 μas (details are pro-
image fidelity. Even though ALMA and vided in EHTC et al., 2019b).
the Atacama Pathfinder Experiment
(APEX) are extremely close geographi- Besides the EHT, which operates at a
cally, ALMA’s superior sensitivity wavelength of 1.3 mm (i.e., a frequency of
allows the EHT to detect signals with 230 GHz, ALMA Band 6), complementary
1 2 3 4 5 6 the required 10-second integration VLBI observations with ALMA were also
(Brightness temperature (10 9 K))0.5 times between all baselines, i.e., to find conducted at 3.5 mm or 86 GHz (ALMA
VLBI fringe solutions, which has a dra- Band 3) in concert with the Global mm-
Figure 5. Black hole model for M87* used for the
image reconstructions shown in Figure 6. This ­specific
matic impact on the imaging capability VLBI Array (GMVA) 3, which consists of up
model has a relatively bright jet footprint appearing of the EHT (see Figure 6). to 18 telescopes located in Europe, North
in front of the photon ring and a more extended jet III. VLBI observations with ALMA also America, and Asia. Figure 1 displays the
emission extending towards west. Note the bright provide connected-element interfero- geographical locations of all the partici-
knot to the south-west at the point where the jet
sheath crosses the photon ring in projection (see
metric data, which are archived, as pating telescopes in the EHT and the
Davelaar et al., 2019). with any standard ALMA project, and GMVA in 2017 (plus additional telescopes
are available to the user in the ALMA that joined after or plan to join in the near
expected maximum time allocation of archive (after the appropriate proprie- future).
~ 5 % of the total ALMA observing time. tary period). As outlined in Goddi et al.
(2019), the calibration of such interfero- The EHT 2017 science observing cam-
metric data allows one to determine paign was scheduled for April when
Impact of ALMA in the EHT array the absolute amplitude calibration Sgr A* and M87 are night-time sources
of the co-located sites ALMA–APEX in and tropospheric conditions tend to be
Owing to the combination of a large physical flux-density (i.e, Jy). This, in the best, averaged over all sites in the
effective aperture, its central location in turn, allows us to bootstrap source array. At that time, ALMA was in a more
the VLBI array, excellent typical atmos- fluxes and calibrate longer baselines compact configuration as required for
pheric conditions and ultra-low noise across the entire array (i.e., network phased array operations. About 40 EHT
receivers, the addition of ALMA drasti- calibration). In addition, since VLBI astronomers travelled to four continents
cally changed the overall capabilities of observations are always performed in
the global EHT array, boosting the full-polarisation mode (in order to Figure 6. Image reconstructions from synthetic
achievable signal-to-noise ratio (SNR) of ­supply input to the polarisation conver- data generated using the M87* model displayed in
VLBI baselines by more than an order of sion process at the VLBI correlators — Figure 5 as input. Each panel shows a reconstruc-
tion from simulated observations with a different
magnitude with respect to the first hori- see Martí-Vidal et al., 2016; Goddi et array: 2017 EHT array (left column); planned 2020
zon-scale detections (Doeleman at al., al., 2019), the ALMA full-polarisation array (middle column); and a future EHT array includ-
2008, 2012). More specifically, the inclu- interferometric datasets can be used ing AMT in Namibia, and LLAMA in Argentina (right
sion of ALMA in the EHT provides three to derive mm-wavelength emission column). A comparison between the top row (with
ALMA) and the middle row (without ALMA) highlights
key advantages: a boost in sensitivity; and polarisation properties of each the importance of ALMA, demonstrating that the
improved baseline coverage; and an ­target observed by the EHT on arcsec- shadow feature can only be recovered if ALMA
accurate measure­ment of the absolute ond scales (Goddi et al., in preparation). is part of the array. A comparison between the mid-
flux-density scale and polarisation frac- dle row and the bottom row (where APEX is also
excluded) clearly demonstrates that, even with a
tions of EHT targets (using standard ­s ingle-dish telescope in the same geographical loca-
ALMA interferometric data). We expand The first global VLBI campaigns with tion of ALMA, it is not possible to recover an image
on these characteristics below. ALMA with sufficiently high fidelity to discern the shadow.
Although with future arrays the quality of the recon-
structed image will improve and it will be possible to
I. The median thermal noise of non- Phased ALMA joined the EHT array for recover new features (for example, the jet), the addi-
ALMA baselines is 7 mJy, and 0.7 mJy the first time in April 2017, performing tion of new stations cannot fully compensate for the
in ALMA baselines; for M87 this cor­ VLBI observations with an array capable loss of ALMA (middle and right columns).

30 The Messenger 177 – Quarter 3 | 2019

EHT2017 EHT2020 EHT2020 + AMT + LLAMA
F. Roelofs, M. Janssen, C. Goddi

70 µas 70 µas 70 µas

0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 1 2 3 4 1 2 3 4

(Brightness temperature (10 9 K))0.5 (Brightness temperature (10 9 K))0.5 (Brightness temperature (10 9 K))0.5

EHT2017, no ALMA EHT2020, no ALMA EHT2020 + AMT + LLAMA, no ALMA

70 µas 70 µas 70 µas

0.5 1.0 1.5 2.0 2.5 3.0 3.5 0.5 1.0 1.5 2.0 2.5 3.0 3.5 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0
(Brightness temperature (10 9 K))0.5 (Brightness temperature (10 9 K))0.5 (Brightness temperature (10 9 K))0.5

EHT2017, no Chile EHT2020, no Chile EHT2020 + AMT + LLAMA, no Chile

70 µas 70 µas 70 µas

0.5 1.0 1.5 2.0 2.5 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0
(Brightness temperature (10 9 K))0.5 (Brightness temperature (10 9 K))0.5 (Brightness temperature (10 9 K))0.5

The Messenger 177 – Quarter 3 | 2019 31

Astronomical Science Goddi C. et al., First M87 Event Horizon Telescope Results and the Role of ALMA

and drove to mountaintops to carry out are provided in EHTC et al. (2019c,d). the black hole was modelled with general
the observations, which included six Multiple independent analyses were also relativistic calculations spanning a wide
­science targets: the primary EHT targets, performed in order to verify the results range of possible accretion states (see
Sgr A* and M87, and the AGN targets, (EHTC et al., 2019e,f). After about two next subsection). By tracing the peak of
3C 279, OJ 287, Centaurus A, and years of dedicated work by many dozens the emission in the ring we can determine
NGC 1052d. Weather is a crucial f­actor of EHT scientists in multiple working the shape of the image, which is close
for VLBI observations at mm wave- groups (from instrument through data to circular with an axial ratio 4:3 (corre-
lengths, which is why the EHT uses flexi- processing to theory), the collaboration sponding to a 10% deviation from circu-
ble observing schedules with windows was finally ready to communicate our larity). The emission in the ring is asym-
that are about twice as long as the breakthrough to the world. metric and is brighter in the south, which
approved number of observing nights. An can be explained as relativistic beaming
array-wide go/no-go decision was made of plasma rotating (close to the speed
a few hours before the start of each The breakthrough: first image of a black of light) in the clockwise direction around
observing night, based on the weather hole shadow the black hole as seen by the observer
conditions and predictions at each site, (i.e., the bottom part of the emission ring
as well as technical readiness at each of The first image of a black hole was is Doppler-boosted towards the Earth).
the participating telescopes. Observa- ­published by the EHT Collaboration on Based on our modelling and information
tions were triggered over a 10-night win- 10 April 2019 in a series of six scientific on the inclination angle of the relativistic
dow 5–14 April 2017 — on 5, 6, 7, 10, publications (EHTC et al., 2019a,b,c,d,e,f). jet (observed on larger scales), we derive
and 11 April. During the whole campaign, The announcement and the historic the sense of rotation of the black hole to
the weather was good to excellent at image were released worldwide in six be in the clockwise direction, i.e., the spin
most stations. In addition to favourable simultaneous press conferences — axis of the black hole points away from us.
weather conditions, the VLBI-specific in Washington, Brussels, Santiago,
technical setup and operations at all sites ­Shanghai, Taipei and Tokyo — with addi- A number of elements reinforce the
were successful, which resulted in fringe tional satellite events in Rome, Madrid, robustness of the result. The data analy-
detections across the entire array. Munich, Leiden and Nijmegen, amongst sis used four independent data sets
others. The core of M87, as imaged taken on four different days (spanning a
By the end of the campaign, 96 disk mod- with the EHT, was renamed M87*, in line one-week observing window) in two
ules were used, each containing eight with the name of the black hole candidate ­separate frequency bands (centered at
helium-filled hard disks (of either six or at the centre of the Galaxy, Sgr A*. 227 and 229 GHz). The top part of Fig­-
eight terabyte [TB] capacity), correspond- ure 3 shows an image of M87* on 11 April,
ing to more than 5 petabytes (PB) of The most striking feature of the image while the bottom panels show similar
(removable) storage; about 4 PB of data (displayed in Figure 3) is a bright circular images from three different days. The
were eventually recorded in total. Given ring with an asymmetric brightness distri- diameter and width of the ring remain
this huge volume of data, the observing bution and a dark region at its centre, stable and the image features are broadly
campaign VLBI data could not be trans- which identifies the black hole shadow. consistent across all four observing days,
ferred over the internet, but were shipped The ring reveals the curvature of space- except the position angle of the bright
from each remote station to the two EHT time due to the extreme gravitational part in the asymmetric azimuthal profile,
correlator centres for processing (the field around a SMBH, which bends light which varies in the range 150–200 degrees
shipping took at least several days, and around it, creating an almost circular measured from north towards the east
many months in the case of the South shadow at its centre. In fact, GR predicts between the first two days and the last
Pole telescope). the shadow to be circular to within a two days.
few percent, whereas alternative theories
of gravity predict distorted, non-circular Overall, the size, circularity, asymmetry,
Path to the image shapes (Younsi et al., 2016; also see Fig- and brightness contrast of the observed
ure 7 of Goddi et al., 2017). image are consistent with the shadow of
The EHT data were correlated at the MIT a “Kerr” black hole as predicted by
Haystack Observatory in Westford, USA The ring has a measured diameter of GR and provide the strongest evidence to
and at the MPIfR in Bonn, Germany. 42 ± 3 μas and the central brightness date of the existence of SMBHs in the
Three independent data calibration pipe- depression has a contrast ratio > 10:1. nuclei of external galaxies.
lines and three imaging pipelines, each The measured angular size, assuming a
using a different software package and distance of 16.8 Mpc (EHTC et al., 2019f),
associated methodology, were used for implies a black hole mass of M = (6.5 ± Modelling and physical interpretation of
the data processing in order to produce 0.7) × 109 M☉, or 6.5 billion times the the black hole image
images independently. This approach mass of the Sun (consistent with one
encourages the data analysis and sci- ­earlier mass measurement — Gebhardt The appearance of M87* has been mod-
ence teams to minimise their biases in et al., 2011). To convert the measured elled using 3D general-relativistic mag-
terms of both methodology and human diameter of the ring into the mass of the netohydrodynamic (GRMHD) simulations,
decision making; further details of this black hole, the radiating plasma around which provide the physical conditions of

32 The Messenger 177 – Quarter 3 | 2019

the plasma and magnetic field surrounding Implications of the black hole shadow the baseline coverage and sensitivity,
the black hole. GR ray-tracing radiative- on tests of GR and complementarity thereby affecting the resulting images, is
transfer (GRRT) codes then take this with LIGO by performing simulated observations.
GRMHD simulation data as input and cal- Instrument simulators were specifically
culate the black hole’s appearance from Simulated images can be used to test built for the EHT to tie theoretical models
the emitted radiation field. Approximately basic properties of black holes as pre- to instrument measurements. In parti­cular,
60 000 simulated images were produced dicted in GR (for example, Psaltis et al., they can generate realistic synthetic data,
in the process (see EHTC et al., 2019e). 2015), or in alternative theories of gravity taking as input GRMHD model images,
Figure 4 showcases the main components (Younsi et al., 2016; Mizuno et al., 2018). and performing synthetic observations
of the M87 SMBH and their characteristic They can also be used to test alternatives using a specific EHT array and observing
scales by comparing observed and simu- to black holes (Olivares et al., 2019). We schedule (see EHTC et al., 2019d,e). As a
lated images (for one specific set of mod- estimate a deviation from circularity of demonstration, in Figure 5 we show a
els). In particular, in the simulation to the < 10 %, so we can set an initial limit on GRMHD simulation of the jet-launching
left (which combines emission at wave- relative deviations from GR. Although it is region of M87 from Davelaar et al. (2019).
lengths of 7, 3, and 1 mm; see Davelaar et difficult to rule out alternatives to black This specific model has a r­ elatively bright
al., 2018 for details), one can see that the holes in GR — a shadow can be pro- jet footprint appearing in front of the pho-
SMBH is embedded in an accretion flow duced by any compact object with unsta- ton ring and a more extended jet emission
and powers a bipolar relativistic jet. ble circular photon orbits (Mizuno et al., extending towards west. We can then test
2018) — we can readily exclude exotic how well the input image (the ground truth)
Zooming in closer to the centre (the alternatives to black holes, such as naked can be recovered with the current EHT
­simulation to the right in Figure 4 shows singularities or wormholes, which predict array and a ­ nalyse the effect of adding new
emission at 1 mm; Younsi et al., in prepa- much smaller shadows than we have stations and/or excluding existing stations.
ration), hot magnetised plasma orbiting measured, whereas others like boson
and accreting onto the black hole creates stars and gravastars need to be analysed Figure 6 shows some examples of recon-
the familiar emission ring structure around with more care (Olivares et al., 2019); also structed images of M87, using the model
the event horizon. As stated earlier, the see EHTC et al. (2019e) for further details. shown in Figure 5 as the input model,
size of the ring is set by the photon cap- produced with one specific EHT synthetic
ture radius: photons approaching the It is worth pointing out that the EHT result data generation pipeline (details are
black hole with an impact parameter b < Rc provides a new way to study black hole reported in Roelofs, Janssen and EHTC,
are captured and disappear into the black spacetimes and is complementary to the submitted). For instance, using the EHT
hole; photons with b > Rc escape to infin- detection experiments of gravitational array and schedule that observed on
ity; photons with b = Rc are captured waves from merging stellar-mass black 11 April 2017, the resulting simulated
on an unstable circular orbit and produce holes with LIGO/Virgo (Abbott et al., 2016). image (shown in the left panel, top row of
the so-called lensed photon ring. There are at least two main complemen- Figure 6) is similar to the one actually
tary aspects between gravitational-wave observed by the EHT. The middle and
While the EHT can resolve down to and electromagnetic observations of bottom rows show simulated images
scales of 0.01 light years (or 3.7 light- black holes: without ALMA, which best showcase its
days), i.e., a region comparable to the 1. Since EHT targets SMBHs and LIGO importance by clearly demonstrating
size of our Solar System, the jet extends mainly targets stellar-mass black holes, that the familiar ring structure cannot be
to much larger scales across several combining measurements from both reconstructed when ALMA is not part of
thousand light years and can be probed methods we can test whether one of the array. Although APEX shares the
using shorter baselines and/or lower the most fundamental properties of same geographical location as ALMA and
frequencies. In Figure 4 (top right), we black holes in GR, that their size scales therefore should provide similar baseline
show an image of the M87 jet, which linearly with mass, actually holds over coverage, the quality of the reconstructed
extends across about 20 arcseconds, eight orders of magnitude. image is not sufficient to discern the ring
corresponding to 5000 light years, 2. G ravitational wave experiments cannot when APEX is in the array and ALMA is
obtained at 1.3 mm using ALMA interfer- rely on the possibility of multiple and excluded. These simulations clearly sub-
ometric data (with maximum baseline repeated measurements of the same stantiate the need for ALMA’s sensitivity,
lengths of only a few hundred metres) source, whereas the EHT can be used which allows for numerous and strong
acquired simultaneously with the EHT to measure the shadow shape of M87 detections of weak signal e.
observations (Goddi et al., 2019 and in with ever increasing precision, leading to
preparation). The science section image progressively better constraints on black By adding new stations, the quality of the
(p. 24) showcases a montage of images hole parameters and their spacetime. reconstructed image improves and new
of the M87 relativistic jet observed at sev- features (for example, the jet) can be
eral radio wavelengths with multiple inter- recovered (see middle and right columns
ferometers at progressively higher angu- Importance of ALMA in imaging M87 in the top row of Figure 6), but these
lar resolution overlaid on the HST optical new stations cannot fully compensate for
image: the VLA at 21 cm, VLBA at 7 mm, The most straightforward way to visualise the loss of ALMA (see middle and right
GMVA at 3 mm, and EHT at 1.3 mm. how the loss of specific stations changes columns in the middle and bottom rows).

The Messenger 177 – Quarter 3 | 2019 33

Astronomical Science Goddi C. et al., First M87 Event Horizon Telescope Results and the Role of ALMA

Future directions the EHT’s capabilities over the coming measure parameters like black hole spin.
years will bring more exciting scientific While current terrestrial VLBI at 1.3 mm
The first EHT image of M87 has provided results. In 2018, the Greenland Telescope can resolve the black hole shadow only
very strong evidence for the existence (GLT) joined the EHT (and GMVA) and in Sgr A* and M87, adding satellites
of an event horizon and supports the VLBI observations were conducted as in space would significantly expand the
notion of SMBHs being located at the part of ALMA Cycle 5 (the analysis of range of sources that can be resolved
centre of galaxies. SMBHs present a new these observations is still ongoing). In on horizon scales. Combining ground-
tool to explore gravity at its most extreme 2019, EHT observations were abandoned based VLBI at 0.87 mm with space-
limit and on a mass scale that was because of operational difficulties at a based VLBI at longer wavelengths would
­hitherto inaccessible. Ongoing analysis of small number of key EHT sites. For 2020, provide better matching beam sizes,
existing data and future EHT observa- observations are planned during ALMA which are important for spectral index
tions will further help us understand Cycle 7 and will include new telescopes: and rotation measure studies. Studying
the nature of black holes and will provide Kitt Peak National Observatory (KPNO) in nearby low-luminosity AGN could fill
even more stringent tests of GR. Arizona, and the NOEMA interferometer the gaps in black hole mass, accretion/jet
in France. These new stations will provide power, and host galaxy type between
Future observations and detailed analysis intermediate baselines (<~ 1000 km) in Sgr A* and M87. Therefore all these devel-
of M87 data will explore the shape and Europe (NOEMA–IRAM-30m) and impor- opments will open up very exciting and
time variability of the shadow more accu- tant short baselines in the USA (KPNO– new scientific possibilities in the coming
rately. The EHT is in the process of study- SMT) (<~ 100 km), thus further extending decades.
ing the magnetised plasma around the baseline coverage for both M87
M87 in polarised light, which will allow us and Sgr A*. The possible future addition Finally, if we were to discover a radio pul-
to investigate the mechanism by which of the Africa Millimeter Telescope (AMT) sar on a tight orbit (period < 1 year)
black holes launch and power their rela- in Namibia (Backes et al., 2016) and around Sgr A*, this would allow us to
tivistic jets. the Large Latin American Millimeter Array measure the black hole properties (mass,
(LLAMA) in Argentina will add further distance and spin) more accurately
As for Sgr A*, the mass-to-distance ratio baseline coverage, including the long than currently possible with orbiting stars
is accurately measured from stellar baselines oriented east-west (AMT–ALMA) targeted by the AO-assisted, two-object,
orbits in the near-infrared (Gravity Collab- and the intermediate baseline LLAMA– multiple beam-combiner interferometric
oration et al., 2019), so measuring the ALMA (180 km). In particular, the addition VLTI instrument, GRAVITY, leading to a
shadow shape and diameter provides a of short/intermediate baselines of the clean test of the no-hair theorem (Psaltis,
null hypothesis test of GR (Psaltis et al., order of a few hundred km, which are Wex & Kramer, 2016). The detection of
2015). Since its mass is three orders sensitive to extended emission from the a magnetar at a projected distance of
of magnitude smaller than that of M87*, jet on scales > 100 µas, may enable us to 0.1 pc from Sgr A* (Eatough et al., 2013)
the dynamical timescales are minutes trace the jet down to the SMBH and suggests that finding a pulsar in a close
instead of days; therefore observing the directly image the jet launching. See the orbit around Sgr A* should be possible,
shadow of Sgr A* will require accounting middle and right columns of Figure 6 to and the recent detection of the Vela pul-
for this variability as well as the mitigation evaluate the impact of these new stations sar with phased ALMA (Liu et al., 2019)
of scattering effects caused by the inter- in recovering the jet structure; the location opens up the possibility of pulsar
stellar medium (Johnson, 2016). Time-­ of the new stations is also displayed in searches with ALMA at high frequencies
dependent non-imaging analysis can Figure 1. (where the effect of interstellar scattering
potentially be used to track orbits of hot is lower). The combination of the far-
spots near the black hole (Broderick & Higher-resolution images can be achieved field measurements (100s–1000s rg )
Loeb, 2006; Doeleman et al., 2009b; by going to a shorter wavelength (0.87 mm based on pulsars and stars, with the
Roelofs et al., 2017), as reported recently or 345 GHz, i.e., ALMA Band 7). A future near-field tests from imaging of black hole
on the basis of interferometric observa- array that combines observations at both shadows (10s rg), has the power to
tions in the near-infrared (Gravity Collabo- 1.3 and 0.87 mm will improve the imaging reveal deviations from the Kerr metric and
ration et al., 2018). Real-time movies may dynamic range, while multi-frequency provide a fundamental test of GR (Goddi
also become possible via interferometric VLBI would also open up spectral index et al., 2017), potentially leading to a
dynamical imaging (Johnson et al., 2017). and rotation measure studies. breakthrough in our understanding of
Time-domain studies and movies of black physics in the strong gravity regime.
holes can then be used to study black In the more distant future, extending
hole accretion and to map the black hole VLBI into space would provide the
spacetime, leading directly to measure- increased angular resolution necessary Acknowledgements
ments of black hole spin and tests of the to image finer structures and dynamics The authors would like to acknowledge all the
“no hair” theorem (Broderick et al., 2014). near the black hole shadow (Fish et al., ­scientists, institutes, observatories and funding
2019; Palumbo et al., 2019; Roelofs et al., agencies who are part of and collectively support
Although the focus of this article is the 2019). An order of magnitude increase the EHT project. The APP was supported by a Major
Research Instrumentation award from the National
first EHT result from the 2017 campaign, in angular resolution would allow us Science Foundation (NSF; award 1126433), an ALMA
it is noteworthy that enhancement of to perform precision tests of GR and to

34 The Messenger 177 – Quarter 3 | 2019

North American Development Augmentation award, Doeleman, S. et al. 2009a, astro2010, 68 Roelofs, F., Janssen, M. and EHTC, submitted to
ALMA North America (NA) Cycle 3 and Cycle 4 Doeleman, S. et al. 2009b, ApJ, 695, 59 A&A
Study awards, and an ALMA NA Cycle 5 Develop- Doeleman, S. 2010, EVN Symposium Proceedings, Shaver, P. A. 2003, Proc. of the workshop “The
ment award. The EHT project has been supported 53 Mass of Galaxies at Low and High Redshift”,
by multiple grants from many independent funding Doeleman, S. et al. 2012, Science, 338, 355 ESO Astrophysics Symposia, ed. Bender, R. &
agencies, including the ERC Synergy Grant “Black- Eatough, R. et al. 2013, Nature, 501, 391 Renzini, A., 357
HoleCam: Imaging the Event Horizon of Black Holes” EHT Collaboration et al. 2019a, ApJ, 875, L1 Tilanus, R. et al. 2014, arXiv:1406.4650
(Grant 610058) and several USA NSF grants (includ- EHT Collaboration et al. 2019b, ApJ, 875, L2 Wright, M. et al. 2001, ALMA Memo, 382
ing AST-1310896, AST-1440254, and OISE-1743747). EHT Collaboration et al. 2019c, ApJ, 875, L3 Younsi, Z. et al. 2016, Phys. Rev. D, 94, 084025
For the complete list of funding grants and acknowl- EHT Collaboration et al. 2019d, ApJ, 875, L4
edgements please see EHT Collaboration et al. EHT Collaboration et al. 2019e, ApJ, 875, L5
(2019a); they have not been reproduced here for EHT Collaboration et al. 2019f, ApJ, 875, L6 Links
­reasons of space and readability. We gratefully Escoffier, R. P. et al. 2007, A&A, 462, 801
acknowledge the support provided by the staff of Falcke, H., Melia, F. & Agol, E. 2000, ApJ, 528, 13  he Event Horizon Telescope webpage:
the ALMA observatory. Falcke, H. et al. 2012, The Messenger, 149, 50
Falcke, H. 2017, J. Phys., Conference Series 942, T he Global mm-VLBI Array webpage: http://www3.
This paper makes use of the following ALMA data: 012001
ADS/JAO.ALMA#2016.1.01154.V. Fish, V. et al. 2013, arXiv:1309.3519 A stronet 2007:
Fish, V. et al. 2019, arXiv:1903.09539 archives/oldpdfs/Astronet_ScienceVision_lowres.
ALMA is a partnership of ESO (representing its mem- Gebhardt, K. et al. 2011, ApJ, 729, 119 pdf
ber states), NSF (USA) and NINS (Japan), together Gillessen, S. et al. 2009, ApJ, 692, 1075
with NRC (Canada), MOST and ASIAA (Taiwan), and Goddi, C. et al. 2017, IJMP, 26, 1730001
KASI (Republic of Korea), in cooperation with the Goddi, C. et al. 2019, PASP, 131, 075003 Notes
Republic of Chile. The Joint ALMA Observatory is Gravity collaboration et al. 2018, A&A, 618, 15
operated by ESO, AUI/NRAO and NAOJ. Gravity collaboration et al. 2019, A&A, 625, 10  e refer to Luminet (2019) for a comprehensive
Janssen, M. et al. 2019, A&A, 626, 75 review of the history of early numerical simulations
Johnson, M. 2016, ApJ, 833, 74 of black-hole imaging during the period 1972–2002
References Johnson, M. et al. 2017, ApJ, 850, 172 and Falcke (2017) for a review of past, current and
Krichbaum, T. et al. 1997, A&A, 323, 17 future efforts to image black holes.
Abbot, B. P. et al. 2016, Phys. Rev. Lett., 116, Krichbaum, T. et al. 1998, A&A, 335, 106 T he strategic importance of millimetre-VLBI for
061102 Krichbaum, T. et al. 2004, EVN Symposium event-horizon-scale imaging of the Galactic Centre
Backes, M. et al. 2016, Proc. of High Energy Astro Proceedings, 15 was also recognised in the European Science
physics in Southern Africa (HEASA 2016), Liu, K. et al. 2019, submitted to ApJ Vision for astronomy, Astronet 20073.
South African Astronomical Observatory (SAAO), Luminet, J.-P. 1979, A&A, 75, 228 In building the science case for phased ALMA,
(Cape Town, South Africa), 029 Luminet, J.-P. 2019, arXiv:1902.11196 astronomers from all ALMA regions were asked for
Bardeen, J. M. 1973, in Black Holes, ed. DeWitt, C. & Martí-Vidal, I. et al. 2016, A&A, 587, A143 input at a variety of meetings in the USA and
DeWitt, B. S., 215 Matthews, L. D. et al. 2018, PASP, 130, 015002 Europe, including an ESO workshop (see Falcke et
Broderick, A. E. & Loeb, A. 2006, MNRAS, 367, 905 Moscibrodzka, M. et al. 2014, A&A, 570, 7 al., 2012).
Broderick, A. E. et al. 2014, ApJ, 784, 7 Mizuno, Y. et al. 2018, Nature Astronomy, 2, 585 In addition to being scientific targets, AGN obser-
Davelaar, J. et al. 2018, Computational Astrophysics Olivares, H. et al., arxiv:1809.08682 vations are very important to facilitate the intensity
and Cosmology, 5, 1 Padin, S. et al. 1990, ApJ, 360, 11 and polarisation calibration of the entire VLBI array.
Davelaar, J. et al. 2019, arXiv:1906.10065 Palumbo, D. et al. 2019, ApJ, 881, 62 A PEX participation is nevertheless extremely
Doeleman, S. & Krichbaum, T. 1999, 2nd millimeter- Psaltis, D. et al. 2015, ApJ, 814, 115 important as it provides the ALMA–APEX short
VLBI science workshop, ed. Greve, A. & Psaltis, D., Wex, N. & Kramer, M. 2016, ApJ, 818, baseline (2.6 km), which allows the robust absolute
­K richbaum, T. P., (St. Martin d’Heres, France), 73 121 calibration of visibility amplitudes (i.e., telescope
Doeleman, S. et al. 2005, ASPC, 340, 605 Roelofs, F. et al. 2017, ApJ, 847, 55 sensitivities).
Doeleman, S. et al. 2008, Nature, 455, 78 Roelofs, F. et al. 2019, A&A, 625, 124
Y. Beletsky/ESO

The groundbreaking ALMA array is composed

of 66 giant antennas situated on the Chajnantor
­Plateau in the Chilean Andes.

The Messenger 177 – Quarter 3 | 2019 35

Astronomical Science DOI: 10.18727/0722-6691/5151

The Physics at High Angular resolution in

Nearby GalaxieS (PHANGS) Surveys

Eva Schinnerer 1, 2 Dyas Utomo 3 HII regions, and star clusters. For the
Adam Leroy 3 Schuyler van Dyk 14 first time, the ESO facilities ALMA
Guillermo Blanc 4, 5 Jacob Ward 13 and MUSE, in combination with HST,
Eric Emsellem 6, 7 Brad Whitmore 26 offer the opportunity to survey the
Annie Hughes 8 properties of these regions and clusters
Erik Rosolowsky 9 across a large sample of galaxies,
Andreas Schruba 10 1 
Max Planck Institute for Astronomy, ­capturing the range of diverse galactic
Frank Bigiel 11 Heidelberg, Germany environments found in the local uni-
Andres Escala 5 2 
Associate Scientist, National Radio verse. Guided by theoretical models
Brent Groves 12 Astronomy Observatory, Charlottes- and simulations, the PHANGS collabo-
Kathryn Kreckel 1 ville, USA ration has begun an endeavour which
Diederik Kruijssen 13 3 
The Ohio State University, Columbus, aims to reveal the physical processes
Janice Lee 14 USA controlling the process of star formation
Sharon Meidt 15 4 
The Observatories of the Carnegie in galaxies.
Jerome Pety 16 Institution for Science, Pasadena, USA
Patricia Sanchez-Blazquez 17 5 
Universidad de Chile, Santiago, Chile
Karin Sandstrom 18 6 
ESO Understanding the physics of star
Antonio Usero 19 7 
Univ. Lyon, Univ. Lyon I, ENS Lyon, ­formation from detailed studies of
Ashley Barnes 11 CNRS, CRAL, Saint-Genis-Laval, nearby galaxies
Francesco Belfiore 6 France
Ivana Bešlić 11 8 
CNRS/IRAP & UPS-OMP, Toulouse, For more than 10 Gyr, most stars have
Rupali Chandar 20 France been formed in galactic disc-like systems
Dimitris Chatzigiannakis 11 9 
University of Alberta, Edmonton, in a secular mode. Star formation in
Melanie Chevance 13 Canada galactic discs, including in our own Milky
Enrico Congiu 4 10 
Max Planck Institute for Extraterrestrial Way, is thus intimately linked to the for-
Daniel Dale 21 Physics, Garching, Germany mation of structure. As far as we know,
Christopher Faesi 1 11 
AIfA University Bonn, Bonn, Germany the overwhelming bulk of star formation
Molly Gallagher 3 12 
Australian National University, in these galaxy discs occurs in cold, well-
Axel Garcia-Rodriguez 19 ­Canberra, Australia shielded giant molecular clouds (GMCs).
Simon Glover 22 13 
ARI/ZAH University Heidelberg, Therefore, the birth of stars in GMCs is
Kathryn Grasha 12 Germany connected to the structural and chemical
Jonathan Henshaw 1 14 
CalTech/IPAC, Pasadena, USA evolution of galaxies, and the build-up
Cinthya Herrera 16 15 
Ghent University, Belgium of their stellar mass. Understanding this
I-Ting Ho 1 16 
Institut de Radio Astronomie Milli­ process — its triggers, efficiency, key
Alexander Hygate 1 métrique, Saint Martin d’Hères, France timescales and dependence on environ-
Maria Jimenez-Donaire 23 17 
Universidad Autónoma de Madrid, ment — is a crucial goal of modern
Sarah Kessler 3 Spain astrophysics.
Jenny Kim 13 18 
University of California San Diego,
Ralf Klessen 22 La Jolla, USA Despite decades of intense observational
Eric Koch 9 19 
Observatorio Astronómico National and theoretical studies (for example, see
Philipp Lang 1 (IGN), Madrid, Spain the review by Kennicutt & Evans, 2012),
Kirsten Larson 14 20 
University of Toledo, Toledo, USA major questions remain. How do the
Alexandra Le Reste 8 21 
University of Wyoming, Laramie, USA properties of these GMCs depend on
Daizhong Liu 1 22 
ITA/ZAH University Heidelberg, their parent galaxy and on their location
Rebecca McElroy 1 Germany within the galaxy? How do the efficiency,
Joseph Nofech 9 23 
CfA/Harvard & Smithsonian, Cam- duration and output of star formation
Eve Ostriker 24 bridge, USA depend on the hosting GMC, the dynami-
Ismael Pessa Gutierrez 1 24 
Princeton University, USA cal environment in the galaxy, and the
Johannes Puschnig 11 25 
Johns Hopkins University, Baltimore, properties of the host galaxy? Do differ-
Miguel Querejeta 5, 19 USA ent clouds and different environments
Alessandro Razza 5, 6 26 
Space Telescope Science Institute, produce different cluster or stellar popu-
Toshiki Saito 1 Baltimore, USA lations? How do clouds evolve? Does star
Francesco Santoro 1 formation accelerate over time or pro-
Sophia Stuber 1 ceed at a steady pace? Which feedback
Jiayi Sun 3 A major advance in understanding the mechanisms dominate the destruction of
David Thilker 25 process of star formation will come clouds in which environments? How does
Jordan Turner 21 from charting the connections between stellar feedback shape the larger struc-
Leonardo Ubeda 26 cold (molecular) gas and young stars on tures of gas, metals, and eventually stars
Jose Utreras 5 the scale of individual molecular clouds, inside a galaxy?

36 The Messenger 177 – Quarter 3 | 2019

These remain open questions because 1.5 Figure 1. The PHANGS
addressing them requires surveys that All local galaxies
combine high resolution and high sensi- PHANGS ALMA targets
tivity. GMCs are the fundamental units of PHANGS HST all of the closest mas-
the cold interstellar medium (ISM). They sive, star-forming galax-
sit at the interface between the large-
focuses on a key subset
scale physics of galactic discs and the of these targets. Both
small-scale physics of star formation. ­s amples achieve good
With sizes of ~ 100 pc, studying these 0.5
coverage in the M-SFR*
plane, covering the
clouds requires high physical resolution.
log10 SFR (M๬ yr –1)

kinds of systems where

Over the last decade, observations of most stars form.
individual galaxies, including the Milky The galaxies targeted by
Way, have replaced the old idea of a uni- ALMA (blue filled
0.0 ­circles), MUSE (red cir-
versal population of GMCs with a more
cles), and HST (black
diverse, dynamical view of clouds. But squares) are shown with
these previous-generation studies have respect to the distribu-
been confined to individual galaxies, and tion of local galaxies.
– 0.5
often to small regions inside those galax-
ies. Before now it has simply been too
expensive to survey molecular clouds
across the whole galaxy population, and
as a result, we still lack a rigorous sta­ –1.0
tistical characterisation of GMC popula-
tions and properties across a represen­
9.0 9.5 10.0 10.5 11.0
tative sample of star-forming galaxy
log10 M෪ M๬ )
discs. Consequently, our understanding
of the link between GMC properties and
star formation is also in its infancy. The transformational power of ALMA of view, fantastic sensitivity, wavelength
and MUSE coverage, and sampling of the other
Our understanding of the time evolution stages of the star formation process,
of star formation regions has been simi- The advent of two ESO flagship facilities MUSE represents the perfect comple-
larly constrained by observations. Over dramatically changed the observational ment to ALMA.
time, an individual star formation region landscape in this field. The Atacama
evolves from a GMC to an exposed stellar Large Millimeter/submillimeter Array
cluster, with turbulence, gravitational (ALMA) can map cloud-scale CO (2–1) PHANGS ALMA and PHANGS MUSE
­contraction, and stellar feedback all play- emission — a key tracer of molecular observations
ing key but weakly constrained roles. gas mass and kinematics — over the
Again, the key physics is accessible only entire disc of a nearby (d ~ 15 Mpc) gal- Recognising this opportunity, the
from highly resolved observations. Here, axy in about two hours of main array PHANGS collaboration proposed ambi-
a multi-wavelength approach is required. time. This is an improvement of roughly tious observational campaigns that aimed
Observations of optical line emission two orders of magnitude in survey speed to use ALMA and MUSE to address the
from ionised gas are indispensable to compared to previous instruments, open questions in this field. The PHANGS
probe the HII regions created by young ­opening the transformational opportunity ALMA and PHANGS MUSE surveys sam-
stars, optical and ultraviolet observations to survey GMCs across the whole nearby ple the full time sequence of the star for-
probe the young stellar populations galaxy population. mation process at resolutions matched to
themselves, while longer wavelength individual clouds across a representative
observations measure the cold gas. To Meanwhile the Multi Unit Spectroscopic sample of local galaxies. By combining
capture the violent cycling between ISM Explorer (MUSE) at the VLT can capture resolution and physical detail using a sur-
phases, all of these observations must the full optical spectrum with the same vey type approach, these programmes
at least isolate individual star-forming resolution and field of view as ALMA. aim to link the detailed physics of star
regions. Again, these requirements have MUSE spectral maps reveal the location, formation to our understanding of galaxy
been so strict that most key work on this kinematics, and physical properties of evolution.
topic has so far been restricted to case HII regions, supernova remnants (SNe),
studies, mostly focused on the Milky Way and planetary nebulae (PNe). At the same
or Local Group targets. time, MUSE captures the underlying A representative sample of local
­stellar structures (spirals, bars, clusters) star-forming galaxies
and populations that represent the
dynamical drivers and outputs of the star Detailed studies of clouds, HII regions,
formation process. With its large field and cloud evolution have so far

The Messenger 177 – Quarter 3 | 2019 37

Astronomical Science Schinnerer E. et al., The PHANGS Surveys

102 corresponds to this key cloud scale. At

this resolution, both instruments can still

cover an area that includes most of

Antennae the massive star formation in each target.

Reaching these GMC scales in nearby

galaxies allows us to connect detailed

Galactic studies of Milky Way clouds to
PHANGS global galaxy properties, to make meas-
Velocity dispersion (km s –1)

+ M51 urements that test theories and numerical

prescriptions for star formation and feed-
back, and to resolve the time evolution of
M31 & M33 the ISM.
Resolving the star-forming units in
nearby galaxies

Molecular clouds across galaxy discs


PHANGS ALMA resolves the molecular

gas reservoir into individual GMCs
across the full disc in ~ 80 targets. The
survey focuses on mapping CO(2–1) line
10 0 emission, which arises from the cold,
molecular gas that forms stars. When
10 0 101 102 10 3 104 complete, PHANGS ALMA will character-
ise more than 100 000 massive GMCs,
Surface density ∑ (M๬pc – 2 )
roughly 100 times the number of clouds
known in the Milky Way. Key science
Figure 2. Molecular clouds in PHANGS ALMA. Line easily observable with ALMA and MUSE goals of PHANGS ALMA include:
width at a fixed size scale (120 pc) as a function of
(–75 < declination < 20 degrees). These
gas surface density for the first ~ 10 PHANGS ALMA
targets (blue) and several literature targets including criteria yield ~ 80 galaxies (see Figure 1) A. Uncover the dependence of molecular
the Antennae galaxies (NGC 4038/39). The figure, that sample the local M* –SFR relation cloud populations on host galaxy and
adapted from Sun et al. (2018), shows the properties very well. local galactic environment.
of ~ 30 000 individual cloud-scale measurements
Case studies of cloud properties in
(i.e., the equivalent of ~ 30 000 GMCs). The clouds
span a wide range of surface density, line width, and individual galaxies have already indi-
internal (turbulent) pressure, but a relatively modest Cloud-scale surveys cated that the distribution of cloud
range of dynamical state (ratio of kinetic (K) to poten- mass and the corresponding gravita-
tial (U) energy). The visible variations in cloud proper-
The key physics described above tional state are changing as a function
ties can be mapped back to the locations of clouds
inside the galaxy and the properties of the host galaxy. plays out at the cloud scale, which is of environment. The statistics provided
50–150 pc. This is about the scale by PHANGS ALMA will not only allow
been mostly restricted to individual case height of the cold gas disc, about the for a systematic assessment of these
­studies. Gas content, galaxy growth, and scale at which supernovae are expected changes, but also identify the param­
­star formation are now understood to to stir turbulence, and about the radius eters that set or control the structure
vary in important ways as a function of of a massive GMC. Current models of of the cold ISM. Figure 3 shows
stellar mass (M*) and specific star forma- star formation and feedback predict a results from a pilot study using our first
tion rate (sSFR). A key goal of PHANGS is deep link between GMC properties, star ~ 10 targets (Sun et al., 2018). These
to link the detailed physics of star forma- formation, and feedback, setting the already reveal strong, systematic varia-
tion to our understanding of galaxy evolu- ­conditions for star formation to occur, its tions in the physical state of molecular
tion. To achieve this, both surveys aim to efficiency and duration. Violent cycling gas across the galaxy population.
sample the so-called main sequence of star- between stages of the star formation pro-
forming galaxies, i.e., the M* –SFR corre- cess also becomes visible at roughly this B. Measure the efficiency of star forma-
lation that persists across redshift. To resolution. tion by comparing the rate at which
this end, PHANGS targets all massive molecular gas forms stars to the gravi-
(9.5 < log(M* /M⊙) < 11.0), actively star- ALMA and MUSE are efficient survey tational freefall time.
forming (log(sSFR/yr –1) > –11) galaxies instruments thanks to their ~ one arcsec- This efficiency per freefall time is a
within D ~ 17 Mpc that are not too ond resolution. We designed the sample central parameter in many star forma-
edge-on (I < 75 degrees) and which are to be close enough that this resolution tion theories. It captures the degree

38 The Messenger 177 – Quarter 3 | 2019

2.5 MW (Lee+ 16)
MW (Evans+ 14)
MW (Vutisalchavakul+ 16) 49:00.0
M51 (Leroy+ 17) CMZ (Barnes+ 17)
(Schruba+ 18b) 30.0

Normalised histogram

60 pc beam

80 pc beam 15:47:00.0

100 pc beam
120 pc beam


This work
0.5 500 pc

52.0 50.0 48.0 46.0 44.0 42.0 1:36:40.0 38.0 36.0 34.0 32.0
0.0 Right ascension
– 3.2 – 3.0 – 2.8 – 2.6 – 2.4 – 2.2 – 2.0 –1.8 –1.6 –1.4 –1.2 –1.0 – 0.8
log10 (εff )1.3 kpc

Figure 3. The distribution of star formation efficiency The dominant feedback mechanism Figure 4. PHANGS ALMA and MUSE observations
per freefall time eff. The fraction of gas mass con- show the evolution of star-forming regions. ALMA
and timescale for cloud dispersal in
verted to stars per gravitational freefall time in the observations of the cold gas reservoir (blue; CO [2–1])
first ~ 10 PHANGS ALMA targets (from Utomo et al., different environments are still overlaid with the PHANGS MUSE view of ongoing
2018). This efficiency per freefall time eff is a key unknown and will be constrained via star formation activity (orange; Ha) in the the nearby
benchmark for theory, capturing the inefficiency of modelling. spiral galaxy NGC628 (Kreckel et al., 2018). The
star formation relative to gravitational collapse. It molecular clouds seen in blue often appear visibly
is uniquely accessible to PHANGS ALMA because offset from the HII regions created by massive young
the high-resolution ALMA imaging allows us to esti- At the time of writing, the observations stars. As gas flows through the spiral arms, cold,
mate the mean density of the molecular gas. for the Large Programme that forms the dense clouds visibly evolve into young star-forming
core of PHANGS ALMA are almost regions. A key goal of PHANGS is to use statistical
and dynamical modeling of these data to constrain
to which star formation is slowed or ­complete. All calibrated data products,
the life cycle of molecular clouds.
curtailed by feedback, turbulence, and cloud catalogues, and a host of other
other means and represents a specific high-level data products are expected to
prediction of many models. Because be released to the community in 2020. roles played by spirals, bars, and central
this measurement requires knowing Science papers addressing our key goals stellar structures revealed by their stellar
the gas density, it also requires high- have already begun to appear. population distributions and kinematics.
resolution imaging. Already, using Resolving star-forming regions and stellar
our first ~ 10 galaxies, PHANGS ALMA discs with PHANGS MUSE (which started
has provided the most definitive meas- Star-forming regions and stellar as a Large Programme in ESO Period 100)
urement of this quantity to date in ­structures with PHANGS MUSE will provide us with a dynamic view
­normal local disc galaxies (Utomo et of star formation, stellar feedback and
al., 2018). Star formation and feedback are violent, chemical enrichment of disc galaxies,
rapid processes, with key roles played allowing us to address the following:
C. Quantify the life-cycle of molecular by ionising radiation, radiation pressure,
clouds. stellar winds, and supernovae — all A. Estimate the timescales of the star
At high resolution, star-forming regions ­phenomena traced through the ionised ­formation process. This is closely
appear in discrete evolutionary states gas phase. These sources of feedback related to modelling the life cycle of
using ALMA and MUSE – that is as impact, destroy, and reshape the cold molecular clouds. MUSE and ALMA
clouds, HII regions, and young star gas that will form the next generation working together (illustrated in Fig-
clusters (beautifully demonstrated in of stars. Meanwhile, the young stars ure 4, adapted from Kreckel et al.,
Figure 4, adapted from Kreckel et al., deposit new metals into the ISM, while 2018) offer the chance to address
2018). Via statistical and dynamical new material for future star formation open questions: How long does star
modelling, such observations con- must flow in through the galaxy disc. Star formation take to set in once a molec-
strain the evolution from diffuse gas to formation itself may be shaped by the ular cloud has formed? How long
dense clouds to HII regions to clusters. underlying stellar potential, with important does stellar feedback take to disperse

The Messenger 177 – Quarter 3 | 2019 39

Astronomical Science Schinnerer E. et al., The PHANGS Surveys

8.450 8.475 8.500 8.525 8.550 8.575 8.600 8.625 8.650

– 59°13ಿ30.0ೀ

Declination (J2000)





46m00.00s 50.00s 40.00s 4h45m30.00s

Right ascension (J2000)

Figure 5. The PHANGS MUSE view of starlight, ion- and out of galaxies, and the redistribu- The PHANGS HST Large Programme
ised gas and metals. PHANGS MUSE observations
tion of metals from their birth sites to builds on the successful Legacy Extra­
of NGC1672. Left: Simulated gri three-colour image
derived from MUSE, showing starlight and prominent the surrounding medium. Galactic Ultraviolet Survey (LEGUS)
dust lanes. Middle: Emission line maps (red: Ha, Treasury programme, using a similar
green: [OIII], blue: [SII]) revealing active star formation PHANGS MUSE targets 19 nearby star- five-filter observing strategy with the Wide
and diverse physical conditions. Right: For each
forming galaxies that are also targets Field Camera 3 (WFC3) camera.
HII region in the galaxy, we use the MUSE spectra to
estimate the metallicity. Even before removing of PHANGS ALMA. We expect to identify PHANGS HST expands from LEGUS
the radial gradient, spatial correlations in metallicity, and characterise ~ 15 000 resolved towards more massive (more Milky Way-
including azimuthal variations, are apparent. star-forming regions, measuring metallic- like) galaxies and focuses on regions
These trace the diffusion of metals and the flows of
ity and other physical diagnostics for ­covered by ALMA and MUSE (see Fig-
enriched and pristine gas across the galaxy.
each. MUSE also captures a host of PNe ure 6). When completed, it will cover
and SNe remnants (for example, Kreckel 38 disc galaxies. We expect the high-­
a cloud? Is star formation triggered et al., 2017) and yields exquisite stellar resolution ultraviolet and optical imaging
within a galactic disc (i.e., in spiral and gas kinematics, offering signatures of from PHANGS HST to yield catalogues
arms) or is it a stochastic process? stellar feedback and (with ALMA) the of 100 000 star clusters and associations.
­relative motion of different phases of the
B. Quantify the strength of stellar feed- ISM. Full spectral fitting of the MUSE Combining HST with ALMA and MUSE
back across environment and scale. datacubes will produce “movies” of stellar will dramatically improve our answers to
Stellar feedback comes in many forms, mass in several bins of ages and metallic- the questions listed above (for example,
including ionising radiation, radiation ities. The survey is currently approxi- from much improved knowledge of the
pressure, stellar winds, and supernova mately three-quarters done, with a full young stellar population) and also con-
explosions. The high physical reso­ release of major data products expected strain: (a) the timescale for the removal of
lution and wide wavelength coverage by 2021. gas from young stellar clusters (YSC); (b)
provided by MUSE allow us to sepa- the relation between the YSC and GMC
rate and characterise HII regions, mass functions; (c) the link between
SNe remnants and PNe (for example, PHANGS beyond ALMA and MUSE cloud properties and the fraction of stars
Kreckel et al., 2017, apply this method formed in clusters; and (d) the connection
to our pilot data). Together with Young stars and stellar clusters with between the multi-scale structure visible
the ALMA data, the MUSE cubes can PHANGS HST in young starlight and cold gas structure.
directly assess the interactions The release of joint HST-ALMA data
between the warm gas (10 000 K), the Direct observations of young stellar clus- products revealing the detailed properties
cold gas (< 100 K), and the young stel- ters offer a powerful complement to of stellar clusters and gas clouds will
lar population. the observations of clouds by ALMA and be a key part of the PHANGS legacy and
those of HII regions and integrated star- will help lay the scientific groundwork
C. Measure the diffusion of metals and light by MUSE. These clusters have for future facilities like the James Webb
chemical enrichment. The details of ­t ypical sizes of a few parsecs, and so Space Telescope (JWST).
metal production and diffusion remain observing them requires the resolving
major open questions in the chemical power of the Hubble Space Telescope
evolution of galaxies. With metallicity (HST). As of 2019, PHANGS HST (Princi- Other major efforts
estimates for ~ 15 000 HII regions pal Investigator: Janice Lee, Infrared
across 19 galaxies, PHANGS MUSE ­Processing and Analysis Center [IPAC] at Building around the core ALMA, MUSE,
gives us a powerful handle on how Caltech, USA), has been observing and HST samples, the PHANGS team
metals are built and distributed. Pat- an overlapping sample with the goal of is committed to building a complete char-
terns in the small-scale distribution ­connecting young star clusters to the acterisation of the stellar, gas, and kine-
of metals (see Figure 5) constrain the cold and warm gas traced by ALMA and matic structure of our sample. Key efforts
flows of pristine and enriched gas in MUSE. include: (1) making high quality maps of

40 The Messenger 177 – Quarter 3 | 2019

the stellar mass distribution for each scale data set, with a focus on moving Figure 6. HST and ALMA probing the life-cycle of
star formation regions. PHANGS HST and PHANGS
­target using Spitzer and Wide-field resolved studies of nearby galaxies into
ALMA observations of the nearby spiral galaxy
­Infrared Survey Explorer (WISE data); the “big data” regime. The team is also NGC1559. The high angular resolution afforded by
(2) assembling new and archival nar- running dedicated state-of-the-art simu- HST allows for the identification of individual young
row-band imaging of the Ha emission line lations. Our observations capture the star clusters that have just emerged from their
birth cloud as seen by ALMA. Optical light (HST red:
for the whole PHANGS ALMA sample; underlying physical processes filtered
white light, green: H-band, blue: near ultraviolet; left),
(3) using the Indian Space Research through a complex combination of pro- with the molecular gas distribution added in red
Orga­nization (ISRO) AstroSat satellite to jection, chemistry, and radiative transfer. (middle). The distribution of molecular gas (ALMA;
obtain high-angular-resolution imaging Implementing a realistic forward model- right) is remarkably similar to that of the dust lanes
seen in optical light.
of far-­ultraviolet emission for a subset of ling perspective into the PHANGS theo-
­targets; (4) new and archival Very Large retical efforts is therefore key to both
Array observations of the 21-cm line uncovering the underlying physics that as leading follow-up proposals and
to trace the atomic gas reservoir; (5) con- drives star formation and constructing ­planning the next generation of PHANGS
struction of detailed environmental masks; new innovative tracers. projects.
(6) reprocessing, fitting, and analysis of
archival infrared (Spitzer, Herschel, WISE)
and ultraviolet (Galaxy Evolution Explorer, A modern scientific collaboration Acknowledgements
GALEX) observations; and (7) obser­ We would like to thank all the staff at ESO, JAO, and
vations of high critical density tracers of The PHANGS collaboration brings NRAO who have been supporting the preparation,
“dense gas” using ALMA, the Institut de together experts on ISM physics, dynam- conduction and delivery of these unique datasets,
radioastronomie millimétrique (IRAM) ics, stellar populations, and galaxy and more generally, devoting efforts to make such a
challenging observational campaign a reality.
facilities, and the Green Bank Telescope. ­evolution. It includes expertise distributed
These efforts leverage a diverse, distrib- across different wavelengths and com-
uted team and promise to pair the bines both observational and theoretical References
unprecedented ALMA and MUSE data points of view. PHANGS started in
Kennicutt, R. C. & Evans, N. J. 2012, ARA&A, 50,
with the most complete view of stellar 2015 with a small group of enthusiasts 531
and gas structure for any sample to date. ­dedicated to seizing the opportunities Kreckel, K. et al. 2017, ApJ, 834, 174K
described above. Today, PHANGS is Kreckel, K. et al. 2018, ApJ, 863L, 21K
a medium-sized collaboration distributed Sun, J. et al. 2018, ApJ, 860, 172S
Utomo, D. et al. 2018, ApJ, 861L, 18U
Synergy of observations and theory around the globe (with substantial rep-
resentation in Australia, Europe, Chile,
The combination of ALMA, MUSE, HST and North America). Links
and the deep supporting observations 1
 HANGS survey webpage:
represents a complex, high-dimensional The team is committed to diversity. We
data set. The full exploitation of such a aim to fill scientific leadership roles with
unique resource requires a close synergy a mixture of junior and senior scientists
between careful observational analysis, and to enhance the visibility of female
modelling and statistical analysis tech- scientists. Currently about half of all lead-
niques, numerical simulations, and ana- ership positions are occupied by female
lytic theory. With this in mind, PHANGS scientists. Following good practice of
has pursued development of new tools other large collaborations, “builder sta-
and close comparison with numerical tus” has been granted to junior scientists
modelling and theory as a core activity. who have spent significant time gene­
The development of new statistical tools rating data products for use by the astro-
focuses on robust statistical characteri- nomical community. Junior scientists
sation of the full multi-dimensional, multi- have also taken on high profile roles such

The Messenger 177 – Quarter 3 | 2019 41

Astronomical News

Construction has started on the foundations

of ESO’s Extremely Large Telescope (ELT) at
Cerro Armazones. Once complete, the telescope
will be the largest ground-based telescope in
­operation, weighing 3400 tonnes.

On 28 August 2019, ESO, MetricArts and

Microsoft Chile received the 2019 award for
“Digital Transformation and Industry 4.0” from
the Chilean Association of Information Technology
Mónica Solis

Companies (ACTI). This was in recognition of

their efforts to integrate artificial intelligence into
operations at Paranal Observatory.

42 The Messenger 177 – Quarter 3 | 2019

Astronomical News DOI: 10.18727/0722-6691/5152

Total Solar Eclipse Over La Silla

Laura Ventura 1 The solar eclipse of 2 July 2019, whose outline of the eclipse project was drafted
Claudio Melo 1 path of totality included La Silla Obser­ in 2013, and a project plan for a big
Lars Lindberg Christensen 1 vatory, provided a stunning culmination ­public event was prepared and approved
Mariya Lyubenova 1 to the celebrations of the 50th anniver- by the ESO Directors Team in June 2018,
Fernando Comerón 1 sary of La Silla, ESO’s first observatory, one year before the event.
which was inaugurated on 25 March 1969.
The thousands of visitors who came to The La Silla site provides a stunning
ESO La Silla for the eclipse were only a small ­setting, combining an astronomical obser-
fraction of the hundreds of thousands vatory and a beautiful landscape; in
of people who travelled from all over the ­addition, given the high likelihood of good
On Tuesday 2 July 2019, in the late world to the narrow strip of land that observing conditions during the event,
afternoon, a total solar eclipse took place would find itself under the shadow of the it became a desirable location to enjoy
over ESO’s La Silla Observatory; totality Moon for almost two minutes. the eclipse. For that reason, ESO decided
lasted 1 minute and 52 seconds. For to open the observatory to as many visi-
this very special event, ESO decided to By 2011, ESO’s education and Public tors as the infrastructure and logistics
open the doors of the observatory to Outreach Department had already would allow.
the public, providing over one thousand received the first requests from the public
visitors with a unique vantage point to come to La Silla to witness the 2019 Figure 1. La Silla during the total solar eclipse,
from which to witness this spectacular eclipse and/or to take pictures, videos which resulted in almost two minutes of totality at
natural phenomenon. and make precise measurements. An 20:39 UT.
ESO/R. Lucchesi

The Messenger 177 – Quarter 3 | 2019 43

Astronomical News Ventura L. et al., Total Solar Eclipse Over La Silla

Figure 2. The President

ESO/M. Zamani

of the Republic of
Chile, Sebastián Piñera,
pictured with Chilean
high school students
while visiting La Silla.

In addition to the general public, La Silla past 50 years, ESO had the pleasure of is difficult even from space-based probes.
was honoured to welcome the President inviting a large group of students and La Silla hosted an array of scientific and
of the Republic of Chile, Sebastián Piñera senior citizens from the municipalities of outreach observations carried out by
and his wife, Cecilia Morel. They were La Higuera, La Serena, and Coquimbo teams that used dedicated equipment to
accompanied by the Ministers of Science, to witness the eclipse from La Silla. The perform a variety of observations,
Technology, Knowledge and Innovation, group, who had a special meeting with described in Christensen et al. on p. 47.
Andrés Couve, and of Education, Marcela President Piñera and the First Lady, Notably, even the NTT was used to obtain
Cubillos, as well as by the Undersecret­ engaged in a lively conversation with them. spectroscopy of the solar corona, an
aries of Foreign Affairs, Carolina Valdivia, In addition to the students from the observation that critically depended on
and of Tourism, Mónica Zalaquett, along Coquimbo Region, La Silla also hosted the perfect synchronisation — a miscalcula-
with other authorities from the Ministries sixteen winners of a contest organised tion could have resulted in direct expo-
of Science and Foreign Affairs and by CONICYT, the National Council of Sci- sure to the radiation of the solar surface
­members of the National Congress. ence and Tech­nology of Chile, among and severe damage to the telescope
Although the President and the First Lady children selected from schools across the optics and instrument. The experience of
could not stay at La Silla for the eclipse, country. the telescope operators ensured that this
they had the opportunity of visiting the was expertly avoided (see Dennefeld
3.6-metre telescope, where they were Most of the members of the ESO Science et al., p. 54). Another notable experiment
received by the Director General of ESO, Outreach Network (ESON), ESO’s network included the use of the Rapid Action
the President of the ESO Council, and of outreach representatives in the Member Telescope for Transient Objects (TAROT)
other members of the ESO management. States and beyond, were present at to reproduce the famous Eddington
After this visit, both the President of La Silla as well. For them, the eclipse was experiment during the historical eclipse of
Chile and the ESO Director General gave the highlight of a tour of the ESO facilities 1919, when the deflection of the light
speeches at a tent set up to host over in Chile that included Paranal and the from stars near the line of sight of the
60 media representatives — both national Atacama Large Millimeter/­submillimeter Sun was used to verify the predictions of
and international — who had registered Array (ALMA). Other members of this tour Einstein’s general theory of relativity.
to provide live coverage of the event. group included eight social media influ-
They then moved to the VIP area, by the encers selected from 300 participants in Aside from the special guests mentioned
New Technology Telescope (NTT), where a #MeetESO social media competition 1 above, over 700 people travelled to
they met with diplomatic representatives including the winner of the La Silla Total La Silla. Around a tenth of these visitors
of ESO Member States and Partners, and Eclipse Public Com­petition 2. came from Chile, with the rest travelling
other distinguished guests. from abroad — mostly Europe and
A total solar eclipse offers a rare possi­ North America. An area with a wonderful
As a symbol of the deep appreciation of bility to carry out scientific experiments view of the eclipse was prepared for the
ESO toward the local community which ­targeting the physical properties of the public and was equipped with a large tent
has hosted its first observatory for these lower solar corona, observation of which with seats, where snacks and beverages

44 The Messenger 177 – Quarter 3 | 2019

Figure 3. Some of the 1000 visitors to La Silla are

S. Lowery
seen here observing the eclipse from an area near
the Swedish–ESO Submillimetre Telescope (SEST),
during the performance by Steve Rothery and friends.

were served. Polyclinic services were

installed at the former control building of
the Swedish–ESO Submillimetre Tele-
scope (SEST), and two ambulances fully
equipped to provide first aid were posi-
tioned along the public viewing area. Infor-
mation on the safe viewing of the eclipse
had been given in advance to all those
attending, and glasses with filters certified
for solar viewing were distributed.

As part of the programme of activities

offered for all the visitors, solar tele-
scopes were set up for public observa-
tions before and during the partial phases along, resulting in a truly impressive dis- band, appropriately entitled “La Silla”.
of the eclipse. The visitor centre at play of telescopes, cameras and a wide The performance was a collaboration
La Silla offered a brand-new exhibition variety of imaging aids. with Rick Armstrong, who played bass
about ESO, including informative panels, guitar. It was an honour to host Rick
the simulation of an old control room and ESO made sure that the spectacle ­Armstrong, the elder son of astronaut
a collection of astronomical instrumenta- enjoyed at La Silla could be followed Neil Armstrong, particularly in the same
tion pieces rescued from the observatory across the rest of the world, offering a live month as the 50th anniversary of the
over the course of its 50-year history, webcast that combined high-resolution first lunar landing.
showing the evolution of detectors from images of the eclipsed Sun with views of
photographic plates to CCDs. Two itiner- the site from several vantage points, con- Despite the expectation of the cold tem-
ant exhibitions were also installed for the veying the atmosphere at the observatory peratures typical in July at La Silla, and
occasion: the already very popular audio- (see Figures 3 & 4). In addition, a team of the non-negligible probability of adverse
visual show “ALMA sounds”, created from expert photographers from ESO ensured meteorological conditions, the weather
millimetre-submillimetre signals detected that excellent images of high technical on the eclipse day was especially compli-
by ALMA; and paintings by Chilean artist and artistic quality were obtained, a sam- ant, adding to the special character of the
Silvana Zúñiga that, using luminescent ple of which are shown in this article. event. The sky remained clear and cloud-
paint, illustrate concepts like light pollu- free for the whole day, temperatures were
tion and multi-wavelength astronomy. A special treat was a concert by British mild and virtually no wind blew, making
The visitor centre also hosted public talks musician Steve Rothery (of Marillion fame) the outdoors experience even more
given by ESO experts on popular astron- and his band, who performed against the pleasant and memorable.
omy topics. Finally, public tours to the impressive backdrop of the Swedish–
NTT and the 3.6-metre telescope were ESO Submillimetre Telescope (SEST)
Figure 4. The moment worth waiting for: totality.
organised. The area reserved for public antenna. The concert ended with the pre- With the Sun only 14 degrees above the horizon, day
viewing offered ample space to set up miere of a new record by Steve R­ othery turns into twilight. The VIP platform at the NTT is
the equipment that many visitors brought and Riccardo Romano, a member of his seen on the right.
ESO/M. Zamani

The Messenger 177 – Quarter 3 | 2019 45

ESO/P. Horalek
Astronomical News Ventura L. et al., Total Solar Eclipse Over La Silla

Figure 5. A composite photo with exposures of Moon completely covering the disc of Contemplating one of the most majestic celestial
v­ arying duration showing filamentary details in the phenomena was enhanced spectacularly by excel-
the Sun. The corona appeared bright and
solar corona. lent meteorological conditions and the wonderful
compact, typical of the period near mini- landscape of ESO’s first observatory. While those
mal solar activity, and, through the tele- particular factors were beyond ESO’s control, the
As the partial eclipse phase started and scopes, a protuberance could be seen success of the experience for visitors on that day
required planning, coordination and organisation
progressed during the early afternoon, an toward the northwest of the solar disc.
and the effort of colleagues across ESO. The
atmosphere of mounting excitement built Most of those witnessing the phenome- authors would particularly like to thank the staff in
up in anticipation of the extraordinary non reacted emotionally, with abundant the Department of Communication, La Silla Logistics
moments to come. People witnessed the exclamations of wonder, hugs and tears. and operations, IT support, ESO’s Representation in
Chile, La Silla Paranal Safety, and everybody else
progress of the covering of the solar disc As the first Baily’s beads appeared, sig-
who worked hard in the preparation and execution of
with dedicated protective glasses at the nalling the end of the total eclipse phase, a unique event in ESO’s history.
beginning, but as totality approached, the the 1 minute and 52 seconds of totality
changes in illumination of the surrounding seemed to many way too short.
landscape became increasingly obvious,
giving the scene an unreal appearance. The deep impressions left by totality 1
#MeetESO social media competition: https://www.
The blue of the sky became deeper and almost turned the last partial eclipse
deeper while a band of intense orange — phase — during which the solar disc pro- 2
The La Silla total eclipse public competition:
often seen in the middle of twilight — gressively reappeared from behind the
encompassed the entire horizon. At the Moon — into an anti-climax, despite the 3
ESO press release for the total solar eclipse:
same time, a noticeable drop in the tem- beautiful show put on by the partially
perature was felt as the solar irradiation eclipsed Sun advancing toward the horizon
decreased. until the end of the eclipse near sunset.
Further impressions of the event can be
Finally, the shadow of the Moon reached seen online2. a
 his is the only total solar eclipse that will be visible
La Silla. In a matter of seconds, the sliver from an ESO observatory for more than 212 years,
when La Silla will be close to the north of the path
of the uncovered solar disc thinned until
of totality during the total solar eclipse of 28 August
disappearing as the sky quickly dark- Acknowledgements
2231. An annular eclipse will be visible from La Silla
ened, stars became visible, and the solar The authors feel privileged to have witnessed this
next century, on 9 January 2187.
corona shone in stark contrast with the once-in-a-lifetime experience at La Silla, as the next
surrounding sky and the black disc of the total solar eclipse visible from there will be in 2231a.

46 The Messenger 177 – Quarter 3 | 2019

Astronomical News DOI: 10.18727/0722-6691/5153

Science & Outreach at La Silla During the

Total Solar Eclipse

Lars Lindberg Christensen 1

Gerardo Ávila 1
Wahab A. Baouchi 2
Michel Boer 3
Jean-François Le Borgne 4
Christian Buil 5
Manuel Castillo-Fraile 6
Eric Denoux 7, 5
Valérie Desnoux 5
David Elmore 8
Loic Eymar 3
Robert F. Fisher 9
Carlos Guirao 1
Alain Klotz 4, 10
Adrien Nicolas Klotz 10
Julien Lecubin 11
Kyle A. Motl 9
Darío Pérez 12
Miguel Pérez-Ayúcar 13
Wouter van Reeven 13
Xavier Regal 14
Yoann Richaud 14
Rico Sautile 14
Alexandre Santerne 15 17 
NASA Solar System Ambassador, Figure 1. Some members of the ESA/CESAR team
at their observing spot. From left to right: Manuel
Roy Wellington 16 Nashville, USA
Castillo, Wouter van Reeven, Miguel Pérez-Ayúcar,
Theo Wellington 16, 17 18 
ESA (European Space Agency), Joe Zender and Darío Pérez de Carlos.
Padma A. Yanamandra-Fisher 9 Noordwijk, the Netherlands
Joe Zender 18
Along with cameras and telescopes
Total solar eclipses are rare phenomena, brought by 700 members of the public
ESO only occurring in a specific location and more than 60 journalists, five differ-
University of Colorado, Boulder, USA once every 360 years on average. His- ent groups conducted outreach, educa-
ARTEMIS-CNRS/OCA/UNS, Nice, torically, total solar eclipses have only tion and science experiments on the
France been observed twice from large pro­ day2. They were assigned a place on the
IRAP-Observatoire Midi Pyrénées, fessional observatories, allowing spe- side of the La Silla mountain just below
Toulouse, France cific science experiments to take place. the Visitor Centre (formerly called the Ritz
Association AUDE, Paris, France On this occasion, ESO invited nearly building) as well as on the New Technol-
Serco for ESA (European Space 25 scientists, communicators and ogy Telescope (NTT) platform. The two
Agency), Madrid, Spain ­educators to observe and document 4-metre-class telescopes at La Silla were
Observatoire Cor Caroli, Caussade, the eclipse and benefit from La Silla’s also pointed near the Sun during the
France clear skies and its infrastructure eclipse. The observations using the New
Association of Universities for Research and resources. This article presents an Technology Telescope (NTT) are
in Astronomy (AURA), Washington D.C., overview of these various activities. described in this issue of the Messenger
USA (Dennefeld et al., p. 54). The eclipse was
The PACA Project, Space Science also captured with a small solar tele-
Institute, Boulder, USA Introduction scope called the HARPS Experiment for
Université Paul Sabatier, Toulouse, Light Integrated Over the Sun (HELIOS)
France It is very rare that a total solar eclipse which is installed on the catwalk of the
OSU PYTHEAS, Marseille, France passes over an existing observatory with ESO 3.6-metre telescope, and fed into
GTlinkers, Madrid, Spain large telescopes — in fact, in the last fifty the High Accuracy Radial velocity Planet
Aurora Technology for ESA (European years there have only been two such Searcher (HARPS) instrument via a fibre.
Space Agency), Madrid, Spain opportunities: in 1961 over l’Observatoire The high-precision spectroscopic data
Observatoire de Haute Provence/OSU de Haute-Provence in France, and in look promising and are currently under
PYTHEAS, Saint Michel l’Observatoire, 1991 over Mauna Kea on the island of analysis. Two national telescopes,
France Hawai’i. A separate article in this issue of Exoplanets in Transits and their Atmos-
Aix-Marseille University/CNRS/CNES/ the Messenger (Ventura et al., p. 43) pheres (ExTrA) and the Rapid Eye Mount
LAM, Marseille, France provides an overview of the total solar telescope (REM), also attempted obser-
Barnard-Seyfert Astronomical Society, eclipse event at La Silla (also see ESO vations but encountered technical
Nashville, USA press release1). problems.

The Messenger 177 – Quarter 3 | 2019 47

Astronomical News Christensen L. L. et al., Science & Outreach at La Silla During the Total Solar Eclipse

The team also completed most of its sci-


entific objectives; at the time of writing

this was the status of the (ongoing)
– Polarisation measurements were com-
pleted using the Eclipse K-corona
POLarimeter (EKPOL) from the Turin—
INAF Observatory using two different
setups; first, the polarisation intensities
of the corona were measured using
observations at four different polarisa-
tion angles (0, 45, 90 and 135 degrees).
From these measurements, the polari-
sation brightness is measured and the
electron density in the corona derived.
The second objective was to obtain the
polarisation intensities at more polarisa-
tion angles to decrease the overall
uncertainty in the computation of the
polarisation brightness. EKPOL (see
Figure 2. This image is a combination of polarised tele­scope. These were sent live (at a rate Zangrilli et al., 2009) is based on an
images obtained during totality to bring out the
of two per minute) and posted online4 as optical telescope supplemented by an
details of coronal structures.
soon as they were processed through electronic controllable liquid-crystal var-
the servers at the European Space iable retarder together with a specific
ESA/CESAR Activities Astronomy Centre (ESAC) in Madrid. CCD camera (funded by the ESA Fac-
–L  ive streaming on YouTube5 was carried ulty). EKPOL was developed as a tech-
The primary goal of the European Space out with a 1-metre focal length tele­ nology demonstrator for the Metis coro-
Agency (ESA) project Cooperation through scope and a Sony Alpha 7 SII camera. nagraph on Solar Orbiter (INAF) and the
Education in Science and Astronomy –A  live Google hangout6 included live Association of Spacecraft for Polarimet-
Research (CESAR) was to carry out sci- connections to the La Silla team and ric and Imaging Investigation of the
entific observations of the solar atmos- talks and presentations related to solar Corona of the Sun (ASPIICS) on Proba-3
phere and the Earth’s ionosphere, as well projects and science. This was coordi- (ESA/Royal Observatory of Belgium).
as general observations for outreach nated from the CESAR facilities at
and education. The results obtained to ESAC enabling a professional multime-
date are summarised below and online3. dia p ­ roduction (including mixing live
video streams from different places, live
The CESAR teama (Figure 1) completed images of the eclipse, presentation
the following education and outreach slides, etc.). Figure 3. A composite image of the solar eclipse
made from ground and space observations: Proba-2
objectives: – In the days following the eclipse, ESA (SWAP) solar disc, CESAR corona in grey, SOHO/
– Still images of the inner corona were published images from the event on the LASCO C2 outer corona in red, SOHO/LASCO C3
obtained with a 1-metre focal length main ESA webpages7 (see Figures 2–5). extended corona in blue.
Proba-2: ESA/Royal Observatory of Belgium

48 The Messenger 177 – Quarter 3 | 2019

kins University Applied Physics Lab

(simple photodiode sensors equivalent

to those flown on the Mars Science
Laboratory Curiosity and the Beagle 2
lander). Additional low-cost ultraviolet
and visible flux measurements were
carried out, including measurements of
a decline in all-sky brightness during
totality — in a silicon solar cell, the
brightness declined to < 0.02% of post-
eclipse values.

The TAROT observations

One century ago, on 29 May 1919,

a seminal experiment led by Arthur
Eddington confirmed Albert Einstein’s
prediction that light can be deflected by
mass. Eddington used the Sun as a mass
deflector and nearby stars as the targets
with which to measure the light deflec-
tion. The experiment becomes feasible
during a total solar eclipse because sky
brightness drops to twilight levels and
Figure 4. The Sun’s chromosphere and prominence lengths from 4000 to 10 000 Å, allowing stars close to the Sun can be observed.
taken with a visible-light telescope (~ f/1000), with no
the identification of features that are
filter and using a Canon 550D, during the Moon’s
exit (third contact). present in the photosphere but not in In 1915, Einstein predicted that the
the chromosphere. apparent position of a star is shifted radi-
– Independent polarisation measure- –T he CESAR team also completed the ally away from the centre of the Sun. He
ments (Figure 2) were completed with first ever measurements of the Earth’s predicted an angular displacement of
simple polarisation filtering of white light ionosphere using a Galileo receiver to 1.751 arcseconds when a star grazes the
(with improved equipment and proce- try to record changes caused by the Sun’s limb (twice the amount expected
dures derived from experience gained transit of the Moon’s shadow over the by Newtonian mechanics). This displace-
during the total solar eclipse in the area of observation. The analysis is ment decreases as the inverse of the
USA). A filter wheel with four polarisa- ongoing; by using multi-band and multi-­ distance to the centre of the solar disc.
tion angles (0, 45, 90 and 135 degrees) constellation data from the Global
was used. A fifth filter position without a ­Navigation Satellite System (GNSS), it Télescope à Action Rapide pour les
polariser allowed the capture of unpo- is possible to analyse the total electron Objets Transitoires (TAROT; see Figure 6)
larised images. The objective of this density perturbations with enough time is a robotic telescope that was installed
experiment was to calculate the differ- resolution to reveal ionospheric irregu- at ESO’s La Silla Observatory in 2006
ent polarisation rates in the inner solar larities during the eclipse. For this pur- (Klotz et al., 2013). The primary goal of
corona to obtain estimates of the elec- pose, the team has a close collabora- the TAROT teamb observations was to
tron content and the magnetic tion with a GNSS research group at the repeat the Eddington Experiment. The
field. These estimates, together with Universitat Politècnica de Catalunya telescope has an aperture of 25 cm, a
physical models of the corona, can give (UPC). The equipment provided by the
Figure 5. The TAROT telescope pointing north
information about the temperature and Galileo Science Office at ESAC is the towards the Sun during totality. The enclosure of
flow speed of coronal electrons. same as that used by the Galileo Experi- the NTT can also be seen. This frame is taken from
– A flash spectrum of the chromosphere mentation & Scientific Tests in Antarctica TAROT’s webcam sequence of the full eclipse 8.
was attempted but was unsuccessful project to study the effect of solar activ-
owing to the unfortunate failure of the ity in the ionosphere at high latitudes.
The TAROT team

camera two minutes before totality. It –S urface ultraviolet irradiance measure-
consisted of a telescope projecting an ments were also completed to observe
image of the Sun through a high-quality Earth-atmospheric evidence of asym-
transmission diffraction grating onto a metric ultraviolet opacity over the
digital SLR camera without an infrared eclipse. This could have a bearing on
blocking filter. This configuration permits how observations of the corona are
one to image the emission spectrum interpreted. This was done in collabora-
of the chromosphere covering wave- tion with Ralph Lorenz from Johns Hop-

The Messenger 177 – Quarter 3 | 2019 49

Astronomical News Christensen L. L. et al., Science & Outreach at La Silla During the Total Solar Eclipse

on sky during the eclipse can conven-


iently be observed at any period roughly
six months before or after the eclipse.

The extended atmosphere of the Sun,

the corona, poses another difficulty. The
corona adds diffuse light which is not
homogeneous because of its filamentary
structure. Although very beautiful, coro-
nal features can reduce the accuracy of
the positional measurements of the stars.
Six months before the eclipse, a series of
images of field stars were recorded at
exactly the same elevation as during the
eclipse. The analysis of these images
demonstrated the ability to measure stel-
lar positions with an accuracy on the
order of ± 0.25 arcseconds, correspond-
ing to an error of ± 15% in the value of
the expected angular displacement that
Figure 6. Stars during totality; 12 stars become visi- lated to the Sun’s gravitational field. has been predicted by general relativity.
ble through the extended corona, at separations less
Removing or accounting for these effects Unfortunately, the TAROT CCD camera
than six solar radii on sky.
is crucial to allowing the accurate meas- failed one month before the eclipse and
urement of the gravitational deflection had to be replaced two days before the
focal length of 850 mm and a CCD and confirming the general theory of eclipse. In order to have proper calibra-
­camera that provides a field of view of relativity. tion images, the optical setup will be kept
1.8 × 1.8 degrees with a spatial sampling the same as during the eclipse, with the
of 3.29 arcseconds per pixel. The most common way to calibrate these aim of recording new calibration images
additional effects consists of recording in January 2020.
The stellar positions are measured exper- images of field stars during a night with
imentally on an image using Cartesian similar conditions to those during the The night before the eclipse, about
coordinates from the CCD sensor. The eclipse: i.e., the same elevation of the 200 images were recorded in the same
main difficulty is to link the Cartesian coor- stars, same optical setup, same tempera- direction as the eclipse in order to com-
dinates of the sensor to celestial coordi- ture, etc. In the case of the TAROT obser- pute the calibration coefficients between
nates on the sky, particularly as the light vations, the advantage is that, as the tele- Cartesian coordinates and celestial coor-
from stars recorded on the CCD is also scope stays in the same place, the same dinates. A second, more complex, cali-
affected by other physical effects unre- stars that would be close to the Sun bration method is ongoing which involves
computing the optical deformations for
each image and merging all of the indi-

vidual calibrations onto a master frame.

During the hour before totality, the partial

eclipse was recorded with TAROT using
an additional aperture solar filter. This
­filter was manually removed at the begin-
ning of totality. A software script was
specially written to record images during
totality and to point the telescope away
from the Sun when totality was over.
Owing to the inherent dangers to the
equipment, all of the steps were prac-
tised many times in the hours before the
eclipse. In the end, image acquisition
worked perfectly during the eclipse. A
first analysis of the images shows that

Figure 7. Part of the PACA_SolPol19 setup on one

side of the La Silla mountain.

50 The Messenger 177 – Quarter 3 | 2019

stars are detected on images with expo- Figure 9. The CAOS

ESO/C. Guirao/G. Ávila

group comprised Carlos
sure times of 5 and 10 seconds. The
Guirao and Gerardo
1-second images do not saturate the Ávila (both from ESO),
corona but stars are almost undetecta- seen here on their
ble. To illustrate the principles behind this observing post on the
NTT platform.
experiment an image has been synthe-
sised using one 1-second image and with
stars visible from the 5- and 10-second
images (see Figure 7). The next step will
be the January 2020 observations
needed for calibration.

The PACA_SolPol19 Activities

The Pro-Am Collaborative Astronomy

project formed a a professional-amateur
collaboration PACA_SolPol19c is to A major unanswered question in astro- and derivation of the polarisation follow
measure the linear polarisation of the physics is how the corona is heated. Cur- procedures used to calibrate the data
K-corona during the total eclipse. The rent coronagraphs block much of the from the 2017 total solar eclipse in the
solar corona, extending far from the Sun, inner corona, making eclipses the sim- USA. The end goal is to measure the
is hotter than the photosphere. The outer plest way this region can be investigated polarisation brightness, degree of linear
part of the corona becomes the solar from the ground. Since the inner corona polarisation and angle of linear polarisa-
wind that moves outward through the (K-corona) is dominated by electron scat- tion images (Elmore et al., 2000; Lites et
Solar System and interstellar space, inter- tering, which is linearly polarised, obser- al., 1999; van de Hulst, 1950; Quémerais
acting with planetary atmospheres and vations of polarised brightness during an & Lamy, 2002).
other Solar System objects, creating eclipse provide information about the dis-
space weather. The corona exhibits radial tribution of polarisation and the polarisa- The team successfully imaged the
filamentary structure — bright long tion brightness, pB, which is related to the ­K-­corona, revealing equatorial streamers
streamers at all latitudes during periods local electron density. (as expected for a quiet Sun), and pro-
of high solar activity. duced a map of the polarisation bright-
PACA_SolPol19 consisted of four mini- ness (see Figure 9). The colours in the
During low solar activity, as during this teams, three of them located at the polarisation brightness map represent the
eclipse, streamers are mostly limited to La Silla Observatory (Figure 8), using one angle of linear polarisation, with red being
lower latitudes. The corona, being thin imaging telescope and two polarimetric the maximum. The quiet Sun exhibited
and tenuous, is only observable during setups (one with a programmable polaris- polarimetric minima at the solar poles
total solar eclipses or with the use of a able sensor and the other with a polar- and polarimetric maxima at the solar
coronagraph to block the disc of the Sun. ised sensor), and one imaging setup at equator, with the solar prominence exhib-
sea level at Punta de Choros, La Higuera. iting low polarisation.
Figure 8. Left: The average total intensity of the All setups used the same software to
corona. Middle: Sobel-filtered image of the solar acquire imaging and polarimetric data, In addition, the solar prominence on the
corona. Right: The product of polarised brightness
(pB) and the angle of linear polarisation, colour-
taking advantage of similar observing north-west limb of the Sun and two stars
coded with red indicating the maximum polarisation conditions and initial data reduction tech- were imaged, one towards the north east
and green minimum polarisation. niques. Detailed flat-fielding, calibration, and one to the south west (tentatively

The Messenger 177 – Quarter 3 | 2019 51

Astronomical News Christensen L. L. et al., Science & Outreach at La Silla During the Total Solar Eclipse

10 100

ESO/C. Guirao/G. Ávila

Fe X 6374.4

Ca II K 3933.6

Hα 6562.6
Fe XIV 5302.6 (corona)
Fe IX 4358.5 (corona)

Fe X 6374.4 (corona)
Ca II H 3968.4
Fe XIV 5302.6
Fe IX 4358.5

Telluric lines

Telluric lines
He I 5876.4

Hα 6562.6

He I 47130.0
He II 4685.4

He I 6678.1
He I 5876.4
He I 4471.3

He I 7066.2


Hγ 4340.3

Hβ 4861.1

Hδ 4101.6

Telluric lines

Telluric lines
1 1
5000 6000 7000 4000 5000 6000 7000
Wavelength (Å) Wavelength (Å)

Figure 10. The two total solar eclipse spectra from the ESO–MIDAS software. The corona incorporated the webcast into public
the CAOS group: the spectrum of the corona (left,
spectrum shows the “classical” iron events set up for the occasion of the
10 exposures of 5 seconds) and the prominence
(right, 3 exposures of 5 seconds). (Fe XIV and Fe X) and Hα lines. The He eclipse9 and members of the public
line at 5876 Å is barely visible. viewed it online in high definition on
ESO’s website and on ESO’s YouTube
identified as HD 48805 [SAO 78717] and However, in the spectrum of the promi- channel 10. The 4-hour live webcast was a
HD 262616). nence, we found some interesting fea- “raw” feed without commentary regularly
tures: Ca II, five helium lines and the four switching between sources, and featuring
hydrogen lines of the Balmer series. views of the Sun from three small tele-
CAOS group scopes and two cameras showing spec-
tators at La Silla. The webcast finished
The group from the ESO Club of Amateurs ESO webcast
in Optical Spectroscopy (CAOS) d (Fig- Figure 11. Part of the ESO webcast team in their
ure 10) was observing from the NTT plat- Starting at 19:56 CEST on 2 July 2019, a working area in the shade under the Ritz (old control
form using a commercial 11-inch Schmidt- teame deployed by ESO provided a live room building, now visitor centre). Left is engineer
Lionel Gauze (APICAL), in the middle producer
Cassegrain telescope and a spectrograph webcast of the 2019 La Silla Total Solar François Glasser (APICAL) and to the right ESO
with a resolving power of 11 000 and Eclipse (see Figure 12). Organisations Photo Ambassador Alexandre Santerne (Aix-­Marseille
spectral range of 3930–7070 Å. A com- in the ESO Member States and beyond University/CNRS/CNES/LAM).
mercial SBIG ST1603ME CCD camera
recorded the spectra, which were cali-
ESO/M. Zamani

brated with a thorium-argon hollow cath-

ode and halogen lamps linked with a
300-µm optical fibre, for spectral calibra-
tion and order identification respectively.

Before totality, the telescope was covered

with a Mylar sheet with an optical density
of OD-5 to protect the instrument and
our sight. In total, 13 exposures of five
seconds were taken during totality. The
slit of the spectrograph was placed close
to the north pole of the Sun, and three of
these 13 exposures showed partial illumi-
nation of the slit by a solar protuberance.
The remaining 10 exposures recorded
only the much weaker corona spectrum.

Both the protuberance and corona

­spectra (Figure 11) were reduced with

52 The Messenger 177 – Quarter 3 | 2019

with a beautiful view of the Sun setting The TAROT telescopes were built and are main- c
The PACA_SolPol19 team consists of Padma A.
tained thanks to the technical and financial support Yanamandra-Fisher, Robert F. Fisher, David Elmore,
over the Pacific Ocean to conclude a day
of CNRS-INSU (ARTEMIS, IRAP), CNES and OSU Wahab A. Baouchi, Kyle A. Motl, Roy Wellington,
of amazing experiences at La Silla. Pytheas. Theo Wellington and Andrei Ursache.
The CAOS team consists of Carlos Guirao and
Considerable planning, including simula- The PACA_SolPol19 Team gratefully acknowledges Gerardo Ávila.
the support and assistance of ESO’s Department of The ESO webcasting team consists of François
tions, was carried out in the months
Communications, especially Fernando Comerón, Glasser, Lionel Gauze, Alexandre Santerne and
before the eclipse. Different potential van- and is grateful for having been selected and given Lars Lindberg Christensen.
tage points were analysed in detail using the opportunity to carry out both our science and
the online application The Photographer’s outreach experiments at La Silla.
Ephemeris 11; this also revealed limitations
in the tool’s sunset times, which are most References 1 
Total Solar Eclipse:
likely due to atmospheric refraction not news/eso1912/
being properly considered. The webcast Elmore, D. F. et al. 2000, SPIE, 4139, 370 2 
Overview of experiments conducted at La Silla
van de Hulst, H. C. 1950, Bulletin of the Astronomical during the Total Solar Eclipse: https://www.eso.
was seen by tens of thousands of viewers
Institutes of the Netherlands, 11, 135 org/public/announcements/ann19031/
on 2 July and has since accumulated Klotz, A. et al. 2013, The Messenger, 151, 6 3 
Results webpage from the CESAR team: http://
more than 250 000 views (see Figure 13). Lites, B. W. et al. 1999, Solar Physics, 190, 185
Quémerais, E. & Lamy, P. 2002, A&A, 393, 295 Solar%20Eclipse%202019%20results
Zangrilli, L. et al. 2009, Solar Physics and Space Archive of CESAR images:
Weather Instrumentation III, Proceedings of the monitor/archive/ra/visible/2019/201907/20190702/
Acknowledgements 5 
SPIE, 7438, 74380W CESAR images broadcast live during eclipse:
All authors are grateful to ESO’s management for 6 
CESAR live-streaming webcast from the event:
maximising the potential presented by this unique
occasion. 7 
ESA released images from the CESAR team:
 he ESA/CESAR team consists of Manuel Castillo,
ESA/CESAR wishes to thank the teams that made
Wouter van Reeven, Miguel Pérez-Ayúcar, Joe search?SearchText=%2Beclipse+%2Bce-
the event possible, including the ESO Department
Zender, Darío Pérez de Carlos, Ralph Lorenz, sar+%2Bjuly+-lunar&img=1&SearchButton=Go
of Communication, the ESA-CESAR observing 8 
Michel Breitfellner, David Cabezas, Donald Merrit, Webcast from La Silla:
team at La Silla and at ESAC, the ESA science direc-
and Santa Martínez. watch?v=wEiyvX9GEDI&
torate (ESA Director of Science Günther Hasinger), b 9 
The TAROT team consists of Alain Klotz, Adrien Announcement of ESO webcast: https://www.
ESA-Communications, the ESA-GNSS Galileo
Nicolas Klotz, Jean-François Le Borgne, Eric
­Science Office, the University of Torino and ESA 10 
Denoux, Christian Buil, Valérie Desnoux, Yoann The ESO webcast page:
­faculty for the EKPOL instrument and camera and
Richaud, Rico Sautile, Xavier Regal, Julien Lecubin, public/events/astro-evt/solareclipse2019/webcast/
Ralph Lorenz (ultraviolet measurements), Robert 11 
Loic Eymar, and Michel Boer. The Photographer’s Ephemeris: app.photoephem-
Nufer and Xavier M. Jubier (cameras control SETnC
and Solar Eclipse Maestro) and the BepiColombo,
Solar Orbiter, Proba-2, Proba-3 and SOHO projects
from ESA. Figure 12. Frame from the webcast at the time of

The Messenger 177 – Quarter 3 | 2019 53

Astronomical News DOI: 10.18727/0722-6691/5154

Pointing the NTT at the Sun: Studying the Solar Corona

During the Total Eclipse

Michel Dennefeld 1

P. Sinclaire/ESO
Serge Koutchmy 1
François Sèvre1
Hassan Fathivavsari 1
Frédéric Auchère 2
Frédéric Baudin 2
Shahin Abdi 2
Peter Sinclaire 3
Ivo Saviane 3
Francisco Labraña 3
Linda Schmidtobreick 3

Institut d’astrophysique de Paris (IAP),
Sorbonne Université, France
Institut d’astrophysique spatiale (IAS),
Université Paris-Sud, France

As soon as we realised that the total solar

eclipse of 2 July 2019 would be visible
from the La Silla Observatory, we saw a
rare opportunity to point a 4-metre-class
telescope at the Sun to obtain spectra of
its corona with unprecedented angular
and spectral resolution. Despite the pes-
simistic reactions of many colleagues —
“You are crazy, ESO will never accept...”
— we pursued the idea and opened dis-
cussions with ESO in early 2018 to see
how this type of observation could be
carried out in practice. Our team pre-
sented a strong argument: this would not
be the first time that a large tele­scope
would be pointed at the Sun during an
eclipse — it had previously been done
successfully with the Observatoire de
Haute-Provence (OHP) 1.93-metre tele-
scope in 1961 (Wlérick & Fehrenbach,
1963) and the 3.6-metre Canada France
Hawaii Telescope (CFHT) in 1991
(Koutchmy et al., 1994). The science
case was to observe emission lines with
different ionisation potentials at different
positions across the corona with arc­ as low as 13 degrees above the horizon, Figure 1. The NTT pointing at the Sun, with a tempo-
rary control system set up directly in front.
second angular resolution, in order to where totality would take place. A
analyse the coronal heating mechanism. detailed procedure was arranged with the
In the end, the team assembled included La Silla team to ensure the telescope the Sun directly in front of you?! Thus, our
a number of people with extensive expe- could be used without risk of fire. There programme using the ESO Faint Object
rience in both solar physics and observa- was, of course, no question that the tele- Spectrograph and Camera v.2 (EFOSC2)
tional techniques. scope should be operated from a remote at the NTT was proposed in Period 103
location (such as the Ritz control room), (Programme ID 0103.D-0139) and
The question of which telescope to use so computer consoles were erected right approved by the Observing Programmes
then arose, and how we could obtain in front of the telescope (Figure 1). This Committee!
several high-resolution spectra during the allowed close monitoring, quick reaction
short duration of totality (1m 48s). In times in case of problems, and, as a The main difficulty with the observation,
practice, the New Technology Telescope bonus, a privileged view of the eclipse — once the above problems had been
(NTT) is the only telescope able to point but did you ever try to read a screen with addressed, was the comparatively long

54 The Messenger 177 – Quarter 3 | 2019

execution time needed to obtain a spec- S
Figure 2. Slit positions
Auxiliary plotted on a coronal
trum using the standard procedure of spectograph
slit image taken by Petr
obtaining an image before moving to the Horálek (ESO). Totality
slit — this would take more than a minute occurs when the lunar
and there was no way to do the image disc reaches the second
contact point, C2 ,and
analysis beforehand. We therefore had
lasts until it moves out
to rely on parameters obtained the previ- Apparent
at C3.
motion of
ous night and point the telescope close the Moon

to the Sun (but not at it), with the shutters C2

closed, to first configure the mirror. It prom.

was then moved into position — defined NTT

only by coordinates and simulated coro- prom.
nal images — a few minutes before total-
ity. We pointed to the west of the Sun,
taking advantage of the moon protecting
us from direct sunlight (Figure 2). C3

Exposure times had been determined

weeks before through observations of the N
Moon; a sequence of 1-second expo- W
sures was launched with the telescope
tracking normally (the Sun’s motion is (Stellmacher & Koutchmy, 1974), as the ESO staff who contributed to this suc-
rela­tively small over the short duration of Moon only covers the photosphere. A cess, both on the technical aspects and
totality). The first exposures were in fact more careful analysis is under way to on the managerial side. A total solar
dark exposures, the Nasmyth shutter remove all artefacts, a challenging task in eclipse is literally an extraordinary event,
being finally opened only just before total- light of the unusual observing conditions requiring a lot of preparatory work, but
ity (C2) to obtain coronal spectra. The and the fact that the telescope had to well worth the effort.
camera was commanded directly, with- be pointed away quickly before calibra-
out the use of Observation Blocks, allow- tions could be obtained. Figure 3 (bot-
ing short total execution times of 25 s per tom) shows a spectrum obtained with the References
spectrum in fast read-out mode, and the small auxiliary spectrograph; the lower Wlérick, G. & Fehrenbach, C. 1963, The Solar
telescope was pointed away immediately part of the slit shows the corona, with Corona, Proceedings of IAU Symposium 16,
at the end of totality. We obtained five [Fe X] dominating on the left edge. The ed. Evans, J. W., (New York), 199
good exposures of the corona, with the upper part of Figure 3 shows the spec- Koutchmy, S. et al. 1994, A&A, 281, 249
Stellmacher, G. & Koutchmy, S. 1974, A&A, 35, 42
sixth being overexposed as Baily’s beads trum of a weak prominence recorded by
appeared at the very end of totality (C3, chance (see the lower side of Figure 2),
Figure 2). which is dominated by hydrogen and Figure 3. Top: Two EFOSC2-Gr#20 spectra of the
helium lines (order superposition is also corona taken at the same position but at two differ-
ent times. The signal is dominated by the continuum
A high-resolution grism (Gr#20) was used present). Altogether, we obtained unique of the K corona; note the presence of parasitic Hα.
with a specially manufactured offset slit spectroscopic data during this eclipse, Bottom: Spectrum from the auxiliary spectrograph
(kindly provided by colleagues in Paranal) and we would like to warmly thank all the during totality.
to obtain velocities and line profiles, albeit
at the expense of spectral coverage. A 16 [Fe X] Corona
lower-­resolution, table-mounted auxiliary
spectrograph was used in parallel outside Atm
the NTT to record the full coronal spec- 12
Flux (× 10 6 )


trum over a wider field of view but at a 10

much lower spatial resolution. While the 8
data reduction is still ongoing, preliminary Hα
results show that the coronal emission
was quite weak; this is not surprising 4

near solar activity minimum. The domi- 6100 6200 6300 6400 6500 6600 6700 6800 6900 7000 7100 7200
nant line in the NTT spectra is the [Fe X] Wavelength (Å)

6374 Å line (Figure 3). Perhaps more sur- Coronal + Prominence spectrum
prising is the Hα line, in spectra taken ­
Hβ Hα He I He I
at the beginning and end of totality. It Moon background O2 atm
is of chromospheric origin, and is due to Fe X
the variable illumination of the line of Coronal spectrum
sight crossing low atmospheric layers

The Messenger 177 – Quarter 3 | 2019 55

Astronomical News DOI: 10.18727/0722-6691/5155

Report on the ESO Workshop

KMOS@5: Star and Galaxy Formation in 3D — Challenges in KMOS 5th Year

held at ESO Headquarters, Garching, Germany, 3–6 December 2018

Eleonora Sani 1 topics ranging from early stellar evolution available so far. In particular, the one-di-
Michael Hilker 1 to stellar populations and even including mensional spectra produced by classical
Lodovico Coccato1 exoplanets. spectroscopy are insufficient for a further
Suzanne Ramsay 1 investigation of the, often complex, gal-
Chris Evans 2 Over the next few years, the community axy morphologies and dynamics. Exactly
Myriam Rodrigues 3 will have access to basic data on a large the same limitation occurs in many other
Linda Schmidtobreick 1 number of sources across a wide range fields, for example, studies of metallicity
Ray Sharples 4 of redshifts thanks to the upcoming or gradients and dynamics in stellar clus-
already available capabilities of survey ters. This is indeed the framework within
instruments on powerful telescopes (for which KMOS is expected to play a funda-
ESO example, the Sloan Digital Sky Survey mental role.
ATC, Royal Observatory Edinburgh, UK [SDSS], Wide Field infrared Camera for
University of Oxford, UK UKIRT [WFCAM], Visible and Infrared The KMOS@5 workshop aimed to bring
Durham University, UK Survey Telescope for Astronomy [VISTA], together scientists working on all areas
VLT Survey Telescope [VST], High Acuity of star and galaxy formation and evo­
Wide field K-band Imager [HAWK-I], and lution using near-infrared IFU spectros-
The K-band Multi-Object Spectrograph Large Synoptic Survey Telescope [LSST]). copy. The five-year milestone offered a
(KMOS) is one of the second-­generation The high-redshift samples will enable a perfect opportunity to assess the impact
instruments at the VLT, and has been new look at the early stages of the Uni- of KMOS on its core science cases. The
operating for five years. To celebrate verse when galaxies were young or still workshop format allowed the exchange
this anniversary this workshop brought forming. Although the ensemble data of strategies and ideas for analysis of the
together astronomers to present scien- already available — which have mostly KMOS data, as well as fruitful discussion
tific results from KMOS and comple- been gathered by means of photometric of future programmes. Finally, practical
mentary instruments. The topics ranged techniques — are already providing valu- tutorials and demonstrations were offered
from star formation in the Galactic centre, able insights, a deeper understanding of and the users, together with the Instru-
to stellar populations in globular clusters, the detailed physics underlying the for- ment Operations Team (IOT), revised the
to galaxy formation and evolution at vari- mation and growth of galaxies requires priorities to improve and optimise the
ous redshifts, and feedback from active more information about their individual performance of KMOS over the next five
galactic nuclei (AGN). Another goal of properties. The coverage, in terms of star years.
the workshop was to assess the impact formation rates and redshifts, of the cur-
of KMOS on its core science goals and rent KMOS Guaranteed Time Observa- In the following sections we summarise
to develop new strategies and pro- tion (GTO) surveys is shown in Figure 1. some of the interesting talks and high-
grammes, also in light of future integral Such knowledge, in particular for high-­ lights from each session.
field unit (IFU) instruments. About 60 redshift objects, was not sufficiently
researchers from the astronomical
Lookback time (Gyr)
community and members of the Instru-
ment Operations Team participated in 0 2 4 6 8 10 12
the workshop and discussed the above –0.4
topics; these discussions served to KMOS3D
identify the highest priority improve-
ments that could increase the scientific
return of KMOS in the future. KGES KDS

log ψ(M๬year –1 Mpc –3)

Motivations –1.2

Any advances in observational astronomy KLENS

are ultimately based on the quality of
observational data obtained with ever-­ –1.6
improving, increasingly sophisticated, KASHz
telescopes and instruments. KMOS ena-
bles deeper insights especially in areas
related to galaxy formation and evolution, –2
as well as a wider variety of scientific
Figure 1. Cosmic star formation rate as a function of 0 1 2 3 4 5 6 7 8
redshift and lookback time (plot from Natascha Förster-­
Shcreiber’s talk; data from Madau & Dickinson, 2014). Redshift

56 The Messenger 177 – Quarter 3 | 2019

1800 Figure 2. Requested and allocated time on KMOS.
The vertical red line marks the small amount of
time requested in Period 103, before the KMOS@5
workshop. The subsequent significant increase in
Period 104 may be seen as the positive reaction of
the community after the workshop.
Time (h)



92 93 94 95 96 97 98 99 100 101 102 103 104 Requested time (h)
Observing period Allocated time (h)

Workshop opening centrated and that there exists a broad can be used as cosmic abundance probes
metallicity distribution of stars in the to construct new mass-metallicity scaling
The meeting began with a welcome talk by Nuclear Star Cluster (NSC) of the Milky relations based on red supergiants.
Bruno Leibundgut, who presented the ris- Way based on ~ 700 KMOS spectra
ing publication statistics of KMOS, which (Feldmeier-Krause et al., 2015, 2017).
are currently dominated by GTO data, Anna McLeod also demonstrated that Galaxy assembly, dynamics and
a natural consequence of the recently with the use of the mosaic mode of evolution
expired GTO time. He also showed that KMOS it is possible to examine the early
the demand for KMOS as measured by stages of stellar feedback in star-forming Day 2 was dedicated to galaxy assembly,
proposal numbers and time requested is molecular clouds within the Milky Way, and galaxy dynamics and evolution.
steadily decreasing, reaching a minimum which are only observable in the infrared The results of the main KMOS GTO pro-
of only ~ 750 hours in Period 103 (see because of extinction. In particular, it has grammes were presented, with particular
Figure 2). Bruno warned the participants become possible to trace the ionisation focus on KLEVER (talks by the Principal
that under-­requested instruments may front and the molecular gas content at Investigator Michele Cirasuolo and Mirko
not be supported by ESO and have an the same time. KMOS observations of Curti), KROSS and KGES (Mark Swin-
increased risk of being decommissioned. massive star clusters reveal evidence for bank and Alfred Tiley talks), and KMOS3D
This had the effect of shaking up the par- outflows and wind-blown bubbles in the the largest GTO programme (talks by the
ticipants, and the wakeup call animated environment of the clusters. Principle Investigator Natascha M.
all subsequent speakers, further motiv- Förster-­Schreiber, Hannah Übler, Philipp
ating them to illustrate the importance of Francesco Ferraro described how the Lang, David Wilman). Star formation
KMOS for their science. It also triggered Multi-Instrument Kinematic Survey of rates, resolved kinematics and metallici-
lively discussion of the future of KMOS Galactic globular clusters (MIKiS) allowed ties of more than 1000 star-forming gal-
throughout the conference. the determination of stellar kinematics axies at redshifts between 0.5 and 3 have
across their full radial range. This survey been determined and have established
was designed in synergy with the VLTI the following:
Galactic and Local Volume science instruments, the Fibre Large Array i. Most galaxies (> 70%) are rotationally
Multi Element Spectrograph (FLAMES), supported and the Hubble sequence
The first scientific session concentrated KMOS, and the Spectrograph for INtegral emerged at around redshift z ~ 1.5
on Galactic and Local Volume science Field Observations in the Near Infrared (Stott et al., 2016, Swinbank et al., 2017).
with KMOS. It was convincingly shown (SINFONI). Resolved kinematics of ii. In terms of metallicity, non-axisymmetric
that KMOS can be efficiently used to thousands of stars, combined with patterns are revealed from resolved
trace massive star formation and its feed- internal proper motions measured metallicity maps, while azimuthally-­
back in both nearby star-forming galaxies from Hubble Space Telescope (HST) averaged metallicity gradients are flat.
as well as in the Galactic Centre and in campaigns (inner regions) and Gaia (out- It is also possible to characterise out-
star-forming regions of the Galactic disc. skirts), provide the first 3D kinematic flow statistics; while the incidence of
maps of Galactic globular clusters. Sur- star-­formation-driven outflows depends
Thanks to the multiplex capabilities of prisingly, they do not rotate as rigid bod- on star formation properties, the frac-
KMOS, as well as the mosaic mode ies, but rather follow a Keplerian law with tion of AGN-driven outflows depends
with a monolithic IFU, it is possible to no significant evidence of intermediate on the stellar mass and its concentra-
cover a large field of view (up to almost mass black holes (see Ferraro et al., tion (Förster-Schreiber et al., 2019;
1 square arcminute) and ~ 5 pc 2 of the 2018; Lanzoni et al., 2018). The flexibility Harrison et al., 2016).
Galactic centre have been observed. of KMOS allows studies of young mas-
Anja Feldmeier-­K rause discussed how sive star clusters in nearby galaxies, and A controversial topic in this session was
young massive stars are centrally con- Ben Davies showed how such clusters whether the outer discs of galaxies at

The Messenger 177 – Quarter 3 | 2019 57

Astronomical News Sani E. et al., Report on the Workshop “KMOS@5: Star and Galaxy Formation in 3D”

redshift z ~ 1–2.5 have flat or falling rota- velocity to escape from the galaxy halos — clusters are older than their analogues in
tion curves. rather it will be re-accreted, contributing to the field; (2) gradual mass-growth mech-
the build-up of stellar mass and angular anisms like minor mergers are favoured
In his contribution, Philipp Lang dis- momentum of the galaxies. (Beifiori et al., 2017); (3) ram pressure
cussed how a representative rotation stripping might compress gas in the inner
curve for high redshift galaxies can be Trevor Mendel’s contribution dealt with discs, thus triggering star formation; and
obtained with KMOS3D; it is characterised the galaxy evolution mechanisms leading (4) non-circular motions are dominant in
by a significant decrease in velocity in the to massive and passive galaxies. He low-mass cluster members, thus indicat-
outer regions. Such a drop in rotation showed how they formed in a two-phase ing kinematic downsizing and/or interac-
velocity can be explained by the domi- process: an early phase driven by rapid tion processes.
nance of baryons within the shallow inner star formation on the main sequence, fol-
dark matter potential. These results sup- lowed by the assembly of already existing
port the limited role of dark matter on stellar mass (Mendel et al., 2015). Also, KMOS current and future perspectives
disc scales (Lang et al., 2017; Übler et al., such evolution can explain the typical
2018). On the other hand, Alfred Tiley decrease in dark matter fraction within In addition to the scientific sessions,
showed that stellar-scale rotation curves the half-light radius. discussions on the present and future
obtained from the KMOS Redshift One status of KMOS were held. We asked the
Spectroscopic Survey (KROSS) sample The fourth and last workshop day was audience to complete a questionnaire, and
remain flat or continue to rise indepen­ dedicated to the environment and late- the answers were used to drive further
dently of redshift. This implies moderate stage evolution of galaxies. Alessandra discussions. In Figure 3 the results of the
to large dark matter fractions (≧ 66%) in Beifiori, Asmus Bohem and Sam Vaughan online survey are reported. Concerning
star forming galaxies over the last 10 Gyr discussed KMOS Galaxy Cluster surveys the needs of the community, it is clear
(Tiley et al., 2019). like the KMOS Cluster Survey (KCS) and that archival search and pipeline improve-
the KMOS Cluster Lensing And Supernova ments are two key areas of interest:
On Day 3, the focus switched to lensed survey with Hubble (K-CLASH), showing i. Querying the archive for KMOS data is
and very distant systems as well as AGN that: (1) massive galaxies in big and old tricky because so far only the central
feedback and stellar kinematics. Results
were presented from GTO surveys, such What is “most wanted” to improve the KMOS experience?
as KMOS3D, the KMOS LENsing Survey
Straightforward archive search Persistence correction
(KLENS), and the KMOS Lens-Amplified
Spectroscopic Survey (KLASS), related to 40% 16.7%
the cosmic dawn. These surveys help in Support/tips for specific
understanding the properties of distant science use cases
systems. Charlotte Mason showed the first 3.3%
robust constraint on the intergalactic neu-
Straightforward archive search
tral hydrogen fraction at z ~ 8, inferred Pipeline improvement + science-ready data product
from deep spectroscopic limits on Lyα
emission, and Marianne Girard discussed 36.7% 3.3%
how the evolution of velocity dispersion KMOS with AO preferences
depends on stellar mass.

AGN feedback and its effect on different

stages of galaxy evolution were discussed
during the afternoon, with results from Keep the spaxel size, Reduce the FoV/ increase
the KMOS AGN Survey at High redshift i.e., GLAO correction? the spatial resolution
(KASHz) and VIRIAL GTO programmes. 73.3% 26.7%
Chris Harrison, Jan Scholz and Rebecca
Davies argued that: (1) the most extreme
gas kinematics are associated with AGN Which of the following facilities provides the most promising synergy with KMOS?
and outflows driven by low-power jets are
important at low redshift; (2) there is no JWST
evidence of AGN-driven outflows quench- 36.7% ALMA
ing star formation in moderate luminosity 30%
AGN; and (3) the vast majority of the out-
flowing material does not have sufficient

Figure 3. Online questionnaire about the most
desired KMOS capabilities and synergies. 33.3%

58 The Messenger 177 – Quarter 3 | 2019

pointing coordinates of the KMOS field run short on spares for the arms and are KMOS workflow to include external Python
are searchable, rather than those of the not able to build new ones, the solution scripts within the data reduction cascade.
individual targets. Such issues will be could indeed be to reduce the number of
fixed shortly thanks to the new ESO arms, while increasing the field of view.
Archive Science Portal and the data Main conclusions & ways forward
product release. ESO will indeed Regarding synergies between KMOS and
release reduced data cubes for single other facilities, the community is divided The workshop brought together the
targets, which will be retrievable from on whether or not all of the major current KMOS community to celebrate the fifth
the ESO Archive Science Portal2. and future facilities (for example, the anniversary of the first second-generation
ii. Pipeline capabilities should be Atacama Large Millimeter/submillimetre VLT instrument. All GTO and Large pro-
improved. Since several recipes have Array [ALMA], the Extremely Large grammes as well as some dedicated
been recently updated and released, Telescope [ELT], and the James Webb studies produced great scientific results,
the IOT asked the community to test Space Telescope [JWST]) are considered leading to a steadily rising publication
the new releases and provide detailed crucial to fully exploiting the multiplex and citation record.
feedback on specific tasks. capabilities of infrared IFUs like KMOS.
These themes also emerged during the The interaction between the community
Moreover, the audience raised a concern sessions, for example, in the talks by and the KMOS IOT has been fruitful and
regarding the ability of the Exposure Jan Scholz, Chian-Chou Chen, Michele has led to the implementation of new
Time Calculator (ETC) to provide the cor- Cirasuolo, Charlotte Mason and Dominika strategies to broaden the number of users
rect signal-to-noise ratio for the faintest Wylezalek. and plan the future of KMOS and its
objects. The ETC has been extensively operations. Bruno Leibundgut’s warning
tested on standard targets (such as stars During a round table session the discus- regarding a decline in the demand of
and line-emitting regions, both Galactic sion turned to which approach would be KMOS had a very positive effect on the
and extragalactic) and it is able to provide more effective to increase demand for KMOS community. In Period 104, the
a S/N fully consistent with the data. How- KMOS — whether it would be better to requested KMOS time significantly
ever, for those objects with a signal of issue a call for Public Surveys3 or for increased, to a level comparable to the
the same order of magnitude as detector more Large Programmes. The commu- first time the instrument was offered (see
effects (i.e., variable bias, cross talk, nity was in favour of the second option. Figure 3). This can be seen as a great
remanence), the ETC underestimates the success of the KMOS@5 workshop.
S/N by ~ 30% (Mason et al., 2019). The
effect of imperfect sky subtraction cannot Data reduction tutorial
be simulated by simply scaling the S/N Demographics
by a factor √2, and this, together with the In the afternoon of the last workshop day,
above-mentioned factors, leads to the about 25 participants attended practical The Science Organising Committee
conclusion that the S/N for such extremely tutorials with hands-on sessions dedi- sought fair representation from the
faint sources cannot be easily simulated cated to KMOS data reduction, followed KMOS science community in terms of
within the ETC. Nonetheless, the IOT by a final Glühwein and Spekulatius fare- gender, seniority and institutes. The
could provide the ETC with a fudge factor well reception. The aim of the tutorial was committee invited 11 speakers to cover
once the signal reaches a given thresh- to introduce the instrument, the design of all scientific topics and major KMOS
old. A large amount of data (for example, the data reduction pipeline and the data programmes, with a 6:5 ratio of male to
from GTO and Large Programmes) is reduction cascade, and to present the female speakers and 5:6 ratio of senior
needed to determine this threshold, and KMOS ESOReflex (Freudling et al., 2013) (staff) to junior (postdoc level) speakers.
the KMOS GTO teams together with PIs workflow as data reduction tool4, 5. The total number of participants was 60
of Large Programmes are prepared to (with a female fraction of 42%), which
provide feedback to the IOT on this task. Special emphasis was given to the allowed a focused and interactive work-
explanation of different data reduction shop with relaxed time constraints. We
We discussed possible upgrades of the strategies, including removal of telluric therefore had the luxury of accepting all
instrument. The community was largely in features and sky subtraction. Because requested talks (25) and posters (6), with
favour of equipping KMOS with a Ground- KMOS data reduction can be complex, the exception of three submissions that
Layer Adaptive Optics (GLAO) system to and the optimisation of the results can were out of the scope of the conference.
significantly improve the sensitivity, require different strategies and algo- The gender balance for different groups
thanks to the AO correction, while still rithms, members of the astronomical in the conference can be seen in Figure 4.
preserving the spatial sampling and community have come up with their own The female representation was between
hence the current field of view of each solutions for specific datasets over the 33% to 50% in all categories.
IFU. Another possibility considered (albeit years, some of which were included
not shown in Figure 3), is related to the in the KMOS pipeline and workflow. The level of participation from young
deterioration of the arms — by far the researchers was very high, with the fol-
most delicate optomechanical component Therefore, one aspect that was covered in lowing breakdown according to seniority:
in KMOS. In the unfortunate event that we the tutorial session was how to modify the ~ 33% students, ~ 35% postdoctoral

The Messenger 177 – Quarter 3 | 2019 59

Astronomical News Sani E. et al., Report on the Workshop “KMOS@5: Star and Galaxy Formation in 3D”

researchers, and ~ 30% tenure-track or 100% Figure 4. Gender bal-

ance and career stage
tenured faculty. In particular, the junior
statistics for the KMOS
researchers were well-represented in the 80% 2018 workshop.
talks (see Figure 4). Given the nature of
the KMOS science community, most of 60%
the attendees (excluding the LOC mem-
bers) came from European institutes 40%
(~ 90%), and the rest (~ 8%) from the
United States and Australia (~ 2%). The 20%
conference picture (Figure 5) shows the
majority of the participants in front of the 0%
ESO headquarters.

We thank all the participants in the KMOS 2018
workshop for their enthusiasm, and the speakers for 60%
their outstanding scientific contributions. We are
grateful to the Directorate of Science, the Science
Operations Department, the Instrument Science 40%
Department and the User Support Department for
their financial support of this workshop. 20%

References 0%
Participants Posters LOC Invited Chairs Contributed SOC
speakers speakers
Beifiori, A. et al. 2017, ApJ, 846, 120
Feldmeier-Krauser, A. et al. 2015, A&A, 584, 2 Male Female Staff Postdoc Student Other
Feldmeier-Krauser, A. et al. 2017, MNRAS, 464, 194
Ferraro, F. et al. 2018, ApJ, 860, 50
Freudling, W. et al. 2013, A&A, 559, 96
Förster-Schreiber, N. et al. 2019, ApJ, 875, 21 2
Swinbank, M. et al. 2017, Nature, 543, 318  he ESO Archive Science Portal: http://archive.eso.
Harrison, C. et al. 2016, MNRAS, 456, 1195 Übler, H. et al. 2018, ApJ, 854, 24 org/scienceportal/home
Lang, P. et al. 2017, ApJ, 840, 92 3
Tiley, A. et al. 2019, MNRAS, 485, 934 ESO Public Survey Policies:
Lanzoni, B. et al. 2018, ApJ, 856, 11 sci/observing/PublicSurveys/policies.html
Madau, P. & Dickinson, M. 2014, ARA&A, 52, 415 4
ESO Reflex:­
Mason, C. et al. 2017, ApJ, 838, 14 Links esoreflex/
Mason, C. et al. 2019, MNRAS, 485, 3947 5
K MOS Tutorial Session at the workshop: https://
Mendel, J. T. et al. 2015, ApJ, 804, 4 1
 ink to workshop programme and presentations:
Stott, J. P. et al. 2016, MNRAS, 457, 1888 rial.html
Figure 5. The KMOS@5 participants.

ESO/L. Calçada

60 The Messenger 177 – Quarter 3 | 2019

Astronomical News DOI: 10.18727/0722-6691/5156

Report on the ESO Workshop

Preparing for 4MOST — A Community Workshop Introducing ESO’s

Next-Generation Spectroscopic Survey Facility

held at ESO Headquarters, Garching, Germany, 6–8 May 2019

Joe Liske 1 observing time will be available to the cal” information about 4MOST. These
Vincenzo Mainieri 2 ESO community to conduct additional presentations largely corresponded
surveys. Regardless of their provenance, to the articles by de Jong et al. (2019),
however, all 4MOST surveys will be ESO Walcher et al. (2019) and Guiglion et al.
Universität Hamburg, Germany Public Surveys. (2019) in the above-mentioned 4MOST
ESO issue of The Messenger, and all of the
The process of selecting the Community information presented is also available
Surveys will be initiated by a Call for on the 4MOST website2. The 4MOST PI,
The 4-metre Multi-Object Spectroscopic Letters of Intent for Public Spectroscopic Roelof de Jong, presented an overview
Telescope (4MOST) is a state-of-the- Surveys, to be issued by ESO by the end of the 4MOST project and the technical
art, high-multiplex, fibre-fed, optical of 2019. The workshop reported on here characteristics and capabilities of the
spectroscopic survey facility currently was jointly organised by ESO and the facility. In a separate presentation he also
under construction for ESO’s 4-metre 4MOST Consortium to prepare the laid out the constraints and principles of
Visible and Infrared Survey Telescope broader ESO community for this exciting the survey strategy.
for Astronomy (VISTA). During the first scientific opportunity, to help potential
five years of operation 4MOST will Principal Investigators (PIs) respond Jakob Walcher provided a very lucid,
be used to execute a comprehensive successfully to the Call, and to foster high-level account of the complex
programme of both Galactic and extra- scientific collaborations between the 4MOST operations scheme. This was a
galactic Public Surveys, and 30% of the community and the 4MOST Consortium. particularly important talk because sev-
observing time during this period will eral aspects of this scheme, including the
be available to the community. The pur- Specifically, the goals of the workshop role of the 4MOST Consortium in opera-
pose of this workshop was to prepare were: (i) to provide the ESO community tions, and the concepts of a shared
the ESO community for this exciting with up-to-date information regarding focal plane and participating and non-­
scientific opportunity. the 4MOST facility, its capabilities, survey participating Community Surveys, are
strategy, data reduction and science new to the world of ESO operations and
pipelines, the 4MOST Consortium’s were thus unfamiliar even to experienced
ESO has a long history in survey astron- scientific plans, and the application ESO users. This talk was further comple-
omy, dating all the way back to its original and selection processes for 4MOST mented by Nic Walton’s presentation on
mission. A new chapter will be added Community Surveys; (ii) to provide the 4MOST data reduction and scientific
to this history by the advent of 4MOST, ESO community with an opportunity to analysis pipelines, and data products.
a spectroscopic survey facility featuring present their scientific ideas for 4MOST Sofia Feltzing and Joe Liske explained
a field of view large enough to survey a Community Surveys; and (iii) to provide the concept of the 4MOST Science Team
large fraction of the southern sky in a few a platform for discussion, networking in some depth, i.e., the organisational
years, and a multiplex of 2400 fibres ena- and collaboration between potential entity within which all of the Consortium
bling surveys of tens of millions of objects Community Survey PIs and the 4MOST and participating Community Surveys
(de Jong et al., 2019). 4MOST will spec- Consortium, and to explore complemen- work together to plan, execute and
troscopically complement a number tarities between Consortium Surveys and exploit the 4MOST survey programme.
of current and future facilities, including potential Community Surveys. Finally, Vincenzo Mainieri, the ESO
Gaia, eROSITA and Euclid, and will 4MOST Project Scientist, detailed the
address a wide range of science areas, To prime the workshop, the 4MOST process by which ESO will select the
from the structure of the Milky Way to Consortium had published a series of Community Surveys.
cosmology. 13 articles in the March 2019 issue of
The Messenger1 (i.e., two months before The second category comprised 10 pres-
To enable its science goals, 4MOST was the workshop), describing the facility, its entations, one for each Consortium
specifically designed as a facility for exe- operations, the survey strategy, and each Survey, in which the 4MOST Consortium
cuting large surveys. Hence, for a period of the 10 Consortium Surveys. laid out its scientific plans. Briefly, the
of at least five years, VISTA will be dedi- Consortium Surveys consist of: four sur-
cated exclusively to observations with Broadly speaking, the presentations at veys complementing Gaia and targeting
4MOST, and the 4MOST facility will in the workshop fell into three categories, the bulge/disc and halo components of
turn be dedicated entirely to a compre- which are described in the following the Milky Way at low and high spectral
hensive, five-year programme of both sections. resolutions, respectively; a survey of the
Galactic and extragalactic surveys. 70% Magellanic Clouds; two surveys following
of the observing time during this five-year up galaxy clusters and active galactic
period will be awarded to the 4MOST Providing information about 4MOST to nuclei (AGN) detected by the X-ray tele-
Consortium in return for delivering and the community scope e-ROSITA, respectively; a galaxy
operating the facility. This time will be evolution survey; a cosmology survey;
spent on a set of 10 distinct yet interlock- The seven presentations in the first cate- and, finally, a survey dedicated to the
ing surveys, collectively known as Con- gory essentially provided the community follow­-­up of extragalactic transients.
sortium Surveys. The other 30% of the participants in the workshop with “techni-

The Messenger 177 – Quarter 3 | 2019 61

Astronomical News Liske J., Mainieri V., Report on the ESO Workshop “Preparing for 4MOST”

ESO/M. Zamani
Figure 1. Participants at the Preparing for 4MOST conducting a comprehensive survey of ties and chemical compositions. A full
workshop outside ESO headquarters.
120 globular and 1500 open clusters. characterisation of this high velocity
population would allow us to trace the
Meanwhile, Henri Boffin argued the Galactic potential and constrain the
Each of the talks in this category case specifically for observing young environment of the Galactic centre,
presented the survey’s scientific context, stellar clusters in order to unravel their whence these stars were ejected. Finally,
specific goals, currently planned survey connected formation histories in large Giampaolo Piotto discussed the crucial
area and target selection, and its data complexes. Similarly, Nicholas Wright role of 4MOST in characterising the target
quality requirements. The goal of these outlined a survey of high- and low-mass sample of the PLATO mission.
talks was to provide potential Community young stars and ionised nebulae across
Survey PIs with enough information to several massive star-forming complexes
decide whether the scientific questions to extend our view of star formation and Community proposals: extragalactic
they had in mind (i) are already addressed early stellar evolution beyond the most science
explicitly by the Consortium, (ii) can be nearby and most clustered stellar sys-
addressed with the Consortium Surveys’ tems to the entirety of such star-forming Kicking off the extragalactic part of the
data (all of which will be made public), complexes. This limitation on our current community talks, Hans Böhringer put
or (iii) require a new survey. Since the view of star formation also motivated forward a proposal for a 4MOST redshift
Consortium Surveys were already Germano Sacco to propose an unbiased survey of candidate members of galaxy
succinctly described in the above-­ survey of all pre-main sequence and clusters in the redshift range 0.4–0.8,
mentioned 4MOST issue of the upper main sequence stars within selected initially from KiDS and VIKING,
Messenger we will not discuss these 500 pc. Pre-main sequence stars were and later from Euclid. He argued that this
talks further here. also on Giacomo Beccari’s mind, but in survey would support the main Euclid
the context of protoplanetary discs, and cluster science by extending the cluster
he discussed how 4MOST data of such sample below a redshift of 0.8 (where
Community proposals: Galactic science stars in young starburst clusters could Euclid’s own infrared spectroscopy is
add chemical and kinematic information ineffective) and calibrating the Euclid
The third category consisted of 16 pres- to existing Hα, Gaia and WISE data to cluster selection function and cluster
entations in which members of the com- comprehensively understand disc frac- masses with high precision. On the topic
munity presented their science cases for tions and lifetimes. of AGN, Gandhi Poshak described his
Community Surveys. Eight of these were efforts to construct a complete sample
concerned with Galactic science, and a Moving away from clusters and star of AGN within 250 Mpc, which 4MOST
recurring theme among them was that formation, Carme Gallart addressed the could support by providing spectroscopic
4MOST’s wide field of view is singularly formation history of the Milky Way by dis- follow-up of infrared- and X-ray-selected
well suited to studying stellar clusters, cussing age distributions (derived from candidates.
associations, star-forming complexes and Gaia colour-magnitude diagrams) of
their larger scale environment. Sara geometrically defined halo and disc sam- Moving on to the domain of galaxies,
Lucatello and Antonella Vallenari jointly ples and proposed a similar analysis for Arjen van der Wel summarised some
proposed complementing the chemo-­ samples defined by abundances and results from the recently completed VLT
dynamical studies of field stars by the kinematics using 4MOST. Tommaso survey Large Early Galaxy Astrophysics
Consortium Surveys with a commensu- Marchetti made an interesting case for a Census (LEGA-C) and shared his
rate effort to understand the formation 4MOST survey of thousands of candidate thoughts on the possibility of comple-
of stellar clusters, their evolution, and hypervelocity stars identified from Gaia menting the Consortium’s galaxy redshift
their relation with the field population by data in order to obtain their radial veloci- survey (Wide-Area VISTA Extragalactic

62 The Messenger 177 – Quarter 3 | 2019

Survey — WAVES) with an intermediate The presentations from the three catego- (25%) talks contributed by the community
redshift, high-S/N extension, thus allow- ries above were interspersed with one being delivered by women. The latter
ing stellar populations, star formation another and spread throughout the ratio was unfortunately slightly lower
histories and kinematics to be derived programme3. Also, each day of the than the ratio of 7:24 (29%) among the
with the same precision as in the local workshop included two question and requested talks, but the SOC agreed that
Universe. Along the same lines, Amata discussion sessions to provide the the three submissions by women that
Mercurio presented her proposal for a opportunity for participants to clarify any were not selected did not sufficiently
southern (i.e., 4MOST) extension of the remaining issues, and to jointly discuss address the goals of the workshop. In the
high-S/N Stellar Populations at intermedi- the connections between the scientific second category (i.e., science presenta-
ate redshifts Survey (StePS) which will ideas presented by the community tions by the Consortium) the gender
be carried out with WEAVE, a wide-field and the Consortium’s plans. balance was slightly better; while the
multi-­object and multi-IFU facility on the female fraction should have been 40% in
William Herschel Telescope. The distin- this category according to the initial plan-
guishing feature of StePS-South would Conclusions ning, it became 30% owing to the late
be its synergy with WAVES, i.e., the arrival of one of the female speakers. This
combination of deep spectroscopy The workshop was a resounding suc- percentage is still representative of the
and detailed environmental information. cess. As was expected, the community 31% of Consortium Survey PIs who are
identified a number of exciting scientific women. Full gender parity was achieved
Another type of synergy — that between themes, ranging from protoplanetary among the session chairs.
4MOST and MeerKAT — was discussed discs to galaxy evolution and cosmology,
by Kenneth Duncan. He described the that can be addressed with 4MOST in In terms of academic age, the speaker
wealth of information that could be addition to the science cases proposed roster was heavily biased towards senior
derived from the combination of 4MOST by the Consortium. Many of the commu- people, with only three postdocs and one
spectroscopy with radio continuum nity speakers stressed the complementa- PhD student among the speakers. We
and HI data from the deep extragalactic rity between their science goals and attribute this again to the special nature
MeerKAT surveys MeerKAT International those of the Consortium Surveys, thus of this workshop, where the Consortium
GigaHertz Tiered Extragalactic reinforcing the vision that the scientific was represented by senior figures of
Exploration (MIGHTEE) and Looking at value of the 4MOST survey programme the 4MOST project and the Consortium
the Distant Universe with the MeerKAT as a whole is larger than the sum of its Surveys; similarly, the community was
Array (LADUMA). This would include, inter constituent parts. Furthermore, of the ~ represented by potential leaders of
alia, the star formation history of the 90 workshop participants in total, about Community Surveys. Such positions tend
Universe, the evolution of the cosmic HI 55 were from the community. Consider- to require long-term employment stability,
density, and the fundamental relations ing that this workshop was mostly aimed i.e., permanently employed staff.
between galactic HI content and star for- at potential PIs of Community Surveys,
mation, stellar mass, and environment. we believe that the turnout, as well as the
Lingyu Wang then proposed a multi-­ wide range of science cases presented, Acknowledgements
purpose 110-square-degree redshift demonstrated the strong interest within The organisers would like to thank ESO for providing
survey of 0.5 million intermediate redshift the community in the large-scale survey the financial, administrative and logistical support
galaxies in Stripe 82, a region covered capabilities provided by 4MOST. In turn, for this workshop. Special thanks goes to Stella
by a wealth of multi-wavelength data, the organisers received feedback from Chasiotis-­K lingner and the rest of the LOC for the
smooth and flawless organisation, to Tania Johnston
with the primary aims of constraining the community participants indicating that — the ESO Supernova coordinator — for a very
nature of dark matter and dark energy, the Consortium and ESO had done a enjoyable planetarium show, and to the librarians for
and tracing galaxy and AGN evolution good job of providing relevant information their support in publishing the talks using Zenodo.
over the redshift range 0.2–0.6. Turning to the participants.
to the low-redshift Universe, Edward References
Taylor reported on the forthcoming
Taipan survey and proceeded to make Demographics de Jong, R. S. et al. 2019, The Messenger, 175, 3
the case for a complete hemispheric Guiglion, G. et al. 2019, The Messenger, 175, 17
Walcher, C. J. et al. 2019, The Messenger, 175, 12
4MOST redshift survey of 5.5 million gal- Although the Scientific Organising
axies out to a redshift of 0.1 to study the Committee considered gender balance
baryon lifecycle of galaxies as a function while putting together the programme, Links
of mass and environment, and to map this was not easy to achieve owing to the 1
Link to 4MOST Messenger issue:
out the large-scale density and velocity special nature of this workshop. Of the​
fields. Finally, Dominik Bomans high- total sample of 33 speakers only 8 (24%) toc.html?v=175&m=Mar&y=19
lighted the ability of 4MOST to identify were women. This was due to only 2
Link to 4MOST website:
relatively large samples of rare galaxies 1 out of 7 speakers in the first category Link to workshop programme: https://www.eso.
by focusing on the case of extreme above being female, reflecting a dearth The workshop programme contains links to videos
emission-­line galaxies. of women in leadership positions in the and PDFs of all presentations, as well as videos of
4MOST project, and only 4 out of 16 the discussion sessions.

The Messenger 177 – Quarter 3 | 2019 63

Astronomical News DOI: 10.18727/0722-6691/5157

Report on the ESO Workshop

ALMA Development Workshop

held at ESO Headquarters, Garching, Germany, 3–5 June 2019

Tony Mroczkowski 1 to provide guidance on what directions programme. Within Europe, ALMA activi-
Carlos De Breuck 1 we want the ALMA observatory to go in, ties are coordinated through ESO. Since
Ciska Kemper 1, 2 and how to get there. each region conducts and funds its activ-
ities differently, we focus here on the
EU-ALMA development programme. ESO
ESO Motivations strives to closely involve the Member
Academia Sinica, Institute of Astronomy State institutes in this programme by
& Astrophysics (ASIAA), Taipei, Taiwan ALMA1 is the world’s most sensitive facil- issuing calls for EU-ALMA development
ity for millimetre/submillimetre astronomi- studies every three years.
cal observations, and will soon be fully
The Atacama Large Millimeter/Sub­ operational in all of the originally planned In 2016, we decided to coordinate a
millimeter Array (ALMA) is the most bands (35–950 GHz). Over the last seven workshop on ALMA development with
sensitive observatory spanning milli­ years, ALMA has continuously delivered that year’s call for studies, and found the
metre and submillimetre wavelengths. exciting and often surprising results in all workshop to be an overwhelming success
To maintain this position, however, areas of astronomy2, from observations (Laing, Mroczkowski & Testi, 2016). The
a vibrant and concerted development of the first galaxies to the multiphase gas result was that proposals submitted to
programme is necessary. Since each in large-scale structures, and from form- that call were well focused on ALMA’s
partner region in ALMA conducts ing protoplanetary discs to observations development goals at the time. Following
its own development programme, we of the Sun. this success, we chose to host the 2019
hosted an international workshop at ALMA Development Workshop at ESO’s
ESO to promote further cross-regional ALMA is a collaboration among three headquarters in Garching, and timed the
discussion of our parallel development partner regions — Europe (through ESO), meeting to fall just after the announcement
efforts. As we describe here, an over- North America (USA, Canada and Taiwan), of the 2019 call for development studies.
riding goal for this was to align our and East Asia (Japan, Korea and Taiwan)
development activities with the goals of — with Chile. In order to keep ALMA at
the 2030 ALMA development roadmap the forefront of technology, each ALMA Figure 1. Key science drivers for ALMA, from the
— a report recently produced by ALMA partner has a continuous development ALMA Development Roadmap.

The Working Group proposes the following fundamental science drivers for ALMA developments over the next decade:

J. Pinto & N. Lira — ALMA (ESO/NAOJ/NRAO)


Trace the cosmic evolution of key elements Trace the evolution from simple to complex Image protoplanetary disks in nearby (150 pc)
from the first galaxies (z > 10) through the organic molecules through the process of star star formation regions to resolve the Earth
peak of star formation (z = 2–4) by detecting and planet formation down to solar system forming zone (~ 1 au) in the dust continuum
their cooling lines, both atomic ([C II], [O III]) scales (~ 10–100 au) by performing full-band at wavelengths shorter than 1 mm, enabling
and molecular (CO), and dust continuum, at a frequency scans at a rate of 2–4 protostars per detection of the tidal gaps and inner holes
rate of 1–2 galaxies per hour. day. created by planets undergoing formation.

Achieving these ambitious goals is currently impossible even with the outstanding capabilities of the current ALMA array. These science goals can
be achieved with the upgrades proposed in this document, upgrades that would make ALMA even more powerful and keep it at the forefront of
astronomy by continuing to produce transformational science and enabling fundamental advances in our understanding of the universe for the
decades to come.

64 The Messenger 177 – Quarter 3 | 2019

A major theme of the meeting was where – Enhancing the long-term capabilities correlator. This implies that a significant
ALMA should be by the year 2030. Since of the ALMA archive. number of the most compelling hardware
2016, the landscape for ALMA develop- – Performing exploratory studies on upgrades must be coordinated amongst
ment has changed. Band 1 (35–50 GHz) potential future development paths, several development groups, and there-
production is fully under way in East Asia, where ESO specifically prioritises feasi- fore the exact parameters of any potential
while the final Band 2 (officially 67–90 GHz, bility studies into the extension of the bandwidth improvement must be care-
though efforts towards a much wider maximum baseline length by a factor of fully defined.
bandwidth design are ongoing) is now 2–3, as well as the applicability of focal
entering Phase 1, which sees the con- plane arrays. Frédéric Gueth and Christophe Risacher
struction of six pre-production cartridges. from the Institut de Radioastronomie
With a clear path in place to complete the We note that the first two items have the Millimétrique (IRAM) and Keith Grainge
original suite of ALMA bands, ALMA lead- principal aim of increasing the observing from the Square Kilometre Array reported
ership has begun formalising the require- speed of ALMA, while the latter two on ongoing and planned development and
ments for the next decade of operations. increase the scientific capabilities of construction activities at their respective
This effort is known as the 2030 ALMA ALMA. observatories. We also heard about the
Development Road­map 3 and it was ongoing ALMA Band 1 (35–51 GHz) pro-
released in June 2018. duction, and the prototype development
Summaries of talks and highlights from for ALMA Band 2 (67–116 GHz), both of
The 2030 roadmap outlines three new sessions which had been mentioned earlier in the
key science drivers for ALMA (see context of completing the original suite
Figure 1), and provides four recommen- Here we summarise the invited talks and of ALMA bands. On the second day, the
dations. These are as follows (quoting highlight discussions from dedicated ses- programme included presentations on
directly from the roadmap)3: sions and contributed talks. All talks and the higher-frequency technologies being
1. Origins of Galaxies: Trace the cosmic posters are hosted on Zenodo 4 and are developed, which are especially applica-
evolution of key elements from the first linked to the workshop programme 5. ble to Bands 6–10, and receiver control
galaxies (z > 10) through the peak of star software that will find more optimal oper-
formation (z = 2–4) by detecting their The opening talks were delivered by the ational parameters for the existing ALMA
cooling lines, both atomic ([C II], [O III]) ESO Director General, Xavier Barcons, receivers.
and molecular (CO), and dust continuum, and the ALMA Director, Sean Dougherty.
at a rate of 1–2 galaxies per hour. These overviews were followed by a Next, the topics moved to the backend
2. Origins of Chemical Complexity: Trace summary of the ongoing work to update electronics that will take advantage of the
the evolution from simple to complex the ALMA receiver and backend electron- upgraded receivers. Crystal Brogan intro-
organic molecules through the process ics specifications, originally defined two duced the project to upgrade the current
of star and planet formation down to decades ago, and by regional overviews correlator — currently functioning with
Solar System scales (~ 10–100 au) by from each of the ALMA partners. Overall, decades-old hardware — to have higher
performing full-band frequency scans there was broad agreement with our spectral resolution and a doubling of the
at a rate of 2–3 protostars per day. vision for ALMA’s future. existing bandwidth, followed by much
3. Origins of Planets: Image protoplanetary discussion of more ambitious projects
discs in nearby (150 pc) star formation First and foremost, everyone agrees to build an entirely new correlator.
regions to resolve the Earth-forming that lower receiver noise temperature
zone (~ 1 au) in the dust continuum at is in ALMA’s interest, particularly in the On the final day, we heard about
wavelengths shorter than 1 mm, ena- so-­called workhorse bands — Bands 3, improvements to ALMA’s software
bling detection of the tidal gaps and 6 and 7. However, as we approach the and observing capabilities, including
inner holes created by planets in the limits of technology (current receivers improved proposal generation tools,
process of forming. have 4–10 times the quantum noise limit, improved observing modes (for example,
which is the ultimate limit for standard solar high-cadence and extended base-
The specific development priorities set receiver technologies), the noise in a fixed lines), and more sophisticated and rigor-
out in the ALMA Roadmap, ranked in bandwidth is dominated by the atmos- ous imaging and analysis techniques.
order of priority, are as follows: pheric contribution. The largest gains to Notably, the ALMA Science Archive pro-
– Broadening the receiver bandwidth by at be had in imaging speed might be easiest ject called “Additional Representative
least a factor of two, and upgrading the to obtain by expanding ALMA’s band- Images for Legacy” (ARI-L) is now under-
associated electronics and correlator. width (currently < 8 GHz). For example, way to bring the Cycle 2–4 data in the
– Upgrading the existing receiver bands, doubling the bandwidth would result in a ALMA archive up to the level of Cycle 5
where the highest priority is given to factor of two increase in imaging speed and later, directly addressing one of the
receivers operating in the 200–425 GHz for any observation requiring broad band- ALMA Development Roadmap priorities.
region, followed by receivers covering width or needing continuum sensitivity. We are in an era of data-driven science,
frequencies lower than 200 GHz, and, Any expansion of the receiver bandwidth and ARI-L will make the ALMA archive
finally, higher than 425 GHz. entails expansion of the backend elec- much more conducive to such work.
tronics — digitisers, data transport, and Also, a web-based replacement for the

The Messenger 177 – Quarter 3 | 2019 65

Astronomical News Mroczkowski T. et al., Report on the ESO Workshop “ALMA Development Workshop”

Figure 2. Workshop photo outside the ESO Demographics accepted. Submissions coming after this
­Supernova, showing that we chose a week of
deadline were given posters.
lovely weather in Garching.
The workshop was attended by 79 regis-
tered participants, only 12 of whom were
current Java-based ALMA Observing female (15%). The Science Organising Acknowledgements
Tool (ALMA-OT) is in the works and will Committee sought fair representation We thank Elena Zuffanelli for all her hard work in
improve compatibility going forward. from the community, but sadly the gen- coordinating the workshop logistics, Sandor Horvath
Most users will welcome this, as the vast der balance within the applications for technical support, Herbert Zodet for taking the
majority of user tickets related to the was particularly poor. We feel this reflects group photo, Rein Warmels for help with the work-
shop website and Evanthia Hatziminaoglou and Luca
ALMA-OT are essentially Java support much of the skewed gender ratio in Di Mascolo for help with the registration desk. This
issues, not issues with the tool itself. teams working on ALMA-related tech­ conference was supported by EASC.
nology development, which we are striv-
ing to improve. In future, groups within
Main conclusions & ways forward ALMA should be encouraged to increase References
the visibility of women within their groups,
Laing, R., Mroczkowski, T. & Testi, L. 2016,
Throughout the workshop, it was clear rather than consistently choosing only The Messenger, 165, 47
that ALMA is not a project that will be male group members to present group
neatly completed — which is the impres- efforts.
sion one sometimes gets as we race to Links
build the final bands in the original ALMA The balance for regional representation
band definition. Rather, ALMA is an was better, with attendees coming from ALMA observatory: https://www.almaobservatory.
ongoing and vibrantly active project. all three regions defined in the ALMA 2
ALMA Press Releases: https://www.almaobserva-
Only through continued development can collaboration and Chile in the following
ALMA remain a competitive, ground- percentages: 3
The ALMA Development Roadmap (Carpenter et
breaking observatory. The breadth of – 66% Europe; al., 2018):
work presented in the talks showed that – 14% North America; 4
Zenodo link to the workshop presentations: http://
the next decade will continue to be as – 11% East Asia;, https://
exciting for ALMA as the first has been. – 9% Chile (South America).
To achieve this ambitious goal, the ALMA search?page=1&size=50
The workshop programme:
executives showed a clear willingness Since this was a small workshop, all meetings/2019/ALMADevel2019/program.html
to collaborate. abstracts submitted by the deadline were

66 The Messenger 177 – Quarter 3 | 2019

Astronomical News DOI: 10.18727/0722-6691/5158

Report on the ESO Workshop

The VLT in 2030

held at ESO Headquarters, Garching, Germany, 17–21 June 2019

Antoine Mérand 1 cence ensures that the facility can be steady, albeit limited, flow of new instru-
Bruno Leibundgut 1 operated for many years to come. A ded- ments in the coming decade.
icated instrumentation programme is in
place for the coming decade. Five major topics were chosen for broad
ESO summaries and forward looks into the
next decade. It is always difficult to
Workshop structure predict the future, but the five reviewers
This four-day workshop offered a forum did an excellent job. Matthew Colless
to discuss the scientific future of the How do we make sure that the future of (Australian National University, Canberra)
VLT and VLTI. Overview talks of some of the VLT and VLTI remains science driven? reviewed the current and future plans in
the main scientific topics for the next The workshop was designed to address cosmology. He restricted himself to the
decade were followed by presentations this question by charting some of the determination of cosmological parame-
on the most important facilities operat- most exciting current research topics into ters and astrophysical contributions to
ing in 2030. Several instrument con- the next decade and deriving the neces- fundamental physics but excluded topics
cepts and ideas were presented which sary capabilities based on discussions like galaxy formation and evolution. In
would signifantly enhance the current throughout the workshop programme. the realm of multi-object spectroscopic
VLT and VLTI capabilities. The work- The workshop was attended by about facilities he acknowledged the significant
shop discussions are the basis for the 130 participants. Two approaches were potential of the 4-metre Multi-Object
plans for the VLT after 2025. followed. First, the workshop participants Spectroscopic Telescope (4MOST); he
heard reviews of five central research also pointed out that other projects, for
topics: cosmology, galaxy and black hole example, the Dark Energy Spectroscopic
Introduction evolution, resolved stellar populations, Instrument (DESI) at the National Optical
star and planet formation, and the Solar Astronomy Observatory (NOAO), are
The VLT/I ranks amongst the most pro- System. In the second part, facilities that more advanced and will begin observa-
ductive and most visible astronomical are due to be operating in the second tions sooner. Versatile facilities like the
facilities worldwide. As the world’s pre- half of the next decade were presented. VLT will become very important, should
mier ground-based facility, it provides a These include ALMA, the next generation a deviation from the currently favoured
powerful suite of visible and infrared of ELTs, European Space Agency (ESA) Λ-Cold Dark Matter (ΛCDM) model be
instruments, including unique capabilities space-based observatories, the James discovered. He noted the important con-
like coherent and incoherent combina- Webb Space Telescope (JWST), the tributions to the study of strong gravity by
tions of the four 8-metre Unit Telescopes Large Synoptic Survey Telescope (LSST), the VLT and the VLTI. It was pointed out
(UTs) and a multi-laser-guided adaptive the Cherenkov Telescope Array (CTA) and during the discussion that the VLT has
optics (AO) system. In combination with the Square Kilometre Array (SKA). The played critical roles in addressing some
the Atacama Large Millimeter/sub­ ESO status, in particular the current VLT of the most fundamental cosmological
millimeter Array (ALMA), VLT/I provides and VLTI setup (instrumentation, opera- problems, for example, providing crucial
comprehensive coverage of the electro- tions, calibrations, surveys) was also laid spectroscopy for supernova cosmology
magnetic spectrum from the ultraviolet out. This was followed by contributed projects and to probe the variation in
and visible through the infrared to sub­ presentations on science cases and fundamental constants such as the fine-­
millimetre wavelengths for the European potential new VLT instruments. All ses- structure constant. Looking to 2030,
astronomical community and its partners. sions had ample time set aside for dis- he felt that other facilities such as Euclid
cussions. The workshop programme is would become significantly more impor-
With the advent of ESO’s Extremely available online1. tant for cosmology than the VLT.
Large Telescope (ELT), the VLT and VLTI
will take on a new role. They will continue ESO Director General Xavier Barcons Linda Tacconi (Max Planck Institute for
to serve a large community and provide opened the workshop and gave an over- Extraterrestrial Physics [MPE], Garching)
unique data. The VLT’s and VLTI’s unique view of the current ESO situation and presented the current status of the very
capabilities are due to the versatile plans. He stressed that ESO’s ELT is cen- wide field of galaxies and black holes.
instrumentation on the four 8-metre tral to the organisation’s efforts in the Resolved properties, like rotation curves
telescopes, the spatial resolution achiev- coming years and cautioned the audi- or velocity dispersions, will gain in impor-
able by interferometry with baselines of ence that VLT developments would need tance to the understanding of the dynam-
over 100 metres, access to ultraviolet/ to proceed within the available resources. ics of early galaxies. She emphasised the
blue and optical wavelengths (including Within that envelope there is stable strength of integral-field spectroscopy
support from adaptive optics) and the funding for Paranal instrumentation for such studies and how well the VLT
flexible operations model. The telescopes developments and Paranal operations already caters to this type of research (for
and instruments have been maintained at (the ELT will eventually be integrated into example, the Spectrograph for INtegral
peak performance and new capabilities this operational paradigm). This implies Field Observations in the Near Infrared
continuously developed over the past two that there will not be a new generation [SINFONI], the Multi Unit Spectroscopic
decades. An ongoing programme to of VLT and VLTI instruments, rather a Explorer [MUSE], and the K-band Multi
avoid hardware and software obsoles- Object Spectrograph [KMOS]). Several of

The Messenger 177 – Quarter 3 | 2019 67

Astronomical News Mérand A., Leibundgut B., Report on the ESO Workshop “The VLT in 2030”

the planned instrument developments contrast Exoplanet REsearch instrument Starting in 2022, the LSST will provide
will be extremely helpful for this field (the (SPHERE) and an optical AO system, like many variable objects, which will need
Enhanced Resolution Imager and Spec- MAVIS, would be beneficial. The impor- dedicated follow-up observations. The
trograph [ERIS], and the MCAO-Assisted tance of coronagraphy was stressed. The VLT and many 4-metre telescopes will
Visible Imager and Spectrograph VLTI with GRAVITY and the Multi AperTure be prime facilities for the required spec-
[MAVIS]), and for some other studies pre- mid-Infrared SpectroScopic Experiment troscopy, as outlined by Pierre Astier
sented during the workshop. Of course, (MATISSE) is just beginning to tackle (Laboratoire de physique nucléaire et de
a better understanding of individual exoplanets, and better characterisations of hautes énergies [LPNHE], Paris). The
supermassive black holes is critical and planetary atmospheres can be expected. planning for the best usage of the various
GRAVITY, the AO-assisted, two-object, Transit measurements and follow-­up telescopes has already begun. Pierre also
multiple beam-combiner on the VLTI, observations of PLATO targets clearly presented an interesting connection with
represents a breakthrough in this represent another exciting research gravitational wave observations, suggest-
respect. Linda Tacconi emphasised the theme waiting to be explored. ing that the LSST could find many optical
importance of very high angular resolu- counterparts. The Cherenkov Telescope
tion observations of active galactic nuclei The importance of the VLT for Solar Array (CTA) is under construction and its
(AGN) as probes of galaxy evolution System objects was described by Heike status and plans were presented by
over cosmological times. Rauer (DLR, Berlin). The VLT provides Werner Hofmann (Max-Planck-Institut für
both the versatility and stability needed Kernphysik [MPIK], Heidelberg). It will be
The ESA space observatory Gaia has for regular and continuous observations exciting to search for optical counterparts
fundamentally changed our view of the of Solar System objects. While space of the many ultra-high-energy sources
Milky Way, its various components and missions will always provide more CTA will discover. Anna Bonaldi (Square
the dynamics of Local Group galaxies. detailed views, they are mostly limited to Kilometre Array Organisation) introduced
Eline Tolstoy (Kapteyn Institute Groningen) short periods (of the order of years). the SKA. It will be interesting to combine
presented the many open questions Ground-based observatories can provide the radio detections with optical sources.
stemming from the analysis of Gaia data. steady observations and long-term cov- Sofia Randich (Istituto Nazionale di Astro-
Follow-up spectroscopy will be extremely erage. They can also yield larger statisti- fisica [INAF], Arcetri) summarised the
important to complementing the posi- cal samples for asteroids, comets and Gaia-ESO survey. She underscored the
tional information with radial velocities trans-Neptunian objects. Special and importance of high spectral resolution and
and abundances of the stars to be pro- unforeseen events on timescales that blue efficiency for stellar abundance work.
vided by massive surveys with ground- prevent satellite missions, for example the A summary of the findings of the Public
based facilities. A most important aspect Shoemaker-Levy 9 impact on Jupiter, Survey Panel after its scientific review of
is covering the ultraviolet for the study of will heavily rely on ground-based obser- the completed and ongoing ESO public
the nucleosynthesis of heavy elements. vations. Synergetic observations to com- surveys in May 2019 was delivered by
4MOST already dedicates a large fraction plement future planetary missions, like Bruno Leibundgut (ESO).
of its observing time to Gaia follow-up. the JUpiter ICy moons Explorer (JUICE)
Other instruments of interest are the Fibre or Europa Clipper, will enhance the sci- Several talks provided background infor-
Large Array Multi Element Spectrograph ence return of these space missions. The mation on the current VLT/I situation. The
(FLAMES) and the Multi-Object Optical and best angular resolution is clearly an asset instrumentation planning for the VLTI
Near-infrared Spectrograph (MOONS). But for observations of Solar System bodies. (Antoine Mérand, ESO) and the VLT
there remains a gap in respect of fainter (Bruno Leibundgut, ESO) were followed
stars and the need for higher spectral Summary talks on existing and future by presentations on data flow operations
resolution to better constrain the stellar facilities set the stage for synergies and (Michael Sterzik, ESO), Paranal opera-
parameters and abundances with a complementarity. Ciska Kemper (ESO) tions (Steffen Mieske, ESO) and calibra-
potential future VLT multi-­object high-res- presented some science results based tion plans and issues (Alain Smette,
olution spectrograph. on the synergy between ALMA and VLT ESO). New ideas include atmospheric
data and described the ALMA2030 forecasting to allow planning of observa-
Our knowledge of planet formation con- Development Roadmap. The ELTs will be tions in more detail, or at least imple-
tinues to develop rapidly. Anne-Marie the flagship ground-based near-infrared menting “now-casting” to obtain a full
Lagrange (Institut de planétologie et observatories and their strengths were understanding of the current status for
d’astrophysique de Grenoble, IPAG) outlined by Michele Cirasuolo (ESO). The real-time scheduling. A move away from
reviewed the current situation; many exciting ESA programme for the coming standard stars to physical atmosphere
protoplanetary disks are known but the decade was detailed by Fabio Favata models for calibration of the atmosphere
number of directly imaged planets (ESA). The JWST launch is planned in was also presented.
remains small. Clearly, sensitivity and early 2021 and Gillian Wright (UK Astron-
higher angular resolution are key for this omy Technology Centre [UKATC], Lively discussions followed these ses-
field. The ELTs (and JWST) will be major Edinburgh) described its scientific plans sions. The importance of adequate
players, but there remain various capabil- and capabilities. With a 10-year planned preparation for LSST transient follow-up
ities which the VLT and VLTI can offer. lifetime, the JWST and VLT will comple- was stressed several times. The comple-
An upgraded Spectro-Polarimetric High-­ ment each other for many projects. mentarity of JWST infrared imaging and

68 The Messenger 177 – Quarter 3 | 2019

Figure 1. Workshop par-
ticipants enjoy a break
in front of the ESO
Supernova Planetarium
& Visitor Centre.

low-resolution spectroscopy with ground- ment concepts, a wide range of opinions has become clear that the VLT/I will
based infrared high-resolution spectros- and positions was voiced. The workshop remain a facility that can serve many dif-
copy was highlighted, although this will closed with a summary by Denis Mourard ferent interests and science applications.
probably be more in synergy with the (chair of the Scientific Technical Commit- This versatility should be maintained as
ELTs than with the VLT. In general, better tee [STC], Observatoire de la Côte d’Azur far as possible. At the same time, VLT/I
coordination between ground- and [OCA]); he outlined a first categorisation operations may open up to more specific
space-based observatories was urged. of the different topics, which will be the experiments or projects, for example
The programmatic aspects of VLT and basis for future discussions. through visiting instruments. The planning
VLTI instrumentation were presented by of the future VLT instrumentation comple-
Luca Pasquini (ESO) setting the frame- ment needs to account for the existing
work for the 21 contributed talks outlining Demographics (and aging) instrumentation; some instru-
science cases and concepts of new ments may not be maintainable at rea-
instruments. Individual presentations on The Science Organising Committee was sonable cost for another decade. A sum-
the science plans of MOONS (Michele composed of one Council and four STC mary of the concepts presented at the
Cirasuolo, ESO), MAVIS (Richard members as well as relevant ESO staff. workshop together with a first scientific
McDermid, Macquarie University Sydney) Its gender distribution was 4 female assessment will be prepared for the next
and an ultraviolet spectrograph (Chris and 7 male members. Among the invited STC meeting in October for further dis-
Evans, UKATC Edinburgh) started off this speakers 8 out of 20 were women (4 out cussion. A concrete plan for VLT and VLTI
part of the programme. Several new and of the 5 reviews). However, among the after 2025 will be prepared by the Pro-
exciting instrument ideas on different 45 submitted contributions only one was gramme Scientists and presented for rec-
scales were presented, which will further by a woman and there were 24 women ommendation to the STC in April 2020.
enhance the VLT and VLTI capabilities. among the 130 participants. While this With the long-term planning in hand, the
gender ratio reflects that of the instru- next instrumentation studies can start to
The workshop programme provides an mentation community, we hope that be fully operational some time during the
overview of the newly proposed projects these stark statistics act as a wake-up second half of the coming decade.
and their science cases and we refer the call to ESO instrument builders. On the
reader to that list for a summary. There timescales explored by this meeting,
were also 16 posters2 on display, as not positive action to ensure more inclusion Acknowledgements
all requested talks could be accommo- in our community can and should be We would like to thank the SOC members for their
dated in the workshop programme. It is undertaken. help in preparing the scientific programme and dur-
too early to discuss any of these propos- ing the workshop. The logistics were handled
als, suggestions and ideas in detail. superbly by Svea Teupke.
However, some trends can already be Outlook
discerned. There are a few concepts for
instruments with a wide range of astro- The workshop successfully laid out the
physical applications. Others cater to scientific landscape and explored the 1
Workshop programme:
more specific science cases or have a interests of the community regarding meetings/2019/VLT2030/program.html
relatively narrow science focus. The VLT/I developments. The enthusiasm dis- Workshop poster papers:
workshop finished with an extensive dis- played by the community presentations
cussion. Because of the different levels on the scientific potential and the ideas
of detail in the science cases and instru- for new instrumentation was obvious. It

The Messenger 177 – Quarter 3 | 2019 69

Astronomical News DOI: 10.18727/0722-6691/5159

Fellows at ESO

Chentao Yang Chentao Yang

Growing up in one of the most remote

cities from the sea, Urumqi, deep in
northwestern China, I was fascinated by
astronomy since my early childhood,
just like hundreds of millions of other
boys on this planet. I can still remember
one night more than 20 years ago, at
nine-years old, I was copying a table of
the temperatures and fluxes of the bright-
est stars in the sky from a book with my
friend. I would come to consider that
moment the beginning of my scientific
career in astronomy. My love of astron-
omy just kept growing over my childhood,
partially because of my stubbornness
and persistence.

Four years later, in junior high school, I

started to write letters to the author of
that same book — the then director of
Yunnan Observatory — asking astronomy
questions. I was incredibly grateful that
he responded to all my mails carefully.
Among those letters, I asked how I
could get prepared to be a professional (ALMA) and that it would become the strongly lensed starbursts from large-
astronomer. He replied with detailed most powerful telescope in the near area Herschel SPIRE maps. We quickly
and insightful suggestions covering two future. I dreamed of using it one day. built the largest sample of the H2O-line-
pages that have impacted my life ever After I graduated, I entered the masters detected sources across cosmic time.
since. Later, in high school, my passion programme at the same university.
for astronomy helped me join the astron- After obtaining my masters degree, I
omy club. The teacher responsible for By coincidence Gao Yu happened to entered Purple Mountain Observatory for
the club was Gao Xing — one of the most meet Alain Omont from Paris during that a PhD programme with Yu. And to con-
famous amateur astronomers in China. time. Alain was working on a new project tinue our study of the sample of strongly
With him, we learned a lot about basic that could lead to a masters student lensed high-redshift starbursts, I started
concepts in astronomy and had many ­project. So I was fortunately introduced a co-tutelle (joint) PhD programme
unforgettable experiences of observa- to Alain and started a research pro- between Université Paris-Sud and Uni-
tions such as eclipses, meteor showers, gramme co-supervised by Yu and Alain. versity of Chinese Academy of Sciences
the transit of Venus and ­Messier mara- Around that time, submillimetre water from 2014. I moved to Paris from Nanjing
thons (during which amateur astronomers vapour lines were coming back to peo- that year and started my research pro-
try to find as many Messier objects as ple’s attention because of the launch jects under the supervision of Alain
possible in one night). of the Herschel Space Observatory. Its Omont and Alexandre Beelen. Université
instruments the Spectral and Photo- Paris-Sud is in Orsay, the beautiful south-
Naturally, after high school, I chose one metric Imaging Receiver (SPIRE) and the ern urban area of Paris. I started a sys-
of the few universities in China that has Heterodyne Instrument for the Far Infra- tematic investigation of the physical con-
an astronomy department to pursue red (HIFI) enable us to detect a massive ditions of the cool interstellar medium in
my passion for a career in astronomy. I number of submillimetre H2O lines in the high-­redshift strongly lensed star-
was very serious about becoming an our Galaxy and in nearby galaxies. These burst galaxies using different telescopes.
astronomer. During the last year of my H2O lines are a unique and essential With the 30-metre telescope of the Insti-
undergraduate life, I started to study the diagnostic tool that helps us understand tut de Radioastronomie Millimétrique
cold dust emission in galaxies with data better the conditions in the interstellar (IRAM), one of my favorite telescopes, I
from the Submillimetre Common-User medium. So I started to investigate those had ­several observing runs lasting weeks
Bolometer Array (SCUBA) on the James H2O lines using Herschel Science Archive to study the CO ladder of those galaxies.
Clerk Maxwell Telescope, under the data, by collecting data from all of the We also obtained a massive amount of
­mentorship of Gao Yu from Purple Moun- galaxies that had been observed. At the data from the Plateau de Bure Interferom-
tain Observatory in Nanjing. It was also same time, with Alain, we also started eter, from which I gained most of my
then that I learned about the Atacama to target those bright H2O lines in the knowledge of interferometry. Unfortu-
Large Millimeter/submillimeter Array high-redshift Universe, by picking up nately, I never had a chance to see the

70 The Messenger 177 – Quarter 3 | 2019

Astronomical News

antennas on the plateau. However, on my thrilled that I would be working with acquired images of dust and gas emis-
28th birthday I observed with the James ALMA, the most powerful (sub)millimetre sion at scales of 100 pc for dusty
Clerk Maxwell Telescope and finally saw telescope that I dreamt about back in ­galaxies when the Universe was about
world-famous telescopes like Keck with 2010. My dream came true and I moved two billion years old. I am also using the
my own eyes. It was an unforgettable to Chile in November 2017. NOrthern Extended Millimeter Array
experience. With the Karl G. Jansky Very (NOEMA) of IRAM and ALMA conducting
Large Array (VLA), we studied the At ESO, I spend 50% of my time at ALMA spectral line surveys of high-redshift
­properties of dense gas in those lensed performing functional duties. I still ­galaxies, pushing the limit of astrochem-
­starbursts and, little by little, I realised remember my first trip to the 5000-metre- istry studies to the early Universe.
that my passion is for astronomy obser- high array operations site of ALMA. Besides, the submillimetre H2O emission
vations, including the joy when the pro- The landscape is simply Martian. I have from galaxies has been one of my
posal is accepted, and the excitement of enjoyed participating in the science oper- areas of interest. With the Atacama Path-
checking freshly acquired data. ations at ALMA a lot, where I keep learn- finder Experiment telescope (APEX),
ing every day and work to contribute to we achieved the first detection of the
During the last year of my PhD, when I ALMA. During the other half of my time, I 752-GHz H2O line in extragalactic sys-
received an offer from ESO for a fellow- continue my research into molecular gas tems from the ground.
ship in Chile in January 2017, I was and dust in galaxies. Using ALMA, I

DOI: 10.18727/0722-6691/5160

External Fellows at ESO

In addition to the ESO fellowships, a largely unplanned, it has been an immense My transition then began in earnest. I
number of external fellows are hosted at pleasure, and I leave as a more enriched chose to continue at Cambridge with a
ESO. Profiles of two of these fellows are and fulfilled person than when I arrived. master’s degree in astronomy, learning
presented here. the fundamentals of the subject from a
I was born and raised in London, where mostly theoretical perspective. I then
I spent my childhood enjoying football spent a year as a European Space
Prashin Jethwa (Liverpool), Pokémon (Charizard), and ice Agency (ESA) Young Graduate Trainee in
cream (all varieties). Notably absent from Madrid, where I got to experience a more
It could so easily not have happened at this list is astronomy. Perhaps I can “hands-on” side to astronomy. My project
all! My two-year stint at ESO has been a blame urban light pollution, but I cannot at ESA consisted of modelling overexpo-
fortunate coincidence. Sidestepping the claim to have been especially awestruck sures on the cameras of XMM-Newton,
usual route taken by fellows, my voyage by the Universe in my formative years. a space-based X-ray telescope. During
through the seas of ESO has been at the Rather than looking up through a tele- this time, I also enjoyed my first look
command of a brave captain: Glenn van scope, I kept my head down, often in a through a telescope. After an impromptu
de Ven. I joined Glenn’s group, funded mathematics textbook. This is what really 100-kilometre drive south of Madrid with
through a European Research Council absorbed the academic side of my youth- a friend’s 20-inch Dobsonian telescope, I
grant, for a position which was originally ful brain: maths problems, puzzles and... saw Saturn’s rings, Jupiter’s moons and
intended to be hosted at the Max Planck polynomials? This led me to the University made amends for my youthful oversight.
Institute for Astronomy in Heidelberg. of Cambridge, where I completed an Having ticked this box, gained some sub-
However, along with Glenn, my position undergraduate degree in mathematics. stantial research experience, and seen
moved to ESO Garching, where I have It was a broad curriculum, spanning part of the wider astronomy community, I
been based since October 2017. No aspects of pure and applied maths as felt ready to move on to the next step.
sooner have I learnt to navigate through well as theoretical physics. The latter
the ESO headquarter buildings, however, topic dominated my choice of courses For my doctoral studies, I returned to the
than my time here has come to an end. in the final year, reflecting my evolving Institute of Astronomy in Cambridge.
I will soon move to re-join the newly interest in mathematics: not just as an Under the supervision of Vasily Belokurov
appointed Professor van de Ven, this time abstract puzzle, but a tool for modelling and Denis Erkal, I completed a thesis
at the Institute for Astrophysics in Vienna. real phenomena and solving real problems. about the Milky Way halo. The halo refers
Despite, then, my time at ESO having been to the region out to distances of a few

The Messenger 177 – Quarter 3 | 2019 71

Astronomical News Jethwa P., Oikonomou F., External Fellows at ESO

hundred kiloparsecs from the Galaxy. It is Prashin Jethwa

filled with satellite galaxies, star clusters,
and diffuse clouds and streams of stars,
all orbiting around the Milky Way. Struc-
tures in the halo are remnants of smaller
galaxies merging with the Milky Way over
timescales of billions of years. By dis-
entangling and characterising these
structures we can see into the Galaxy’s
past evolution and growth.

One main result from my thesis concerns

newly discovered dwarf galaxies. Early in
my PhD, two teams raced to report the
discovery of dozens of dwarf galaxies in
the Dark Energy Survey (DES), a photo-
metric survey in the southern hemisphere.
The number of dwarfs discovered in DES of past galactic mergers. Decomposition nar series. I leave with an unexpected
far exceeded expectations from previous, is therefore very powerful, but it can be friendship with the owner of the local
similar surveys in the north. One possible computationally challenging and will Getränkemarkt (bottle shop) having
explanation for this overabundance was become increasingly challenging as data bought hundreds of crates for our weekly
that the new dwarfs were associated with quality improves. Some of the highest after-work Beer Fridays. On display here,
the Magellanic Clouds, the largest of the quality IFU data currently come from every Friday, is the warm atmosphere
Milky Way satellites, which lay close the the MUSE instrument on the VLT and, and friendly, interesting people who really
boundary of the DES footprint. I tested looking to the future, the High Angular make ESO a pleasure to be part of.
this hypothesis by building a dynamical Resolution Monolithic Optical and Above all, I leave ESO with countless
model of the Magellanic Clouds, Milky Near­-infrared Integral-field spectrograph good memories and friends.
Way and dwarf galaxies, and devising (HARMONI) on the Extremely Large
a statistical framework to compare this ­Telescope (ELT) will improve spectra and
model to observations. The results con- spatial resolution yet again by orders of Foteini Oikonomou
firmed that most of the DES dwarfs were magnitude. To prepare for this increase
likely to be associated to the Clouds, and in the size and quality of data, I have I was born and grew up in Athens,
furthermore predicted their velocities, been investigating dimensionality-reduc- Greece. I cannot remember a fascination
several of which have since been corrob- tion techniques to allow us to perform with the night sky, perhaps because of
orated with follow up observations. This decomposition for large datasets. Along- the light-pollution in Athens, until at an
work encapsulates the main tools used side many other examples, this project age of around nine I visited the observa-
in my research, which are mainly theoreti- has expanded my scientific horizons. tory of Penteli during one of the open
cal, with plenty of dynamics, and using evenings. That evening we observed
statistical modelling techniques. In addition to being scientifically eye-­ the Hercules globular cluster. Hearing the
opening, in my two years at ESO I have astronomers there describe the system
Moving to ESO, my scientific focus has grown in several other ways. The stu- we were looking at and explain the time it
shifted slightly. Whilst still interested in dents and fellows I have met in Garching had taken the light to reach us made me,
the evolution and growth of galaxies, I are an exceptionally proactive and for the first time, very aware of the vast-
now focus on those outside of our local engaged group of people. We take excel- ness of the Universe and filled me with
neighborhood. The type of data available lent advantage of opportunities provided a sense of wonder.
for studying more distant galaxies is very to us and take the initiative to create
different from data we have in the Milky further opportunities for ourselves and During my teenage years, my interest in
Way. Some of the richest datasets come future ESO scientists. Seizing these the Universe grew, mostly from reading
from Integral Field Units (IFUs). These opportunities has been a key part in my popular science books and occasional
observe not just 1D spectra, or 2D personal development. I leave with teach- visits to the observatory. Those were the
images, but 3D data-cubes: images ing experience, having co-supervised days without internet, and I can remem-
where every pixel of the image contains a an undergraduate summer student ber that I got into the habit of scanning
spectrum. A powerful method to study in the inaugural ESO Summer Research the local press for information about the
galaxies using IFUs is decomposition, i.e., Programme, and PhD student Meghan next open evening at the observatory,
breaking the data-cube down into parts Hughes who is enrolled in the ESO stu- and the frustration I felt when I couldn’t
that represent physical components. This dentship programme. I leave with find anything scheduled. In high school,
can reveal the history of when stars were enhanced organised skills, having organ- I decided to take mostly science courses
formed, how the chemistry of gas and ised the ESO Garching Science Day, so that I could apply for a university
stars evolves, and uncover the remnants group meetings and one of several semi- degree in astrophysics. I chose my high

72 The Messenger 177 – Quarter 3 | 2019

school diploma thesis to be on black of my PhD, I was offered a postdoctoral blazars, and I was very lucky to be able
holes because they captured my curiosity position at the Pennsylvania State Uni­ to research this topic in the midst of
the most. I think that Stephen Hawking’s versity (Penn State). There, I was able to world experts on blazars and on high-­
popular science books were what sparked work alongside Miguel Mostafa and energy neutrinos.
this interest. I have since confirmed that Stephane Coutu on the Pierre Auger
this was the case for many of the fellow Observatory. This is the largest ultra- At ESO and the surrounding institutes
astrophysics students of my generation high-energy cosmic ray detector ever in Garching, the opportunities to grow as
at university. built. Distributed particle counters cover a scientist seem endless. I heard some-
an area of 3000 square kilometres in where that Munich is the city with the
I went to England to study astrophysics the pampa, in the Mendoza region of largest number of astrophysicists in the
at University College London (UCL). I Argentina. I was lucky to be able to visit world. I do not know if this is true, but
made the decision because there was no the experiment several times and to it certainly feels that way. At ESO and in
bachelor’s degree in astrophysics in ­operate the fluorescence telescopes dur- Garching I enjoy the very rich variety of
Greece. It seemed to me at the time that ing shifts. During my time at Penn State, excellent weekly talks, and other regular
studying physics instead wouldn’t be as I also worked alongside Kohta Murase on events, the Joint Astronomy Colloquium,
much fun. It was in London that I truly got jetted active galaxies, called blazars. Active Galactic Nuclei (AGN) Club,
acquainted with astronomy. The degree I These objects are powered by the Uni- ­Journal Club, and the various weekly
took involved many astrophysics courses, verse’s most massive black holes and seminars of the many institutes in
but also lots of time using telescopes and have long been thought to be sources of Garching. My position is funded by the
analysing astronomical data at the Uni- ultra-high-energy cosmic rays. DFG (German Research Foundation)
versity of London Observatory. I am still ­collaborative research centre SFB1258:
fascinated by the dedication and excite- I next moved to ESO, to my current posi- Neutrinos and Dark Matter in Astro- and
ment of all the astronomers there, for the tion, to work alongside Paolo Padovani Particle Physics. Within this programme,
great research they manage to do, just and Elisa Resconi on blazars, and on I am able to participate in additional
next to the A1 motorway in the suburbs high-energy neutrinos, expanding my ­regular cross-­institute colloquia, lectures,
of the often cloudy London! multi-messenger expertise so as to tackle and workshops, collaborative gatherings
the ultra-high-energy cosmic ray problem with artists, and outreach activities.
In my final years at university, I became from additional directions. The time of my While writing these lines I realised that
most interested in extragalactic astro- arrival in Munich was very opportune not much has changed in my main
physics and elementary particle physics. because, within a week of the start of my ­astrophysics interests since as a teen-
I decided that I’d like to pursue a PhD on position at ESO, the IceCube neutrino ager I got excited by the mysteries of
a topic which combined elements of detector registered an alert for a high-­ the high-energy Universe, and the mys-
both. At the time, UCL launched an energy neutrino coincident with a flare teries of black holes. I feel very ­fortunate
initiative called the Institute of Origins, from the powerful blazar, TXS 0506+056. to be able to nourish this curiosity every
designed to foster collaboration between This event probed in-depth investigations day.
the Particle Physics and Astrophysics on neutrino emission from this and other
groups at UCL. In this platform, I found a
PhD topic that greatly interested me: an
investigation on the origin of ultra-high-
energy cosmic rays. These fascinating
messengers of the extreme Universe,
which are most likely extragalactic, are
the most energetic particles known.
They possess energies ten million times
higher than the particles that can be
accelerated at the Large Hadron Collider.
We do not know by what astrophysical
objects they are accelerated, and to this
day, this question has been the back-
bone of my research endeavours.

I was fortunate enough to be offered

a PhD position at UCL. I still remember
the offer email I received as one of the
happiest moments of my life. I worked
alongside Ofer Lahav, Amy Connolly, and
Kumiko Kotera on the signatures of plau-
sible astrophysical sources of ultra-high-
energy cosmic rays. Towards the end Foteini Oikonomou

The Messenger 177 – Quarter 3 | 2019 73

Astronomical News DOI: 10.18727/0722-6691/5161

Lodewijk Woltjer (1930–2019)

Daniel Hofstadt 1 ity, which he decided should be hosted at in his book “Europe Conquest of the Uni-
the ESO Headquarters in Garching. verse”. With Roger Bonnet, ESA’s Direc-
tor of Scientific Programmes, he wrote
ESO However, clearly his leading achievement “Surviving 1000 centuries. Can we do it?”
remains the Very Large Telescope (VLT) – a perfect illustration of his extended
conceptual design outlined in his blue range of interests.
“Lo” Woltjer (b. Noordwijk 1930, d. 2019) book and subsequently approved by ESO
was ESO Director General from 1975 to Council in 1987. The new observatory site He was not an easy guest on planet Earth.
1987. Despite originally studying geology, was selected on the Paranal mountain. In The suitcase in his hand was as much a
Lo was to become one of the youngest the VLT design, Lo included an interfero- symbol as a travel outfit. He hated public
professors of astronomy in the Nether- metric mode, an option which was not exposure and confrontations, maintaining
lands. From there he went to the USA to met with universal enthusiasm in the sci- a formal distance behind an elegant style.
chair the department of astronomy at entific community at the time but has It was not always easy to guess what he
Columbia University from 1964 to 1974. subsequently proved to be visionary. had in mind, often hiding behind non­
He related to ESO early on and realised chalant gestures and leaving to others the
the potential an international organisation Lo Woltjer became President of the Inter- challenge of interpreting his purpose.
had for European astronomy, in keeping national Astronomical Union and later Highly respected, at times very much
with the spirit of its founders. served as Chair of the ESA Space Science admired for his courage, he could be very
Advisory Committee. He was instrumental stubborn in imposing his will. In his func-
Lo Woltjer’s scientific interests were ini- in the development of the European Astro- tion as ESO Director General he also had
tially focused on the Crab Nebula super- nomical Society (EAS), whose annual lec- to make diplomatic compromises, but
nova remnant. He also made extensive tures carry his name. He was honoured most of the time he managed to impose
contributions to the study of quasars and with the Karl Schwarzschild medal of the his own views. For those who had the
magnetic fields in stars and galaxies. Astronomische Gesellschaft (German privilege to interact with him, whether
Astronomical Society) and was a member under his leadership or privately, he was a
When he was appointed ESO Director of several national science academies. beacon and a reference point.
General in 1975 he reshaped the Organi-
sation in line with a unique ambition, Lo was also a lover of the Sun with a pas-
emphasising in-house technical develop- Who was Lo Woltjer? sion for the outdoors, whether in Saint
ments initiated with the support of CERN. Michel l’Observatoire, on canoeing trips
He later engaged ESO in collaborations Lo was an enigmatic personality with a with his daughter Eleonore, or hiking and
with industry and astronomy institutions strong determination to promote Euro- swimming at Lago Rupanco in the south
across Europe. In his proposal for pean astronomy and raise ESO to the of Chile. Six months ago, his wife Ulla
the ESO directorship he conditioned his status of a world class Institution. He was died. She was of outstanding assistance
acceptance on the creation of a scientific a natural leader with outstanding mana- to him and provided an enjoyable link to
department at ESO in order to anchor gerial skills. If intelligence is defined as his social environment both in Europe
the role of Organisation in the European the ability to convert raw information into and Chile. His health failed over the last
community. practical and challenging developments, few months of his life, his legs gave way
he must be portrayed as an exceptional and he broke a shoulder. With Ulla gone
During his thirteen years as Director Gen- masterminding individual. He fully identi- he lost interest in life and slowly faded
eral ESO became the world-class institu- fied with his mission. Everybody con- away. He was 89 years old.
tion it is today. The financial contribution nected to the Organisation was aware of
of its Member States increased vastly that simple equation: Woltjer was ESO
over his tenure. Italy and Switzerland and ESO was Woltjer. Some of his strug- Figure 1. Lodewijk Woltjer welcoming guests at the
joined the Organisation, followed by other gles and achievements are summarised inauguration of the headquarters at ESO Garching.
European countries. The Max-Planck-
Gesellschaft (MPG)/ESO 2.2-metre tele-
scope and the Swedish–ESO Submilli-
metre Telescope (SEST) antenna saw
their first light at La Silla observatory as
well as an ever-larger set of advanced
instrumentation during his tenure at ESO.
The New Technology Telescope (NTT)
was designed thanks to the joining fees
provided by Italy and Switzerland when
they joined ESO. Lo Woltjer also estab-
lished an interface with the European
Space Agency (ESA) through the Space
Telescope European Coordinating Facil-

74 The Messenger 177 – Quarter 3 | 2019

Astronomical News

Personnel Movements

Arrivals (1 July– 30 September 2019) Departures (1 July– 30 September 2019)

Europe Europe

Adams, Henning (DE) Mechanical Engineer Avila, Gerardo (MX) Engineer/Physicist

Bazin, Gurvan (FR) Web & Advanced Projects Coordinator Chen, Chian-Chou (TW) Fellow
Beuchert, Tobias (DE) ESO Supernova Presenter Circosta, Chiara (IT) Student
Calistro Rivera, Gabriela (PE) Fellow Cosentino, Giuliana (IT) Student
Caseiro de Almeida, Álvaro José (PT) ESO Supernova Technical Assistant Fensch, Jérémy (FR) Fellow
Chaturvedi, Avinash (IN) Student IMPRS Jeřábková, Tereza (CZ) Student
Hayden-Pawson, Connor (UK) Student Käufl, Hans-Ulrich (DE) Infrared Instrument Scientist
Hofer, Josef (AT) Software Engineer Klitsch, Anne (DE) Student
Hoffstadt Urrutia, Arturo (CL) Software Engineer Querejeta, Miguel (ES) Fellow
Huber, Florian (DE) Logistics Officer Sagatowski, Jakob (SE) Software Engineer
Kabátová, Anežka (CZ) Student Whitehouse, Lewis James (UK) Student
Lamperti, Isabella (CH) Student
Miles Páez, Paulo Alberto (ES) Fellow
Pfuhl, Oliver (DE) Optical Engineer
Popping, Gergely (NL) ALMA Regional Centre Astronomer
Ramírez Molina, Andrés (CL) Software Engineer
Richerzhagen, Mathias (DE) Optical Engineer
Szostak, Artur (PL) Software Engineer
Zsidi, Gabriella (HU) Student

Chile Chile

Anania, Andres (CL) Software Engineer Dias, Bruno (BR) Fellow

Cea, Victor (CL) Electronics Group Leader Gallenne, Alexandre (FR) Fellow
Gran, Felipe (CL) Student Hau, George (UK) Operation Staff Astronomer
Herenz, Edmund Christian (DE) Fellow Hüdepohl, Gerhard (DE) Electronics Group Leader
Kravchenko, Kateryna (UA) Fellow Lillo Box, Jorge (ES) Fellow
Macías, Enrique (ES) Fellow Messias, Hugo (PT) Fellow
Olguin, Rodrigo (CL) Head of Engineering Department Minniti, Javier (AR) Student
Ribas, Alvaro (ES) Fellow Razza, Alessandro (IT) Student
Solarz, Aleksandra (PL) Fellow Rojas, Alejandra (CL) Student
Vial, Sofia (CL) Procurement Officer Toledo, Pedro (CL) Software Engineer
Villenave, Marion (FR) Student

The ESO Annual

Report 2018 is
available online
now at www.eso.

The Messenger 177 – Quarter 3 | 2019 75

76 The Messenger 177 – Quarter 3 | 2019

Related Interests