The Messenger 164

Solar activity and instrument data quality
VISIR Upgrade Science Verification

XXL Survey: First results
LEGA-C Survey: First data release
The Messenger
No. 164 June 2016
Telescopes and Instrumentation
Adaptive Optics Facility Status Report: When First Light

Is Produced Rather Than Captured
Robin Arsenault 1
Pierre-Yves Madec1
Elise Vernet 1
Wolfgang Hackenberg1
Domenico Bonaccini Calia1
Paolo La Penna1
Jrome Paufique1
Harald Kuntschner1
Jean-Francois Pirard1
Marc Sarazin1
Pierre Haguenauer 1
Norbert Hubin1
Ignacio Vera1
1
ESO
First light for the 4 Laser Guide Star

Facility (4LGSF) took place in Paranal
on 26 April 2016 with four laser units in
operation for the first time. A combined
test with the first laser guide star unit
and the Ground Layer Adaptive optics
Assisted by Lasers (GRAAL) instrument
in October 2015 demonstrated the whole
acquisition sequence of the Adaptive
Optics Facility (AOF). Many tools that will
support the operation of the AOF for science observations have meanwhile been
implemented. GALACSI was granted
Provisional Acceptance in Europe in
April 2016, completing the system tests
and qualification in Garching of the
adaptive optics modules GRAAL and
GALACSI (Ground Atmospheric Layer
Adaptive Optics for Spectroscopic Imaging), their real-time computers and the
deformable secondary mirror (DSM).
Results of tests both in the laboratory
and on sky are presented. The installation of the DSM and GALACSI will be
completed by early 2017, to be followed
by commissioning of all AOF systems.
ESO Project Team:

P. Duhoux, J.-L. Lizon, S. Guisard, P. Lilley, L. Petazzi,
P. Hammersley, I. Guidolin, L. Kern, T. Pfrommer,
C. Dupuy, R. Guzman, J. Quentin, M. Quattri, R.
Holzlhner, D. Popovic, M. Comin, S. McClay, S.
Lewis, F. Gago, J. Kolb, A. Jost, J. Argomedo, S.
Tordo, R. Donaldson, R. Conzelmann, M. Lelouarn,
R. Siebenmorgen, M. Downing, J. Reyes, M. Suarez
Valles, S. Strbele, S. Oberti, P. Gutierrez Cheetam,
M. Kiekebusch, C. Soenke, E. Aller-Carpentier,
P. Jolley, J. Vernet, A. Manescau-Hernandez, L.
Mehrgan, G. Calderone, A, van Kesteren, G. Chiozzi,
H. Sommers, D. Dorigo, T. Bierwirth, J.-P. Kirchbauer,
S. Huber, G. Fischer, A. Haimerl, S. Lvque, P.
Amico, G. Hubert, S. Brillant, P. Baksai, J. C. Palacio,
I. Munoz, E. Fuenteseca
The Messenger 164 June 2016
Chronicle of recent AOF activities

The Adaptive Optics Facility (Arsenault et
al., 2010; 2014) is a long-term project on
the Very Large Telescope, Unit 4 (UT4) to
provide the adaptive optics systems for
the two UT4 instruments MUSE (the Multi
Unit Spectroscopic Explorer), fed by
GALACSI, and the High Acuity Wide field
K-band Imager (HAWK-I), fed by GRAAL
with a four sodium laser guide star system. After years of planning, construction
and testing, with the completion of reviews
for Provisional Acceptance Europe (PAE)
of AOF modules in the course of 2015,
activities are now shifting to Paranal.
The first laser guide star unit (LGSU#1)
was installed and tested on UT4 in 2015,
with first light in April 2015, and the three
other units were then completed. The
LGSU#1 commissioning was completed
in August 2015 and the unit showed
excellent performance. It confirmed the
good design choices and validated all the
interfaces with UT4. PAE for the 4LGSF
system was granted at the end of 2015
and in January 2016 the three remaining
LGSUs were re-integrated in Paranal and
installed on UT4. On 26 April 2016 first
light for the 4LGSF took place (see cover
image and Release eso1613) and the
commissioning of the 4LGSF in standalone mode could begin.
In the meantime, a two-year-long system
testing phase was concluded in Garching
(February 2016). The adaptive optics (AO)
modules GRAAL and GALACSI were
mounted in that order on the ASSIST

(Adaptive Secondary Setup and Instrument Simulator) test bench and tested in
realistic conditions. GRAAL tests were
completed in early 2015 and GRAALs
PAE was granted in April 2015. Then the
system was prepared for shipment and
re-integrated in Paranal in June. This provided a unique opportunity in October
2015 to undertake a combined commissioning run with LGSU#1 and GRAAL,
allowing many aspects of the acquisition
sequence of the AOF on the telescope to
be debugged and tested under real
conditions. This evaluation was very
instructive and useful to the project and
again validated many design choices.
Then the GALACSI module was installed
on ASSIST and tests continued. By
autumn 2015, the tests in the MUSE widefield mode configuration were completed
and a start could be made with tests of
the MUSE narrow-field mode. These were
completed in February 2016 and the PAE
for GALACSI was granted in April.
In parallel, many tools required to optimise the operation of the AOF have
been developed, ranging from new versions of the GuideCam tool and the
observing tool to the delivery of the laser
traffic control software and the complete
refurbishment of the astronomical site
monitor. Most of these tools are now
operational. The commissioning period
of the AOF itself will ensure a final integration of these tools into the operational
scheme of the AOF. All these tests have
Figure 1. Characterisation of the beam profile

from the optical tube
assembly shown in the
Garching cold chamber,
after integration with the
beam control diagnostic
system and the laser.
1
2
3
AZ offset (arcsecond)
Figure 2. Accuracy of the pointing of the laser guide

stars (LGSU#1). The red circle is the requirement. The
final performance is better than the requirements by a
factor of two, ensuring a short acquisition time for
the AOF and a small overhead with respect to observation without laser guide star. Symbol colour indicates pointing angle from telescope optical axis.
demonstrated full compliance of the AOF

systems with their specification and the
desired performance has been reached
and clearly demonstrated.
Throughout the AOF review process,
and the system testing and installations
in Paranal, the project has placed strong
emphasis on training our Chilean colleagues. Many trips and exchanges took
place between Garching and Paranal.
This spirit of collaboration and enthusiasm will no doubt ensure that the astronomical community receives expert support for the AOF systems and effective
maintenance and optimisation of its performance will take place.
4LGSF highlights
The 4LGSF subsystem of the AOF has
demanded a substantial effort in terms of
manpower, finance and contract super
vision. The engineering systems expertise
within the AOF/4LGSF project has ensured
an efficient overview of the numerous
state-of-the-art components composing
this facility (Hackenberg et al., 2014).
There are five key specifications of the
lasers that represented real challenges
in the early years of the AOF project.
These were:
1. High power output: 22 watt;
LGS Guider FWHM ( )
ALT offset (arcsecond)
Figure 3. The laser

guide star spot size
versus the natural
guide star (NGS)
image full width half
maximum (FWHM)
measured by the
telescope guider.
The two values of
FWHM are plotted
separately.
1.60
r = 0
r = 7.1
r = 43.8
r = 246
1.40
1.20
FWHM x
FWHM y
1.00
0.80
0.50
0.75
1.00
NGS Guider FWHM ()
2. P
ointing accuracy of the laser spot:
<5arcseconds peak-to-valley;
3. L
aser spot size: < 1.35 arcseconds for
1-arcsecond seeing (30 from zenith);
4. R
eturn flux of 5 106 photons m 2 sec 1
at the UT4 Nasmyth focus (for average
sodium density of 4 1013 m 2);
5. Reliability and robustness.
The power figure was reached early in
the development phase and did not represent a major issue for the TOPTICA/MPB
consortium in charge of delivering the
five laser units (four plus one spare). The
robust and simple laser design ensured
the stable behaviour of the laser output
power during its development and all the
test phases in Garching, and today at the
telescope. The most impressive demonstration was the reception of the first laser
unit in Garching; after unloading from the
truck transport (TOPTICA headquarters
are 35 kilometres from Garching), rolling
off into the ESO laboratory, making the
connections and switching on, the power
meter unflinchingly showed a remarkable
and stable dead-on 22-watt output! This
speaks for the lasers robustness, which
was always considered an important feature for a system that was to be operated
in an observatory environment. Also the
numerous systems in operation on the
AOF require a long mean time between
failure in order to fulfil the requirement of
high night-time availability.
The pointing accuracy was also a difficult
requirement; one must remember that
the operational optical quality for all
4LGSF optics is in the diffraction regime.
This demands high optical quality from
the components, careful alignment, stable
1.20
and rigid systems to maintain alignment

and careful engineering to control the
pointing. The system met the specification in the laboratory in Garching with no
margin. At this stage the project decided
to implement a laser pointing camera
(Bonaccini Calia et al., 2014), which would
relieve this specification somewhat. The
laser pointing camera is a small unit
mounted on the telescope top ring that
identifies the laser star, determines its
position with respect to the telescope axis
and updates the laser pointing if it is too
far off. This system was delivered by the
Rome Observatory and its implementation
has been very successful (see Figure 2).
The laser spot size is also within specification and is evidence for the stable and
excellent image quality of both the laser
beam control diagnostic system and the
optical telescope assembly. Figure 1
shows the characterisation of the beam
profile in the Garching cold chamber after
integration with the beam control diagnostic system and the laser.
In early 2008, when the laser call for
tender was launched, the interaction
between the laser beam and the atomic
fine structure of the sodium atoms
was not so well understood, strange as
this may seem in hindsight. Coordinated
efforts between the ESO laser and atomic
physics community helped to improve
this situation (Holzlhner et al., 2008;
2010; 2012). These efforts have been
largely successful as the spectral format
prescribed has delivered the desired
outcome. It was decided to specify to the
laser supplier to inject 20 watts into the
main Na D2a line and 10 % of this power
5
7
9
Arsenault R. et al., Adaptive Optics Facility Status Report
11
12.5
12.1
0
10
8
2
4
Figure 4. Simulation of the expected return flux (for a

simulated laser flux of 16 watt) plotted over the sky.
The variation across the hemisphere is due to the
interaction between the beam polarisation and the
Earths magnetic field; the maximum occurs towards
the south magnetic pole. Measured values range
between about 10 and 20 10 6 photon m 2 s 1 (laser
is actually 22 watt not 16) and follow well the distribution simulated across the hemisphere.
in the Na D2b line. This should prevent the

sodium atoms from recombining in a
state where the main NaD2a laser
line can no longer excite the atoms. The
result is an improved return flux.
The first on-sky results of return flux are
very encouraging, but the final assessment will only be made after a complete
year of monitoring to cover the full range
of seasonal variations of the density of
the atmospheric sodium layer (Figure 4).
GRAAL-LGSU#1 combined tests
The GRAAL module (Arsenault et al.,
2014; Paufique et al., 2012) serves the
HAWK-I imager and provides a ground
layer correction over the 7.5 7.5 arcminute field of view of HAWK-I. It also
provides an option called maintenance
and commissioning mode, featuring onaxis natural guide star adaptive optics.
This mode allows the 1170 degrees of
freedom of the DSM to be fully exploited
and aims at testing and validating the
correction capability of the mirror.
The installation of GRAAL on UT4 took
place in June 2015 (Figure 6) and the
combined commissioning with LGSU#1
in October 2015. Obviously, no adaptive
optics verification could be done in Octo-
ber 2015 without the DSM. However,

many loops and offload schemes could
be tested and the LGSU#1 could be
acquired on one wavefront sensor and
the tip-tilt star selection tested. This
allowed many steps to be executed under
real conditions and offered reassurance
as to the estimated acquisition overhead,
which had already been tested on
ASSIST in Garching. With all the loops
closing (except the main adaptive optics
loop with the DSM), the overhead was
always found to be less than four minutes
(requirement five minutes). The jitter loop
on one laser was closed on this occasion: the jitter mirror in the LGSU#1 was
controlled by the GRAAL laser guide star
wavefront sensor to reduce the tip-tilt of
the laser spot caused by telescope shake.
The pupil alignment on the wavefront
sensor was, however, found to become
decentred by more than we had predicted. This misalignment arises from a
combination of factors including telescope misalignment, but also from the
co-rotator (inside GRAAL) and the
Nasmyth co-rotator. However, simulations
and tests on ASSIST (with GALACSI, see
Figure 7) convinced us that this could be
handled by the scheme of regular command matrix updates.
The GuideCam tool was used for tip-tilt
star selection, and the laser traffic c
ontrol
software proved very useful to predict
and prevent laser collisions with other
telescope beams. The most frequent collisions were experienced with the Visible
and Infrared Survey Telescope for Astronomy (VISTA), the VLT Survey Telescope
(VST) and the Auxiliary Telescopes (ATs).
GALACSI performance
As for the GRAAL module, the tests of
the GALACSI module (La Penna et al.,
2014) on ASSIST clearly demonstrated
that it fulfils the system specifications.
The performance requirements for the
GALACSI wide-field mode is to increase
the received energy per pixel by a factor
of two, over the 60 60 arcsecond
MUSE field of view at 750 nm wavelength
and for 1.1-arcsecond seeing. This is
defined for a given atmospheric refractive
index structure parameter (Cn2 ) distribution. This parameterisation was assumed
Figure 5. The four laser guide star units mounted on

the UT4 centrepiece with their respective laser and
instrument control cabinets.
early on in the GALACSI system simulations and the design concept was
expected to provide the desired correction. This can clearly be seen in Figure 8.
The ASSIST test bench comprises only
three discrete phase screens to simulate
the continuous distribution of turbulence
in the atmosphere. This number of phase
screens and their relative importance
are not fully representative of the specified Cn2 profile. However, when the
ASSIST distribution was simulated, we
could reproduce the measured values of
ensquared energy (dashed line on Figure
8) well. This convinced us that on sky
with the specified Cn2 profile, the
GALACSI module would provide the
required correction. With the specified
atmosphere, simulations predict a factor
of two in improvement in ensquared
energy (EE) with 55 % turbulence in the
first 500 metres above ground.
These tests were carried out relatively
quickly as they implemented a similar
algorithm to the GRAAL ground layer
adaptive optics correction (Arsenault et
al., 2013) and little surprise was expected.
However, the GALACSIMUSE narrowfield mode involved a new, complex algorithm for laser tomography. The complete
Figure 6. The GRAAL module being handled above

the Nasmyth A platform of UT4 (top left of picture)
before being sandwiched between HAWK-I and the
Nasmyth flange.
definition of the command matrix for this

algorithm was defined by the ESO Adaptive Optics Group and performed beyond
expectations. The main difficulty arose
with the measurement of the Strehl ratio
values in the visible. The specification
called for a Strehl ratio of 5 % (goal 10 %)
at 650 nm in a 5-arcsecond field of view
and for 0.6-arcsecond seeing. One should
not be misled by such low values of
the Strehl ratio! This is a true challenge
for adaptive optics due to the short wave-
length (visible, 650 nm); it corresponds

to a 77 % Strehl ratio at 2.2 m and starts
to approach planet-finder performance if
the goal value is considered (82 % Strehl
ratio at 2.2 m).
Figure 7. The GALACSI module mounted on the

ASSIST test bench. A special rig has been produced
to interface the IRLOS (InfraRed Low Order Sensor)
used for tip-tilt and focus sensing with the MUSE
narrow-field mode.
Despite the challenges, the image

improvement is indeed spectacular (see
Figure 9) and measurements of the Strehl
ratio confirm this impression. The tests
performed on ASSIST also revealed a
very robust algorithm. The correction also
worked well in worse seeing conditions
(using the 1.1-arcsecond seeing simulation) and was found to be stable. Besides,
the correction appeared to be robust
enough that the control matrix does not
have to be modified when the turbulence

profile evolves, which will certainly simplify operations.
EE gain in 200 mas @ 625 nm
5
ASSIST
Octopus
20
40
60
% of C n2 in first 200 metres
80
100
Figure 8. The evolution

of the correction on
ASSIST at 625 nm as a
function of the fraction
of turbulence in the
ground layer (first 200
metres). The improvement starts getting substantial when more than
60 % of the turbulence
lies below 200metres
altitude. Good agreement is seen between
the simulation (solid
line, Octopus) and the
ASSIST measurements.
Benefits of the system tests

The choices for the design and construction of the ASSIST test bench (Stuik et al.,
2012) were initially guided by the need
to provide Microgate and ADS (the two
companies in charge of delivering the
new secondary mirror unit to ESO) with a
tool allowing the final optical calibration
of the DSM (Arsenault et al., 2013; M
anetti
et al., 2014 and Briguglio et al., 2014).
At the end of the manufacturing process,
each DSM capacitive sensor has to be
optically calibrated, and the optical shell
flattened to within 7.5 nm root mean
square error. This requires an interferometric setup. Due to the convex shape
of the DSM, this set-up ended up in a
cumbersome 2 2 3 metre tower made
of a 1.7-metre diameter aspherical primary mirror and a much smaller aspherical secondary mirror (see Figure 10).
It was thus decided to add, at the input
of this tower, a simple source simulator
and turbulence generator, and to design
Arsenault R. et al., Adaptive Optics Facility Status Report
1.0
520 mas
0.9
FWHM gain > 10
Peak intensity 30
Strehl ratio > 10%
0.8
0.7
Intensity
0.6
0.5
0.4
0.3
0.2
16 mas
0.1
0.0
500
1000
Profile size (mas)
1500
Figure 9. Image improvement of GALACSI narrowfield mode. Upper left: Uncorrected image; lower
left: with correction. The full width half maximum
gain and peak intensity improvement is shown above.
a VLT focus simulator to be located at the

output of ASSIST. The complete ASSIST
would then not only be able to achieve
the last step of the DSM optical calibration but also to fully characterise GRAAL
and GALACSI before going to Paranal.
The nice optical design was provided by
Bernard Delabre, and NOVA (University of
Leiden) supplied the system and covered
most of the costs and effort.
More than four months were necessary
to complete the critical optical calibration
of the DSM. During the two years of tests
on ASSIST, several birth defects in the
subsystems have been identified and corrected. Four months were spent on fixing
two major failures of the new secondary
mirror unit, time much more fruitfully and
efficiently spent in the laboratory than on
the telescope!
The ASSIST setup with the VLT control
model allowed a configuration that was
very close to that of the final telescope to
be achieved. Therefore, substantial progress could be made on template coding
and testing and on the software infrastructure. The complete acquisition sequence
could be tested and optimised. Finally,
ASSIST has offered a comfortable work
environment with calibrated atmospheric
turbulence, a situation much more stable

and reproducible than on the telescope.
The AOF staff who have operated the
system for these past two years in the
laboratory have gained a superb knowledge and understanding of the system.
No doubt unknowns, new issues and
problems will surface once the system is
operating on UT4, but the experience
gained in Garching will prove invaluable in
solving these efficiently. A number of
Paranal staff have travelled to Garching
and have been using the system to prepare them for the future operation and
maintenance work at the observatory.
AOF operation tools and infrastructure
At the instigation of the AOF Scientist,
many tools were put in place to optimise
the AOFs operation. Many of them have
been mentioned already. The GuideCam
tool has been developed for GRAAL/
HAWK-I to allow tip-tilt natural guide star
selection. A similar version with some
additional functionalities is being developed for GALACSI/MUSE as well.
The laser traffic control system (Amico et
al., 2105) is a piece of software used to
predict optical collisions between the

powerful laser beams of the 4LGSF and
the lines of sight of other telescopes
on the mountain. It has been in operation
for two observing periods and we have
received very positive feedback from
Science Operations in Paranal. A web
interface displays the situation of all the
telescopes on Cerro Paranal and predicts any collision with the AOF lasers.
Priority rules are implemented and
recommendations are made to the telescope operator. Alarms are triggered
when a collision is imminent. The observing tool has also been updated by implementing a query system to the laser
traffic control system to check whether
an observing block risks a collision. This
will serve observers during the night
and help in the preparation and scheduling of observing blocks.
The AOF observing mode to select will
depend on atmospheric conditions. To
this end the astronomical site monitor
went through three major improvements:
1. New atmospheric tools have been
implemented; a multi-aperture scintillation sensor and a differential image
motion monitor have been installed on
a higher structure (7 metres above
ground) so that they are less sensitive
to the near-ground surface layer, which
can cause a systematically more pessi
mistic reading of the seeing than the
telescope guider. A new SLOpe Detection And Ranging instrument (SLODAR)
has also been Paranalised and integrated into the astronomical site monitor infrastructure.
2. T
he database to store the readings of
these new tools has been upgraded
to include additional parameters. This
database is also automatically replicated in Garching.
3. Finally, a new display tool has been
developed and implemented. The
standard setup has been chosen to
faithfully replicate the previous look
and feel, so that many users may not
yet have noticed the difference. But
this new display tool is very powerful
and allows any user, through a login,
to define his/her own preferred setting
and parameters.
The astronomical site monitor will provide
the turbulence distribution in the atmosphere, which will be a critical criterion for
new instrument at the Cassegrain focus

will follow (ERIS Enhanced Resolution
Imager and Spectrograph), which combines the use of the near-infrared integral
field unit SINFONI and a diffraction-limited camera sensitive up to 5 m. All
UT4 instruments can make use of natural
guide stars as well as up to four laser
guide stars for wavefront sensing.
For MUSEs wide-field mode, the adaptive optics module GALACSI will con
centrate the energy of the point spread
function across the field of view, enabling
more occasions with excellent seeing
conditions in r egular observing time to be
available. MUSEs narrow-field mode
opens up a new domain with Strehl conditions of > 5 % in the visible (650 nm),
which facilitates science cases for very
crowded field integral field spectroscopy,
such as for the centres of galaxies and
globular clusters.
Figure 10. The GRAAL module on ASSIST with the
DSM at the top during system tests in Garching.
observation mode scheduling (Kuntschner

et al., 2012). The simple rule is:
Good seeing 0.6 arcseconds and better
implies GALACSI laser tomography
adaptive optics and MUSE narrow-field
mode;
Worse seeing ~1 arcsecond and strong
ground layer, 70 % below 500 metres,
implies GALACSI ground larger adaptive optics and MUSE wide-field mode;
Even worse seeing ~1 arcsecond and
strong and very low ground layer (below
300 metres) implies GRAAL/HAWK-I;
In all other conditions, seeing-limited
operation (about 30 % of the time).
Science outlook
It is planned to first use the AOF in MUSE
wide-field mode for science from mid2017 onward. Other modes such as the
narrow-field mode of MUSE and HAWK-I
ground layer adaptive optics will follow
in 2018. With the adaptive secondary mirror, all foci at UT4 can be provided with
turbulence-corrected images, without the
addition of adaptive modules and sup
plementary optics in front of the instruments. The concept is more far-reaching
than only a deformable secondary mirror,
since the instrument park is optimised
to benefit from this upgrade. In 2020 a
HAWK-I with GRAAL is expected to provide about a factor of two improvement

in the occurrence of good J-, H- and
K-band images (< 0.4 arcseconds).
Improvement is also expected for all seeing conditions, promising almost spacebased observatory quality images in the
near-infrared.
Next phases and milestones
Two main systems remain to be installed
on the telescope. In October 2016 there
will be a UT4 shutdown until December
to install the new secondary mirror unit
with the deformable secondary mirror.
Then, the telescope will be re-validated
in non-adaptive optics mode. In early
January 2017, the GRAAL maintenance
and commissioning mode will be used to
validate the adaptive optics correction
capability of the DSM, while GALACSI will
be re-integrated and installed on UT4.
This operation should be completed by
March 2017.
The AOF team will then conduct regular
commissioning runs throughout 2017
with the goal of delivering the GALACSI
MUSE wide-field mode for Observing
Period 100. After this the project will focus
on commissioning the GRAAL ground
layer adaptive optics and GALACSI laser
tomography adaptive optics modes.
Conclusions
The AOF has completed its activities in
Garching and now the focus is shifting
ever more to Paranal, completing the
stand-alone 4LGSF commissioning and
then moving on to the installation of
the deformable secondary mirror on UT4
in the fourth quarter of this year. With
this facility in place commissioning of
GALACSI in wide-field mode can start,
yielding the first science by mid-2017.
HAWK-I/GRAAL narrow-field mode and
MUSEs ground layer adaptive optics will
follow in 2018. The suite of instruments
is accompanied by a full set of operation
tools and an upgrade of the astronomical
site monitor, which will all support efficient operation of the AOF. The current
progress and laboratory measurements
within specification of the adaptive optics
models for HAWK-I and MUSE establish
high expectations for the onset of science
observations from mid-2017.
Acknowledgements
The AOF project started at ESO in 2006; we thus
celebrate its tenth anniversary this year! We are very
grateful to all the team members who have been
dedicated to this project since its beginning and for
their hard work even in the difficult times of technical
failures and challenges. Also, a great many people
have contributed to the AOF for temporary support
activities, or a review board panel; they are also
warmly thanked for their contributions. We also very
much appreciate all the functional managers and
group leaders who have been helping to solve
resource issues and conflicts during this long effort.
The AOF has also greatly benefited from the expert
support of its industrial partners, and we wish to
thank here especially Microgate, ADS Intl., TOPTICA,
MPB and SAFRAN Reosc.
References
Amico, P. et al. 2015, The Messenger, 162, 19
Arsenault, R. et al. 2010, The Messenger, 142, 12
Arsenault, R. et al. 2013, in AO4ELT Conference

Adaptive Optics for Extremely Large Telescopes,
Florence, Italy, May 2013
Arsenault, R. et al. 2014, Proc. SPIE, 9148
Bonaccini Calia, D. et al. 2014, Proc. SPIE, 9148
Briguglio, R. et al. 2014, Proc. SPIE, 9148
Hackenberg, W. et al. 2014, Proc. SPIE, 9148
Holzlhner, R. et al. 2008, Proc. SPIE, 7015
Holzlhner, R. et al. 2010, A&A, 510, A20
Holzlhner, R. et al. 2012, Proc. SPIE, 8447
Kuntschner, H. et al. 2012, Proc. SPIE, 8448
La Penna, P. et al. 2014, Proc. SPIE, 9148
Manetti, M. et al. 2014, Proc. SPIE, 9148
Paufique, J. et al. 2012, Proc. SPIE, 8447
Stuik, R. et al. 2012, Proc. SPIE, 8447
A Fruitful Collaboration between ESO and the Max Planck

Computing and Data Facility
Nathalie Fourniol 1
Stefano Zampieri 1
Manuel Panea 2
1
2
ESO
Max Planck Computing and Data
Facility, Garching, Germany
The ESO Science Archive Facility (SAF),

contains all La Silla Paranal Observatory raw data, as well as, more recently
introduced, processed data created at
ESO with state-of-the-art pipelines or
returned by the astronomical community. The SAF has been established for
over 20 years and its current holding
exceeds 700 terabytes. An overview of
the content of the SAF and the preservation of its content is provided. The latest development to ensure the preservation of the SAF data, provision of an
independent backup copy of the whole
SAF at the Max Planck Computing and
Data Facility in Garching, is described.
The ESO Science Archive Facility
The SAF is the unique access point to
ESO data via services that are increasingly
used by the community. The ESO SAF
currently contains about 700 TB of ESO
data in the form of more than 40 million
distinct files. About twice this amount
of data is actually stored, as a second
copy of all archived files is present in the
system. The data from over 20 individual
instruments, spread among the three
ESO sites namely Paranal, La Silla,
and APEX are transferred continuously
via the network from Chile to Garching.
Ninety per cent of the files reach Garching
in less than 15 minutes. Once saved in
the SAF, these data are then published,
in the sense that authorised users can
download them. Projections show that
with the arrival of the next few instruments, which will generate large amounts
of data, the SAF will reach its first petabyte in 2018.
The SAF is the one and only access point
to all La Silla Paranal Observatory raw
data and to processed data thereof.
Figure1 shows the current web interface
to the SAF and its collections1. As well
as raw data from the observatory, data
products are generated either by the

community (Arnaboldi et al., 2014) or
produced in-house (Romaniello et al.,
2016a). Recent analysis on downloads
of these data products as well as the
raw data have shown that the SAF is a
very popular science resource in itself,
aggregating a new astronomical community and attracting new registrations to
the user portal at a rate of a couple per
day (Romaniello et al., 2016b).
In order for this mine of scientific data to
be available to users, the archived data
need to be preserved. The SAF content
resides in a data centre at ESO where
machines run continuously (Figure 2). The
ESO archive system is based on software
developed at ESO called NGAS (Next
Generation Archive System; Wicenec et
al., 2007), also used by the ALMA Science Archive2. The system uses the
RAID5 technology, which ensures data
redundancy. The second copy in the system, produced for any newly archived
data within the hour, provides additional
redundancy.
But this of course is not enough. To
ensure data preservation and availability,
one must also plan against major disaster
or long-lasting disruption, while hoping
to never experience such an event. That
is where the collaboration between ESO
Figure 1. The ESO Science Archive Facility homepage.
and the Max Planck Computing and Data

Facility comes in, where a backup copy
of all ESO data now resides.
The Max Planck Computing and Data
Facility
The Max Planck Computing and Data
Facility (MPCDF, formerly known as
Rechenzentrum Garching [RZG]) is the
main computing centre of the Max Planck
Society. The MPCDF provides computing
and data management services for all
Max Planck Institutes across Germany
in many fields of science, including materials and life sciences, theoretical chemistry, polymer research, astrophysics,
plasma physics and others.
Apart from several mid-range Linux computer clusters, the MPCDF operates a
supercomputer, currently an IBM iDataPlex system with Intel Ivy Bridge pro
cessors and Nvidia K20X GPUs. In total
there are about 83 000 cores with a main
memory of 280 terabytes and a peak
performance of about 2.8 petaflop s 1,
making it one of the fastest computers in
Germany. In close collaboration with
scientists, the MPCDF also develops and
optimises algorithms and programs for

high performance computing.
Figure 2. An image of
the ESO Data Centre
where the SAF content
resides.
The MPCDF data science team supports

the development and deployment of
data management solutions for big data
projects. Large amounts of experimental
data and data from supercomputer sim
ulations are managed and stored in three
tape libraries in two different locations,
with a total capacity of about 40 000
tapes. One of the libraries is shown in Figure 3. The combined amount of backup
and archive data stored at the MPCDF
currently totals over 50 petabytes, with a
current growth rate of almost 1.5 petabytes per month.
An important requirement of some of
the stored data is its long-term availability. This is especially true for the humanities because of the irreproducibility of
the data, for example, scans of valuable
documents from the Max Planck Institute
for Art History or recordings of endangered human languages from the Max
Planck Institute for Psycholinguistics.
Through redundancy at the logical and
at the physical level, the storage of data
for a period of time that surpasses the
typical life-cycle of the hardware and
software components employed can be
guaranteed, with the goal of keeping the
data forever. Some of the data still stored
at MPCDF originated in the early 1980s,
almost 40 years ago.
Implementation of a collaboration
The possibility of storing a backup copy
of the ESO archive at the MPCDF was
first discussed between ESO and the
MPCDF in November 2012. The agreement signed thereafter by both parties
stipulated that the MPCDF offered to
save a backup copy in their tape library
of all ESOs archived data. ESO developed in-house an automatic procedure
for transferring the data to the MPCDF.
The replication of archived data then
started on 7 May 2015 and, as planned,
it took eight months to handle the backlog (at that point in time 690 terabytes
of data were replicated). As of 20 January
2016, the backlog has been cleared, and
the replication process is now operational, meaning that all newly archived
data are backed up in near real time at
the MPCDF. This process has been working non-stop and flawlessly since then.
Figure 3. One of the tape libraries at Max Planck

Computing and Data Facility (MPCDF) on the
G arching Research Campus.
Acknowledgements
References
Our first thank-you goes to Felix Stoehr, who first

established a contact between ESO and MPCDF
colleagues, which then triggered the discussions
and the eventual collaboration. We also want to
thank Nicolas Rosse, Laura Mascetti and Michael
Boelter for their dedication in helping to test the
p rocess and then running it in operation. Maurice
klein Gebbinck is also to be thanked for testing the
replication procedure.
Arnaboldi, M. et al. 2014, The Messenger, 156, 24

Romaniello, M. et al. 2016a, in prep.
Romaniello, M. et al. 2016b, The Messenger, 163, 5
Wicenec, A. et al. 2007, The Messenger, 129, 27
Links
1
2
Finally, this article is an opportunity for ESO to say

a big thank you to the MPCDF for this very fruitful
and successful collaboration, which is also, and not
least, of long-term benefit to all of the SAF users.
ESO SAF: http://archive.eso.org/

A LMA Science Archive: http://almascience.eso.org/
alma-data/archive
Solar Activity-driven Variability of Instrumental

Data Quality
Christophe Martayan1
Alain Smette 1
Reinhard Hanuschik 1
Pierre van Der Heyden 1
Steffen Mieske 1
1
Sun lead to a number of consequences

on Earth as listed in Table 1.
ESO
The unexplained variability of the data

quality from Very Large Telescope
instruments and the frequency of power
cuts have been investigated. Origins
for the variability in ambient temperature variations, software, data reduction
pipelines and internal to hardware
could be discarded. The most probable
cause appears to be correlated with
the e
volution of the cosmic ray rate,
and also with solar and terrestrial geomagnetic activity. We report on the
consequences of such variability and
describe how the observatory infrastructure, instruments and data are
affected.
Context
With the improvement over time of detector stability and the increase in frequency
and level of instrument health monitoring,
any deviations in quality control parameters due to larger than normal changes
in ambient conditions (for example, temperature and air pressure) are now easier
to detect. Sometimes, unusual glitches
are registered. Once hardware and software problems have been discarded
as the probable cause, a few events still
remain unexplained. Such problems can
have real consequences for data quality.
The primary cause for such variability lies
in the SunEarth relationship. The solar
maximum occurred around 20122014,
although this cycle (24) is characterised
by a low level of activity. Solar activity and
geomagnetic changes induced by the
There are five classes of geomagnetic

storm, from G1 (minor, ~ 1500 per solar
cycle of 11 years) to G5 (extreme, 0 to
4per solar cycle), as classified by the US
National Oceanic and Atmospheric Service (NOAA1). The consequences of
geomagnetic storms can range from the
appearance of aurorae and disturbances
to migration patterns, to complete blackout of the electrical systems on Earth
and failure of satellite electronics. The
largest recorded geomagnetic superstorm (combination of consecutive severe
to extreme storms) that hit the Earth
occurred in 1859 (the Carrington event2)
with aurorae down to sub-equatorial
regions, although they are usually limited
to the polar circles. The terrestrial apparent magnetic field inverted and some
induced currents caused the telegraph to
fail and to deliver electric shocks to operators. More recently, in 1989, a power
blackout occurred in Canada and northern
USA due to strong induced currents generated by another storm, but three times
weaker than the 1859 one.
Northern Chile lies in the South Atlantic
Geomagnetic Anomaly, where the magnetic field is much lower than everywhere
else on Earth3. As a consequence, the
geomagnetosphere is less effective in
protecting this region from the effects of
cosmic rays and particles coming from
the Sun. Most satellites crossing this
region are set in safe mode to protect
them from the higher cosmic ray and particle flux. In addition, it is known that at
a defined latitude, higher altitude regions
receive more cosmic rays. The increase
is also larger for latitudes ranging from
the equator to 30, including all ESO
observatories; the NOAA World Magnetic
Model gives details4. Therefore, it is per
tinent to ask whether some or all of the
types of activity listed in Table 1 have an
impact on the measured instrument vari-
ability and data quality, as well as in the

dependability of the observatory infrastructure.
There are very few cosmic ray/neutron
monitoring stations in the southern
hemisphere. They represent less than
20 % of the worldwide network covered
by the Neutron Monitor DataBase5. In
addition, there are no neutron monitoring
stations in South America and Africa
and others are located in regions with a
stronger magnetic field. As a consequence we have compared the measurements from the monitoring stations in
the Kerguelen Islands (southern Indian
Ocean, latitude 49) and Terre Adlie
(continental Antarctica) with our data.
Cosmic ray monitoring during the minor
geomagnetic storm 45 September 2015
The monitoring of the health of the instruments at Paranal Observatory is carried
out daily during morning calibrations.
Various parameters6 are obtained concerning the detectors, such as the bias
and dark levels, readout noise, etc. However, X-shooter (Vernet et al., 2011) is
the only instrument to specifically monitor
the number of transient events, from the
detector dark exposures with the nearinfrared arm. The level of the dark current
is partially caused by radioactive decay
within the instrument itself. However, the
main cause of variability for the darks is
caused by the impact of particles originating outside the observatory, such as
real cosmic rays. Unfortunately, as for
other instruments, the monitoring is only
carried out once a day, leading to a low
temporal sampling.
However, during the night 45 September 2015, regular sequences of three
5-minute darks were specifically executed
to determine whether the dark cosmic
ray rate and readout noise (RON) levels
were affected together. Figure 1 shows
Table 1. Solar and geomagnetic activity and their consequences.

Type of solar/geomagnetic activity
Consequences on Earth
Consequences on data and instrument
Solar flare radiation/proton storm (RS)

Coronal mass ejection (CME)
Fast dense coronal solar wind (SW)
Magnetic fluctuations (MF)
Co-rotating interaction region (CIRC)
Aurora, airglow
Radio burst and blackout
Geomagnetic storms
Variation of Earths apparent magnetic field including its axis
Induced Foucault currents, Forbush effect
Increase/decrease in flux of cosmic rays
10
Instrumental parameter level variability

Electronics failure
Power loss
the comparison of the time evolution

of the relative cosmic ray rate measured
by X-shooter (blue squares) vs. the CR
rates from the Kerguelen Islands and
Terre-Adlie stations (red and black curves
respectively, hereafter KERG and TERA).
At the beginning of this minor geomagnetic storm (G1 level), all the curves
show a dip. This paradoxical behaviour
is called the Forbush effect (Forbush,
1937; 1938). The solar magnetic loops,
often extended by the solar matter beyond
the Earth, act like a shield, deviating the
cosmic rays. As these are mostly of extra
Solar System origin, this effect causes
a decrease in the CR rate. However, in
this event, a fast, dense solar wind
brought protons (green curve in Figure 1)
that reached the Earth, generating the
geomagnetic storm.
A linear regression between the X-shooter
CR rate and those of KERG and TERA
indicates a strong correlation, with a
Pearson coefficient larger than 45.

7RGNNSDQ
*DQFTDKDM
3DQQD CKHD
/QNSNMkTW

,)#
The follow-up of this event with X-shooter

illustrated its capability to serve as a CR
rate-monitoring instrument. This monitoring can be carried out at almost no cost
during the day and at night, when the
instrument is not being used for calibrations or scientific observations, without
disturbing operations.
Impact of solar and geomagnetic activity
on X-shooter dark exposures
One can also ask whether the instruments and their subsystems can be
26&
,%
affected by other geomagnetic and/or

solar events besides cosmic rays. Before
investigating this question a few considerations are necessary. Firstly, it should
be noted that the dark current rate of
the X-shooter near-infrared arm shows a
reproducible and slow increase over
several months following each thermal
cycle that is not understood. The same
behaviour is also noticeable in other nearinfrared instruments like the High Acuity
Wide field K-band Imager (HAWK-I).
Therefore, in order to better assess the
impact of the local variations related to
astronomical events in the X-shooter dark
26&
,%&

,%
26
12"1(1 1222!"

l
l
12",$26
2622!"&
l
",$R26
&26",$
l
l

12

-NQL@KHRDCCDUH@SHNMSNSGDLDCH@MC@QJ
l2DOSDLADQ
,HMNQFDNL@FMDSHBRSNQL&
LOKHjDCMNQL@KHRDC"1
Figure 1. Relative evolution of the cosmic ray rate

during the minor geomagnetic storm of 45 September 2015. The x-axis gives the Modified Julian Day
(time) and the y-axis corresponds to the relative CR
rates. The X-shooter (blue squares), Kerguelen
Islands (red), Terre-Adlie (black) CR rates all show
first a decrease then a return to the normal level once
the solar wind intensity is back to normal (Forbush
effect). The density of protons carried out by the fast
dense solar wind, which causes the storm, is shown
in green.
m
",$22!"

,)#

m

m

Figure 2. X-shooter
normalised near-infrared
detector dark level after
removal of the long-term
trend (blue points and
line). The 2, 3 and 4
standard deviations are
shown with horizontal
lines. The events at
more than 3 are annotated with a possible
explanation: CME:
c oronal mass ejection;
CRIR: co-rotating interaction region; Gx: geomagnetic storm of class
x (1 to 5); MF: magnetic
fluctuation; RS: radiation
and proton storm; SSBC:
solar sector boundary
crossing; SW: fast dense
solar coronal wind.
11
Figure 3. Temporal evolution of a few instrumental

parameters (X-shooter VIS readout noise, KMOS
median dark, GIRAFFE readout noise, HAWK-I
median dark and readout noise) which correlate,
over the period 30 October to 12 November 2015.
level, this slow long-term trend was

removed. Secondly, the median and
standard deviation () of the dark rate are
computed and frames showing absolute
deviations larger than 3 are visually
examined. The dark level and its glitches
over the period 19 November 2014 to
29 November 2015 are shown in Figure2.
At the epochs of the noted events indicated on Figure 2, all instrumental and
pipeline problems, as well as variations in
ambient temperature, can be discarded
as the probable cause. Manual inspection of the individual frames showed confirmatory evidence of frames deviating
from the normal.
We found that 94 % of the deviating
X-shooter dark levels can be matched to
specific geomagnetic or solar activity,
as recorded7. Only one event in this oneyear period remains unexplained. The
probability of a random temporal coincidence between dark glitches and solar
events would be 1 % (for example, nonrandom correlation of more than 3).
Sometimes, aurorae and geomagnetic
storms can happen even without obvious
registered phenomena. Often, events like
a CME apparently decrease the dark level
(even if it is difficult to explain). Others,
such as the magnetic fluctuation of the
geomagnetosphere (crack), or recoupling
with the solar magnetosphere, usually
lead to an increase in the dark level. In
some cases, it is difficult to conclude
whether there is a positive or negative
effect of the solar event on the data quality, as the dark level can increase or
decrease. This is especially true if there is
a combination of events including CME,
SW, etc.
However, as already mentioned, the
temporal sampling of X-shooter nearinfrared darks is usually too low, meaning
that the calibrations must be taken at the
time of the event in order to record it. The
duration of the events range from a few
minutes to several days. In most cases,
because no calibrations were taken
during the event, the sampled dark curve
remains unaffected, while it would have
12
Martayan C. et al., Solar Activity-driven Variability of Instrumental Data Quality
*,.2C@QJ
7RGNNSDQ5(21.-
&(1 %%$1.-
' 6*(C@QJ
' 6*(1.-

12
&RSNQL
&RSNQL
TMRDSSKDC

LOKHSTCDQDK@SHUDDUNKTSHNM

,)#
probably shown a glitch at the time of the

event occurrence. Despite this limitation,
one can recognise an event representing
the Forbush effect, as shown in Figure 1,
visible in Figure 2 at MJD ~57270. During
this event, there is a moderate to strong
(2.53) correlation of the near-infrared
dark level, and also of the X-shooter
ultra-violetblue (UVB) and visible (VIS)
detector bias levels and the UVB and VIS
readout noise, with the Kerguelen/Terre
Adlie and X-shooter near-infrared CR
rates.
Impact of solar and geomagnetic activity
on the Paranal instruments and data
quality
Since X-shooter darks show a response
to solar and geomagnetic activity, it is
justifiable to ask whether other Paranal
instruments might react as well. Two
strong limitations are apparent: the first
is the infrequent temporal sampling,
the second the lack of simultaneity of the
calibration for a particular instrument
with the reference data of X-shooter dark/
CR and the Kerguelen/Terre Adlie CR
record. Despite these difficulties, we
chose the time interval from 30 October

to 12 November 2015. During this period,

several CMEs and geomagnetic storms
G1 to G2 occurred, but there was also a
quiet time.
Over this period, a search for a correlation between various detector parameters
and the X-shooter or Kerguelen/Terre
Adlie CR rates was carried out. On average, three parameters (among them
bad pixel number, pattern noise, bias
or dark level, readout noise) per instrument were examined. The instruments
that were tested are: FORS2, KMOS,
GIRAFFE, X-shooter (UVB/VIS arms),
VISIR, SINFONI, HAWK-I, VIRCAM and
OMEGACAM. They were chosen because
at least one is located at each of the VLT
Unit Telescopes and the VLT survey telescopes. As an example, a comparison
of the temporal relative evolution of a few
parameters from four instruments showing a similar trend is displayed in Figure 3.
Linear regressions between the chosen
instrumental parameters and the CR
values were carried out. The regression
and Pearson coefficients were computed.
To define whether the instrumental quantity correlates with the CR, values of the
Pearson coefficient of more than 2 and
Table 2. Correlation of instrument parameter values with evolution of the X-shooter or KERG/TERA CR rates.
Type of CR monitoring
Fraction of instrument parameters

with correlation > 2 (%)
Fraction of instruments with correlation at:

2
X-shooter
54
6/9
2/9
KERG/TERA
60
6/9
3/9
3 were considered. During the events,

instrumental parameter variations of less
than 1 % (no change) to 155 % were found.
The main results are given in Table 2.
Slightly more than half of the tested
parameters per instrument apparently
vary with the CR evolution as indicated
by the percentages in Table 2 column 2.
Not only is the number of bad pixels
affected, but also the dark/bias level
and/or the readout noise. In Table 2
columns 3 and 4, the correlation of the
parameter variation with the CR evolution
is given. It is better than 2 in two thirds
of the instruments and better than 3
in the remaining one third of the cases.
These results indicate that all the tested
instruments seem to be impacted in
some way by geomagnetic and solar
activity. It is also worth mentioning that
some simultaneous disturbance in
several instruments located on different
telescopes occurred, pointing to a common origin. Other more sophisticated
statistical methods seem to confirm the
correlations.
The temporal coverage and the samples
used are still quite small. Ideally, obtaining better sampling would improve the
significance, but this is difficult to reconcile with normal Paranal operations. However, geomagnetic activity cannot be
blamed for all variations or glitches that
occur. In all cases, severely affected calibrations are stored in the archive (but are
not generally available) and the calibration
data are retaken. Such events can occur
during both daytime and nighttime but
monitoring activity is generally confined
to the daytime, so occurrences at night
may be missed. However, one might
take advantage of the Forbush effect
(decrease in cosmic rays) to observe faint
objects and avoid the magnetic fluctuation events (increase in readout noise) as
often as possible.
problems and some protective systems

will mistakenly trip out key assets from
the grid, inducing pipeline currents, etc.,
according to NOAA1. In particular, the
solstice G4+G4 superstorm (2223 June
2015) and the Halloween G5+G5 superstorm (2930 October 2003) generated
power cuts in Sweden, possibly also
in Argentina, induced currents in various
countries, aurorae in tropical areas,
magnetic declination fluctuations reaching 20degrees, etc.
At Paranal, power cuts occurred in
some areas at the times of these storms,
damaging hardware, including on the
telescope platform and systems under
stabilised current control. No other specific reason could be found for these
power cuts. Generally, the main electrical
systems destroyed during these powerful
geomagnetic storms are the transformers, breakers, etc. This is exactly what
happened at Paranal. A correlation
between power cuts and superstorms
corresponds to a random probability of
~0.0005 %, which is highly unlikely. In
addition to monitoring solar and geo
magnetic activity, preventive and corrective actions to be taken when a new
superstorm occurs have been defined.
Prospects
Other impacts on the observatory infrastructure
Paranal instruments, being highly sensitive, are affected by external disturbances,

such as solar and geomagnetic events.
In some cases, the data quality benefits
from them, but not always and it could
be strongly and negatively impacted.
When several solar events occur simultaneously, their consequences on the dayto-day functioning of the observatory, the
instrument health and data quality are
difficult to predict. X-shooter appears to
be able to follow, with adequate temporal
sampling, the evolution of the cosmic
ray rate and some of the solar and geomagnetic events. Such monitoring could
be harnessed to provide added value to
the community.
Beyond the effect on the data quality

and health of Paranal instruments, solar
and geomagnetic activity can also disturb operations in another way. Severe/
extreme storms are known to generate
possible widespread voltage control
The Sun should now evolve towards the

quiet phase of its unusual cycle number
24 and that will help in improving the
stability of the instruments and their data.
Fortunately, because the South Atlantic
Magnetic Anomaly is drifting eastwards
with time, in almost a century Chile

should be free from its effects.
Acknowledgements
Kerguelen Islands and Terre Adlie neutron monitor
data were kindly provided by the French Polar Institute (IPEV, Brest) and by the Paris Observatory.
The results presented in this paper rely on data
c ollected at the Argentine Islands (Akademik
Vernadsky base), Antarctic; Isla de Pascua Mataveri
(Easter Island), Chile; Trelew, Argentina; and Portaux-Francais, Kerguelen Islands, France. We thank
the National Antarctic Scientific Center of Ukraine
(NASC), the Direccin Meteorolgica de Chile
(DMC), the Institut de Physique du Globe de Paris
(IPGP), Universidad Nacional de La Plata (UNLP),
Royal Meteorological Institute of Belgium (RMIB)
and the cole et Observatoire des Sciences de la
Terre (EOST), for supporting their operation and
INTERMAGNET8 for promoting high standards of
magnetic observatory practice.
References
Forbush, S. E. 1937, PhRv, 51, 1108
Forbush, S. E. 1938, PhRv, 54, 975
Vernet, J. et al. 2011, A&A, 536A, 105
Links
1
OAA Space Weather Scales:

N
http://www.swpc.noaa.gov/sites/default/files/
images/NOAAscales.pdf
2
Super solar flares and the Carrington event:
http://science.nasa.gov/science-news/science-atnasa/2008/06may_carringtonflare/
3
Chulliat, A. S. et al., 2014. doi: 10.7289/V5TH8JNW
4
World Magnetic Model:
https://www.ngdc.noaa.gov/geomag/WMM/data/
WMM2015/WMM2015_F_MERC.pdf
5
N eutron Monitor DataBase: http://www.nmdb.eu/
6
ESO Quality control and data processing group:
http://www.eso.org/observing/dfo/quality/
7
Log of space weather:
http://www.spaceweather.com/
8
INTERMAGNET: http://www.intermagnet.org
13
Science Verification for the VISIR Upgrade
ESO
INAF Osservatorio Astronomico di
Padova, Italy
3
L aboratoire AIM, CEA/DSM CNRS
Universit Paris Diderot, IRFU/SAp,
Gif sur Yvette, France
2
The Very Large Telescope spectrometer

and imager for the mid-infrared (VISIR)
was upgraded in 2015 with new detectors and several new modes were
added. Science Verification (SV) is carried out for new ESO instruments as well
as for substantial upgrades to existing
instruments. Sparse aperture masking
and coronagraphy in the mid infrared
have now been added to VISIRs capabilities and during SV these new observational modes, together with the
recommissioned burst mode, were used
to demonstrate the observational capabilities of the instrument. The SV process
for VISIR is briefly described and some
results from the successful observations
are presented. All SV data are publicly
available.
The VISIR upgrade project (Kufl et al.,
2015) replaced the N-band low-resolution
grating by a prism, the old detectors by
new AQUARIUS (1024 1024 pixel)
detectors, and added to the high spatial
resolution capabilities by deploying coronagraphic and sparse aperture masking
modes. In addition, support is now provided for monitoring of the precipitable
water vapour integrated column through
the atmosphere to enable the best midinfrared observing conditions to be
selected. The VISIR burst mode, which
enables very high time resolution sampling, was recommissioned shortly before
the Science Verification run.
Observations
The call for VISIR upgrade Science

Verification proposals was issued1 on
15 November 2015 and advertised
through the ESO Science Newsletter 2
on 18 November. Thirty-five proposals
had been received by the deadline on
15December 2015. The VISIR upgrade
SV team evaluated all the proposals.
Two SV proposals that requested the
same targets and observing modes as
Period97 programmes were rejected.
A few other programmes requesting
instrument modes that were not part of
the upgrade were also rejected. The
cut-off line was defined at 44 hours of
allocated time, which meant that 22 programmes could receive an allocation.
All PIs were informed of the outcome of
the selection on 27 January 2016 and
the Phase 2 material was submitted by
all successful PIs by 12 February 2016.
The VISIR upgrade Science Verification

took place on VLT Unit Telescope 3 in
February and March 2016. Originally,
the nights from 23 to 27 February 2016
were scheduled, but as these SV nights
were very strongly affected by adverse
atmospheric conditions, three half nights
from 19 to 21 March were added. The 22
observing programmes selected required
a total of 43.6 hours of telescope time.
Of these programmes, 12 were fully completed and another six received partial
data. Only four programmes could not be
observed at all. The instrument worked
without major fault and no substantial
loss for technical reasons was recorded.
A wide range of science topics were allocated time. They included the imaging
of Solar System objects (including the
companions of asteroids), exoplanetary
systems, protoplanetary and debris discs
around young stars, the environments of
evolved stars and nearby active galactic
nuclei (AGN).
Precipitable water vapour (mm H 2O)
Proposal solicitation and submission
Figure 1. The precipitable water vapour record from

the LHATPRO radiometer over the two runs of VISIR
SV in February and March 2016.
17.5
2327 Feb 2016
15.0
12.5
10.0
7.5
5.0
SV run 1
2.5
0.0
00:00
00:00
00:00
00:00
00:00
17.5
00:00
00:00
1921 March 2016
15.0
12.5
SV run 2
10.0
7.5
5.0
2.5
0.0
14
The conditions during the first run of SV

nights were extremely poor. Out of the
four nights 1.5 nights were lost to bad
weather with the telescopes closed. The
remaining time suffered from high humidity in the atmosphere above Paranal.
During the 2.7 nights when the telescope
was open, the precipitable water vapour,
as measured by the dedicated LHATPRO
20.0
Precipitable water vapour (mm H 2O)
Daniel Asmus 1
Mario van den Ancker 1
Valentin Ivanov 1
Hans-Ulli Kufl 1
Florian Kerber 1
Bruno Leibundgut 1
Andrea Mehner 1
Yazan Momany 2
Eric Pantin 3
Konrad R. W. Tristram 1
00:00
12:00
00:00
12:00
00:00
12:00
00:00
12:00
00:00
12:00
radiometer installed on Paranal as part

of the VISIR upgrade project (Kerber
et al., 2015) never dropped below 4 mm,
with an average around 10 mm. Figure1
(upper panel) shows the log of the precipitable water vapour measurements
spanning the time of the SV observations.
For comparison, median precipitable
water vapour conditions on Paranal are
around 2.4 mm (Kerber et al., 2014).
Technically, most of the SV run was
unsuitable for mid-infrared observations.
Figure 2. The southern

hemisphere of Jupiter
imaged with VISIR in the
M-band utilising burst
mode.
Given this unsuccessful run, ESO allocated another three half nights between
19 and 21 March 2016 to recover some
of the SV observations. This second
run was much more successful (see Figure1, lower panel) and yielded data for
a majority of the SV programmes. These
data are of excellent quality and in almost
all cases we were able to obtain unique
high-contrast observations. All new VISIR
modes were used during SV and worked
as expected. A few minor technical problems were uncovered, which were remedied for regular operations, starting with
Period 97 (1 April30 September 2016).
scope, allowing us to probe deeper into

Jupiters atmosphere than was possible
with any other instrument. The VISIR
images will also be used by the proposing team to select regions of interest
for closer scrutiny by the Juno mission
once it arrives at Jupiter in late 2016.
The sparse aperture masking mode
was used to image the massive young
stellar object NGC 3603 IRS9A, com
plementing VLT NAOSCONICA sparse
aperture masking data in the Ks- and
L-bands (Sanchez-Bermudez et al., 2016).
VISIR observations of Jupiter were

obtained in burst mode. Figure 2 shows
the southern hemisphere of Jupiter in
the M-band (4.8 m). This is the first
M-band image of Jupiters springtime
hemisphere from an 8-metre-class tele-
6FLHQFHWDUJHW

QBRDBNMC
&DOLEUDWRU

QBRDBNMC
2O@SH@KEQDPTDMBX@QBRDBNMC
(MSDQEDQNFQ@L
A few first results
QBRDBNMC
The VISIR upgrade SV web page3 contains direct links to the raw data. A new
version of the VISIR data reduction pipeline was released and now supports
basic reduction for all the new modes
tested during SV, except for sparse
aperture masking. The pipeline can be
accessed through the SV upgrade web
page3 or via the VLT pipelines page4.
QBRDBNMC
All raw data have been archived and

are publicly available. The automatic data
processing and transfer worked well,
considering the high data rates produced
by the new modes of VISIR, especially
the burst mode, which yields up to
200GB/hour worth of data.
Joel Sanchez (MPIA)
Archive and data processing
Figure3 shows the interferograms at

10.5m of NGC3603 IRS9A (left upper)
and the calibrator star HD96918 (left
lower). The fuzzier appearance of
NGC3603 IRS9A with respect to the calibrator star demonstrates that the science
target is clearly resolved at the baselines
corresponding to the sub-apertures of
the sparse aperture mask. The right panels show the spatial Fourier transform of
the interferogram which can, in a second
processing step, be converted to visibility. In comparison to the calibrator, the
disappearance of visibility at the highest
(MSDQEDQNFQ@L%3


Figure 3. Interferograms
at 10.5 m from VISIR
sparse aperture masking of NGC 3603 IRS9A
(left upper) and the calibrator star HD 96918
(left lower). The righthand panels show the
Fourier transform (FT) of
the interferograms of
both target (upper) and
calibrator (lower), where
the asymmetry of the
envelope of NGC 3603
IRS9A is apparent.
15
Asmus D. et al., Science Verification for the VISIR Upgrade
Figure 4. The mid-infrared extended dust emission

around the WolfRayet star WR104 is revealed by
the VISIR coronagraphic annular groove mask.
Figure 6. VISIR burst mode image of the close

v icinity of the Circinus AGN taken with the PAH2_2
narrow filter.
spatial frequency is a measure of the

spatial extent of the object. The goal of
the observations is to constrain the position angle of the presumed bipolar cavities and the morphology of the outer
envelope surrounding the central source.
A preliminary analysis shows that the
target has an angular size of at least
~450 milliarcseconds, which is consistent with the size of the 11.7 m envelope
reported by Vehoff et al. (2010) and by
Sanchez-Bermudez et al. (2016) at lower
wavelengths. With an angular resolution
of /2D, where D is the maximum diameter of the sparse aperture mask, and
a robust calibration in the closure phases,
the new sparse aperture masking mode
of VISIR is the only technique that enables the different contributions to the
NGC3603 IRS9A morphology at the halfarcsecond angular scale to be disentangled in the mid infrared.
AGN torus and situated in the polar

region, on scales of a few parsec from
the centre. The image shown in Figure6
was obtained using the PAH2_2 filter
at ~12m. It is dominated by the bright
central point-like source, which is responsible for the Airy rings, illustrating the
superb image quality that VISIR reaches
in burst mode. The unresolved emission
comes from warm dust surrounding
the AGN. Owing to the excellent image
quality, an extended, bar-like structure
becomes nicely visible. This emission
traces dust along the edge of the ionisation cone on scales of tens of parsecs,
well outside the expected AGN torus.
The WolfRayet star WR104 is carbonrich and classified as a late WC-type. It

shows an inner structure indicative of
a binary system and an infrared excess
presumably arising from warm and hot
dust components. WR104 was observed
at 12.4 m with the new annular groove
phase mask coronagraph to detect the
extended dust surrounding the central
star. Figure 4 shows the outer structure
with the central star light blocked behind
the coronagraphic mask. An attenuation
of the light of the central star by a factor
of 42 is achieved by the coronagraph.
16
Figure 5. Demonstration of the AGPM coronagraph

on the visual binary star Oph (HR6378) with the
broad 12.4 m filter. The upper image shows the
binary without the annular groove phase mask, and
the lower image with the coronagraph occulting the
left-hand star.
A further example of an observation with

the coronagraphic annular groove mask
is shown in Figure 5. This observation
shows the binary star Oph (HR 6378)
with and without the mask inserted. In the
right image with the mask inserted, the
cancellation of the star beneath the mask
is excellent. These observations were
taken during commissioning of the annular groove phase mask in January 2016.
Circinus is one of the nearest galaxies to
the Milky Way with an active nucleus.
It is hidden behind the Galactic Plane
and oriented almost edge-on, so that its
nucleus can be observed best in the
infrared. The goal of the VISIR burst
mode observations was to investigate a
possible emission of dust outside the
References
Kufl, H. U. et al. 2015, The Messenger, 159, 15
Kerber, F. et al. 2014, MNRAS, 439, 247
Kerber, F., Querel, R. R. & Neureiter, B. 2015,
J. Phys., Conf. Ser., 595, 012017
Sanchez-Bermudez, J. et al. 2016, A&A, 588, A117
Vehoff, S. et al. 2010, A&A, 520, 78
Links
1
Call for VISIR upgrade SV proposals:

http://www.eso.org/sci/publications/announcements/
sciann15081.html
2
ESO Science Newsletter November 2015:
http://www.eso.org/sci/publications/newsletter/
nov2015.html
3
V ISIR upgrade SV page: http://www.eso.org/sci/
activities/vltsv/visir-upgrade.html
4
V LT instrument pipelines: http://www.eso.org/sci/
software/pipelines/
Astronomical Science
FORS2 image of the Galactic star-forming region
RCW 34 (also catalogued as Gum 19). This colour
composite, formed from images in broadband
filters (B, V, R) and a narrowband H filter, emphasises the diffuse emission ionised by the central
O-type star. This young region also has many lowmass stars in the process of emerging from the
parental molecular cloud. See Release eso1521
for details.
A Stellar Census in NGC 6397 with MUSE
Sebastian Kamann1
Tim-Oliver Husser 1
Martin Wendt 2, 3
Roland Bacon 4
Jarle Brinchmann 5, 6
Stefan Dreizler 1
Eric Emsellem 7, 4
Davor Krajnovi 3
Ana Monreal-Ibero 8
Martin M. Roth 3
Peter M. Weilbacher 3
Lutz Wisotzki 3
1
Institut fr Astrophysik, Universitt

Gttingen, Germany
2
Institut fr Physik und Astronomie,
Universitt Potsdam, Germany
3
Leibniz-Institut fr Astrophysik (AIP),
Potsdam, Germany
4
CRAL, Observatoire de Lyon, SaintGenis Laval, France
5
Sterrewacht Leiden, Universiteit Leiden,
the Netherlands
6
Instituto de Astrofsica e Cincias do
Espao, Universidade do Porto, Portugal
7
ESO
8
GEPI, Observatoire de Paris, PSL
Research University, CNRS, Universit
Paris Diderot, Sorbonne Paris Cit,
Meudon, France
The new and powerful integral-field
spectrograph on the VLT, the MultiUnit Spectroscopic Explorer (MUSE),
was designed to search for distant
galaxies to an unprecedented depth,
but it is also capable of opening new
science windows on the Galaxy. To
demonstrate this capability, the globular cluster NGC6397 was observed
during the commissioning of MUSE in
August 2014. We outline how the analysis of this unique dataset allowed us
to assemble the largest spectroscopic
sample of stars in a globular cluster
to date. We also highlight the scientific
applications that benefit from such
MUSE data.
MUSE (Bacon et al., 2012) is an optical
integral-field spectrograph that observes
a continuous field of view of 1 by 1 arcminute on the sky, sampled at a spatial
resolution of 0.2 arcseconds. The instrument splits the field of view into 24 slices,
each feeding a different spectrograph.
18
The spectrographs operate at a medium

resolution, R, of 17003500, which allows
for the inclusion of a large wavelength
range, 4800 to 9300 , in a s ingle
exposure. The data reduction process
which transforms the 24 raw CCD images
into a three-dimensional datacube is quite
complex and summarised in Weilbacher
et al. (2012). A fully reduced cube contains about 300 300 spaxels (spatial
elements), which each consist of about
3600 spectral elements.
MUSEs design as a true spectrophotometric instrument with the capability to
observe several thousand stars simultaneously makes it a very powerful tool for
the investigation of stellar fields. There
are two reasons for the large multiplexing
factor. First, the number of spaxels is
much higher than for any other existing
integral-field spectrograph. Second, the
continuous spatial coverage at a sampling below the atmospheric seeing
allows techniques to disentangle the light
Figure 1. VRI colour image created from the MUSE

mosaic of NGC 6397. The image is 5 5 arcminutes
in size and consists of 23 individual MUSE pointings.
The images have been corrected for varying background levels.
contributions of blended stars to be

used. Such techniques are of crucial
importance, especially in crowded stellar
fields such as the central regions of
globular clusters, and will be described
in more detail below.
The globular cluster NGC6397 is, at a
distance of ~ 2.3 kpc, one of the closest
Galactic globular clusters. It has a mass
of about 105 MA and its metallicity of
[Fe/H] = 2 is low, even when compared
to other Galactic globular clusters (Harris
et al., 1996). The central 5 5 arcminute
region of NGC6397 was observed during
the third MUSE commissioning run (see
Bacon et al., 2014) by means of a mosaic
of 23 pointings (with two outer pointings
missing due to constraints on the commissioning activities). The total mapped
@",#
A'23

Figure 2. Example of the successful extraction of stellar spectra

from MUSE data. Panel (a) shows
a colourmagnitude diagram of
NGC 6397 p
lotted with Hubble
Space Telescope (HST) photometry from Anderson et al. (2008).
As can be seen in panels (b) and
(c), the two stars highlighted in red
and blue appear strongly blended
in the MUSE data and even in an
HST image. Nevertheless, the
extracted spectrum of the blue
star shows the broad Paschen
bands that are characteristic of
hot horizontal branch stars and the
red star shows the strong c
alcium
triplet typical for spectra of red
giant stars.
B,42$

5m(

-NQL@KHRDCkTW
C$WSQ@BSDCRODBSQ@

6@UDKDMFSG

area is shown as a colour image made

from the MUSE data in Figure 1. The
observations of the central 3 3 pointings benefited from very good seeing
(~0.6 arcseconds), whereas the seeing
was higher (~1.0 arcsecond) during the
observations of the outer fields. Further
details about the data collection and
processing are presented in Husser et al.
(2016).
domain of integral-field spectroscopy and

uses the point spread function (PSF) of
the observations to deblend the spectra
of nearby stars. We designed the software package PampelMuse, which we
successfully used to analyse the MUSE
data of NGC 6397, around this concept.
Figure 2 shows that even for stars separated by only 0.2 arcseconds, i.e., about
one third of the width of the seeing,
uncontaminated spectra can be extracted.
Extraction of spectra
As described in detail in Husser et al.

(2016), we could extract 18 932 spectra
for 12 307 stars from the full MUSE
mosaic of NGC 6397, making this the
largest spectroscopic sample obtained
so far in any globular cluster. The spectra
cover a large range of spectral types
and reach down to a magnitude of about
V = 19, several magnitudes below the
main sequence turn-off of NGC 6397.
The spectra are made available online1.
Figure 1 gives a good impression of the

typical stellar crowding in the central
regions of globular clusters, which can
pose a severe limitation for spectroscopic
observations. For example, in a multiobject spectrograph, a fibre is placed on
the image of every star of interest. However, near the centre of a globular cluster
every such fibre will also collect a fraction of light from the stars close neighbours, leading to a contamination of the
observed spectrum. In Kamann et al.
(2013), we introduced the concept of
crowded-field 3D spectroscopy to overcome this issue. It represents a con
tinuation of optimal extraction algorithms
developed for imaging data (such as
DAOPHOT, Stetson [1987]) into the
Spectral analysis
The analysis of the extracted spectra is
a multi-step process that starts with
estimating stellar parameters from photo
metry obtained with the Hubble Space
Telescope (see small inset in Figure 4).

We compare the brightness and colour

of each observed star with an isochrone
that matches the colourmagnitude
diagram (CMD) of the cluster, yielding an
effective temperature and a surface
gravity. Using these parameters, a synthetic spectrum is created and used as
a template for a cross correlation with
the observation in order to derive a radial
velocity.
The actual analysis is performed via a
LevenbergMarquardt optimisation that
finds the best matching template in a
grid of synthetic stellar spectra, using the
previously determined values as initial
guesses. As a result, we obtain stellar
parameters like effective temperature,
metallicity and -element abundance, as
well as a radial velocity. The surface gravity is currently fixed to the one derived
from photometry.
A basic principle for all our analyses is
never to alter the observed spectra,
since every operation, such as re-binning
or normalisation, would also result in a
loss of information. Instead, we leave
the observed spectrum untouched and
only change the model spectra. So, for
instance, we never remove the continuum
flux from the observed spectrum, but try
19
Kamann S. et al., A Stellar Census in NGC 6397 with MUSE

An example of an analysed spectrum is

shown in Figure 3. The observed spectrum in black is overplotted with the model
spectrum in red that has been found to
best match the observation. The residuals are plotted in blue below.

6@UDKDMFSG
( %65$& , &

Figure 3. Steps in the

data processing. In
black the PSF-extracted
spectrum of one of the
brightest stars is shown.
The red line shows the
best fit, including a telluric absorption correction. In blue, the fitting
residuals are displayed
to scale with the data
and fit. The small inserts
zoom into spectral
regions of interest for
ISM analyses. The left
panels show the prominent NaD doublet as
well as diffuse interstellar bands at 5780 and
5797. On the righthand side, the insets
illustrate the success of
the telluric fit where the
weak K I doublet lines
clearly stand out. The
zoom for the residuals is
scaled by a factor of ten.

+TLHMNRHSX++
Results for the globular cluster NGC6397

as a whole are shown in Figure 4 in the
form of an HertzsprungRussell diagram
(HRD), plotting the luminosity as a function of effective temperature. For the
luminosity, we derived V-band magnitudes
from the spectra and applied a bolo
metric correction based on the fitted stellar parameters. All the stars are colourcoded with their corresponding metallicity
from the analysis. While the variation of
the metallicity along the main sequence
is presumably due to low signal-to-noise
ratio (S/N) in this part of the HRD, the
trend on the giant branch may indeed be
real, as it has been observed before by
other groups and instruments. For
instance, Korn et al. (2007) interpreted
this variation as the result of atomic diffusion in the stellar atmosphere. The results
for NGC 6397 are discussed in detail in
Husser et al. (2016).

5%6( %65$& , &

While the results for single stars are

already of good quality, they cannot compete with those from high-resolution
spectroscopy. But the greatest strength
of our MUSE observations lies in the
unprecedented amount of data. Instead
of measuring, for instance, the metallicity
in a few high-resolution spectra, we can
provide a mean value and spread for
the metallicity for thousands of stars,
either for the whole cluster or limited to a
small region in the CMD. Furthermore,
the large number of spectra allows us to
improve the S/N, especially on the main
sequence, by co-adding spectra, either
from multiple visits to the same star or
from neighbouring stars in the CMD.
20

$EEDBSHUDSDLODQ@STQD3DEE*
Figure 4. The small inset shows the colourmagnitude

diagram of NGC 6397. All the stars observed with
MUSE are colour-coded with the signal-to-noise
ratios of their respective spectra. The large plot
shows the HertzsprungRussell diagram using the
stellar parameters from the analysis. Here the colour
indicates the derived metallicity of each star.

,DS@KKHBHSX:,'<
to find a polynomial that, when multiplied

by the model, best matches the observation. Furthermore, instead of removing
the telluric absorption lines by means of
observations of a telluric standard star, we
try to model them. Abundances of water
and molecular oxygen are free parameters in the optimisation as well as a line
shift and broadening for the telluric spectrum. This approach improves the quality
of the derived parameters significantly.
2HFM@KSNMNHRDQ@SHN
Cluster dynamics
+HSDQ@STQD
With respect to the cluster dynamics, the

central region is the most interesting one.
For example, there is an ongoing debate
about the presence of massive black
holes, weighing about 102105 MA, in the
centres of globular clusters (see, e.g., van
der Marel et al. [2010] and Noyola et al.
[2010]). However, a common problem of
spectroscopic studies is that they can
only target isolated stars, where contamination from nearby sources is negligible.
Thanks to the spatial coverage of MUSE
and our deblending algorithm, we are
able to overcome this problem.
Figure 5 shows that our measurements
extend much further into the centre
than previous radial velocity studies of
NGC6397, allowing us, for the first time,
to constrain the presence of a massive
black hole in this cluster. To do so, we
compared our measurements to dynamical models, some of which are depicted
in Figure 5. We found that the velocity
dispersion in the centre is higher than
what would be expected from the gravitational potential of the bright stars alone. A
black hole with a mass of about 600MA
would be an intriguing explanation for this
discrepancy. However, it is not the only
possible explanation. Alternatively, a cen-
,42$

mQJLR l
With a mass of around 10 MA, NGC 6397

is only moderately massive when compared to other globular clusters in the
Milky Way. For example, Omega Centauri
is more than ten times as massive as
NGC 6397. A consequence is that the
internal dynamics of NGC 6397 are dominated by low velocities the central
dispersion is expected to be as low as
5km s 1. This poses a severe challenge
for spectroscopic studies with the low
spectral resolution offered by MUSE,
because they must achieve an accuracy
in radial velocity that is higher than the
intrinsic cluster dispersion. From the
analysis of telluric absorption bands in
the extracted spectra, we could show in
Kamann et al. (2016) that the internal
accuracy of MUSE is stable at a level of
1km s 1, both across the field of view
and over the course of a night. Given the
complexity of MUSE, this is a remarkable
result that confirms the high stability of
the instrument and the excellent quality of
the data reduction pipeline.

,NCDKTRHMFNMKXAQHFGSRS@QR

,NCDKVHSGBDMSQ@KAK@BJGNKD
,NCDKVHSGBDMSQ@KQDLM@MSR

KNFQBDMSQD@QBRDBNMC

Figure 5. Velocity dispersion of NGC 6397 as a function of distance to the cluster centre as measured by
MUSE (green diamonds) and from a compilation of
literature studies (grey squares). The different lines
show the expected velocity dispersion curves based
on the gravitational potential of the bright stars alone

(black solid line), with the addition of a black hole
with 600 MA (blue dashed line), and with the addition
of a central accumulation of stellar remnants with a
similar mass (red dash-dotted line).
tral accumulation of stellar remnants

(such as neutron stars or stellar-mass
black holes), which may form as a consequence of mass segregation in the cluster, could also explain our measurements.
Further details about our analysis and
possible ways to distinguish between the
two alternatives in the future can be
found in Kamann et al. (2016).
of this phenomenon requires the observation of many stars along the main
sequence, because giant stars all have
more or less similar masses, and is therefore extremely challenging. In the MUSE
data, we found a marginal trend for more
massive stars to have a lower central
velocity dispersion. Further studies are
required to settle this issue, but the commissioning data of NGC6397 already
show the potential of MUSE in this
respect.
The diagnostic power of the MUSE data

is not limited to the search for massive
black holes. Thanks to the large stellar
sample, we can also look at the cluster
dynamics in a two-dimensional way. In
doing this, we identified a small rotational
component, with a projected amplitude
of about 1 km s 1 around the centre. In
addition, we could investigate whether
the stellar dynamics change depending
on the masses of the investigated stars.
Such a dependency can be caused by
relaxation processes inside the cluster.
Gravitational encounters between member stars will on average accelerate the
less massive stars and decelerate the
more massive stars, ultimately leading to
mass segregation. The investigation
Interstellar medium
The template matching of the individual
stellar spectra and the comprehensive
sky model fits are quite successful. In
fact, they are robust to such an extent
that we can carry out further studies on
the fitting of the residuals themselves,
which still feature absorption lines and
bands of the interstellar medium (see
Figure 3). This is a field of research for
which MUSE was not even designed.
This study provides a unique insight into
small scale structures in the interstellar
21
Kamann S. et al., A Stellar Census in NGC 6397 with MUSE
Figure 6. Equivalent width map of the NaD doublet.

All residual spectra were combined into Voronoi
tessellated bins. The colour bar shows the equivalent width range as well as the average value for
this species. The uncertainty per bin is of the order
of 8 m.

To reach a higher S/N, we computed

the error-weighted average of ~ 300
residual spectra to form several com
posite spectra of similar S/N. The accuracy of the telluric and stellar fits is very
high. The inset in Figure 3 shows both
doublet lines of the weak interstellar K I.
While K I 7664 sits directly on a strong
telluric band, K I 7699 is hardly affected
by skylines at all. Their independently
fitted ratio remains very stable and provides great confidence in the applied
method. In fact, we can even utilise that
ratio as a diagnostic tool to trace optical
thickness. Since we know that we can
successfully subtract stellar features
as well as sky lines, we continued to systematically analyse other weak ISM features, such as a number of diffuse interstellar bands that we observe in the
broad wavelength range of the MUSE
spectrograph. This analysis will be
described in a third follow-up paper on
these observations by Wendt et al.
Other, stronger ISM lines even had to be
considered during the template matching
itself. A particular challenge is the NaD
doublet. Here, we expect at least three
unresolved contributors: the stellar component, NaD in the Earths atmosphere,
as well as the ISM component(s) along
the line of sight. The first two contributions are subtracted. Figure 6 shows the
equivalent width of the remaining (interstellar) contributions in the 31 Voronoi
tessellated bins (spaxel spectra co-added
to increase the S/N). The colours reflect
the measured total equivalent width of the
NaD doublet for each composite spectrum per bin with an average of about
960 m, and the MUSE pointings are
indicated as black dashed lines; a red
star marks the centre of the globular
cluster. The mapping reveals a compelling small-scale structure in interstellar
NaD that is neither correlated with the
pointings, nor the number of stars per
22
6#DB@QBRDBNMC

medium (ISM) that could not be traced or

spatially resolved with individual isolated
spectra.

$6L

61 @QBRDBNMC
bin, i.e., with the cluster itself. At the distance of 100 pc for the edge of the Local
Bubble, the linear projection for typical
scale sizes is in the order of a milliparsec.
This illustrates how MUSE is uniquely
able to provide an overview of the smallscale structures of the ISM.
Prospects
The example of NGC 6397 has shown
the huge potential of MUSE for the investigation of crowded stellar fields. The
unprecedented number of stars for which
spectra can be acquired simultaneously
enables completely new science cases.
We are currently conducting a large survey of 25 Galactic globular clusters with
the aim of obtaining multi-epoch spectroscopy for several thousand stars per
cluster. In addition to detailed investigations of stellar parameters, the central
dynamics and the ISM, this survey will
also reveal clues about the properties of
binary stars in the clusters.
Following the installation of the Ground
Atmospheric Layer Adaptive Optics for
Spectroscopic Imaging (GALACSI) system (see Strbele et al., 2012), MUSE
observations at a significantly higher spa-

tial resolution will soon be possible. In

crowded stellar fields, this improvement
will even further increase the number of
accessible stars. As such fields are not
specific to globular clusters, but are also
found in the Galactic Bulge or nearby
galaxies, we believe that there are huge
prospects for MUSE observations similar
to those that we have presented for
NGC6397.
References
Anderson, J. et al. 2008, AJ, 135, 2055
Bacon, R. et al. 2012, The Messenger, 147, 4
Bacon, R. et al. 2014, The Messenger, 157, 13
Harris, W. E. et al. 1996, AJ, 112, 1487
Husser, T.-O. et al. 2016, A&A, 588, 148
Kamann, S. et al. 2013, A&A, 549, 71
Kamann, S. et al. 2016, A&A, 588, 149
Korn, A. J. et al. 2007, ApJ, 671, 402
Noyola, E. et al. 2010, ApJ, 719, 60
Stetson, P. B. 1987, PASP, 99, 191
Strbele, S. et al. 2012, Proc. SPIE, 8447, 844737
van der Marel, R. et al. 2010, ApJ, 710, 1063
Weilbacher, P. M. et al. 2012, Proc. SPIE, 8451,
84510B
Links
1
O nline access to extracted spectra: h

ttp://muse-vlt.
eu/science/globular-cluster-ngc-6397/
Pulsating Hot Subdwarfs in Omega Centauri
Suzanna K. Randall 1
Annalisa Calamida 2
Gilles Fontaine 3
Matteo Monelli 4
Giuseppe Bono 5, 6
Maria Luisa Alonso 7
Valrie Van Grootel 8
Pierre Brassard 3
Pierre Chayer 2
Marcio Catelan 7
Stuart Littlefair 9
Vik S. Dhillon 9, 4
Tom R. Marsh 10
1
ESO
Space Telescope Science Institute,
Baltimore, USA
3
Dpartement de Physique, Universit
de Montral, Canada
4
Instituto de Astrofisica de Canarias,
Tenerife, Spain
5
Universita di Roma Tor Vergata,
Department of Physics, Rome, Italy
6
INAF Osservatorio Astronomico di
Roma, Monte Porzio Catone, Italy
7
Pontificia Universidad Catolica de Chile,
Santiago, Chile
8
Institut dAstrophysique et de
Gophysique de lUniversit de Lige,
Belgium
9
Department of Physics and Astronomy,
University of Sheffield, United Kingdom
10
Department of Physics, University of
Warwick, Coventry, United Kingdom
2
The formation and evolution of hot sub

dwarf B- and O-type stars (sdB and
sdO) stars is one of the remaining mysteries in stellar evolution theory. While
it is commonly accepted that the cooler
sdB stars (Teff ~ 20 00040 000 K) are
compact and evolved extreme horizontal
branch (EHB) stars that have lost too
much of their hydrogen envelope near
the tip of the red giant branch (RGB) to
sustain hydrogen-shell burning on the
asymptotic giant branch (AGB) after core
helium ignition, the details of their evolution, in particular the dramatic mass loss,
remain unclear. The hotter sdO stars
(Teff~4000080000 K) are even more
challenging to understand, likely comprising a mixed bag of post-EHB, post-RGB
and post-AGB stars.
One of the most promising ways of testing different evolutionary scenarios is
by inferring fundamental properties, such
as the total stellar mass or the thickness
of the remaining hydrogen envelope,
from asteroseismological analysis of pulsating hot subdwarfs. This has been
done with particular success for the rapidly pulsating subdwarf B (sdBVr) stars
(Fontaine et al., 2012). These are a well-
established class of pulsator found in a

narrow instability strip between ~ 29000
and 36000K, characterised by multiperiodic luminosity variations on a short
timescale of 100200 s. The observed
variations can be explained very nicely in
terms of non-radial pressure mode instabilities driven by an opacity mechanism
associated with a local overabundance of
iron in the driving region.
Although hot subdwarfs are known in
large numbers both among the Galactic
field population and in globular clusters
(where they are identified as EHB or
blue hook stars), the exploitation of their
pulsations has traditionally been limited
to the field star population for the simple
reason that, until recently, no hot sub
dwarf pulsators had been found in a
globular cluster.
A search for hot subdwarf pulsators in
Cen
Like so many observational astronomy
discoveries, the identification of the first
hot subdwarf pulsator candidate in a
globular cluster was serendipitous. We

"2 '23
6%( L
55

1&!

!L@F
We recently discovered the first globular cluster hot subdwarf pulsators in

Omega Centauri ( Cen). These stars
were initially thought to belong to the
class of rapidly pulsating subdwarf B
stars, which are well established among
the field star population and have
become showcases for asteroseis
mology. However, our spectroscopic
analysis revealed the Cen variables to
be significantly hotter than expected,
indicating that they form a new class of
subdwarf O pulsators clustered around
50000 K, not known among the field
star population. Non-adiabatic pulsation
modelling suggests that the driver for
the pulsations occurs via the same iron
opacity mechanism that is at work in
the rapidly pulsating subdwarf B stars.

$'!

,2
6#

!1L@F

Figure 1. Colourmagnitude
diagram of Cen based on
the merged HST ACS/WFI
catalogue. The box shows
the colourmagnitude cuts
applied for our selection of
EHB star candidates, and
the five rapid sdO pulsators
are indicated by red
crosses.
23
This initial discovery paved the way for

an extensive follow-up survey with two
main observational components: timeseries photometry obtained with EFOSC2
and ULTRACAM at the NTT, and multiobject spectroscopy gathered using
FOcal Reducer/low dispersion Spectrograph (FORS2) at the Very Large Telescope (VLT). More details can be found in
Randall et al. (2016).
The time-series photometry amounts to
nearly 100 hours of fast-cadence (~10s)
monitoring of some 300 stars scattered
across off-centre fields in Cen and
identified as EHB candidates based on a
colourmagnitude cut in a merged Hubble
Space Telescope Advanced Camera for
Surveys (ACS) Wide Field Imager (WFI)
catalogue (Castellani et al., 2007). The
corresponding colourmagnitude diagram
of Cen as well as the colourmagnitude
ranges used for our EHB target selection
are shown in Figure 1. It appears that
this simple selection method is quite reliable, since from our spectroscopic sample of 60 stars only two turned out to not
be hot subdwarfs.
From the photometry we were able to not
only confirm the variability of the pulsator
candidate identified with SUSI2, but also
discovered an additional four pulsating
EHB candidates, bringing the total number to five (denoted V1V5 in the order
that they were discovered). The pulsational properties of the variables are quite
similar, each star showing two to three
well-separated oscillations in the 85125s
range with amplitudes up to ~2.5% of
their mean brightness. Figure 2 shows
examples of the Fourier amplitude spectra obtained, the pulsations being clearly
24

LOKHSTCD

5

LOKHSTCD
were obtaining time-series photometry

with the Superb Seeing Imager 2 (SUSI2)
on the New Technology Telescope (NTT)
for an unrelated observing programme
and chose a more or less random field in
Cen as a backup target at the end of
the night. Since at that point the hunt for
sdBVr stars in bright nearby globular
clusters had been going on for some time
without success, we could hardly believe
it when our analysis of this short twohour dataset revealed a 115-second periodicity in a star for which the colour
magnitudes were consistent with a hot
subdwarf.
Randall S. K. et al., Pulsating Hot Subdwarfs in Omega Centauri
5

5

LOKHSTCD

%QDPTDMBXL'Y
visible above the 3.7 detection threshold

imposed. It is worth noting that the highest amplitude peaks show fine frequency
structure corresponding roughly to the
resolution of the dataset, which are attributed to significant amplitude variations of
the pulsations. Such amplitude variations
are well documented for the field sdBVr
stars and are therefore not unexpected.
Surprising results from spectroscopy
Given our knowledge of hot subdwarf
pulsators in the field and the pulsational
properties observed for the Cen variables, we quite naturally assumed we had

Figure 2. Fourier amplitude spectra for three of the

Cen variables based on the combined ULTRACAM
u: light curves. The horizontal dashed line indicates
the 3.7 detection threshold imposed.
found the long-sought-after globular

cluster counterparts to the sdBVr stars.
However, the spectroscopic data of the
new variables painted a different, even
more intriguing picture: rather than the
expected temperatures around 30000K,
our atmospheric analysis yielded significantly higher values around 50000K,
implying that these stars are in fact a
new type of sdO pulsator never observed
among the field star population!
Figure 3. Location in Teff log g space of the Cen

variables compared to the different types of field hot
subdwarf pulsators known. The reference to the discovery paper is indicated for each.
RC!5R
&QDDMDS@K

t"DMRC.R
1@MC@KKDS@K

RC!5QR
2BGTGDS@K

KNFF
This can be fully appreciated from Figure3, where we show the location of
the Cen pulsators compared to the
known field hot subdwarf pulsators. Only
the previously mentioned sdBVr pulsators
and the one lone field sdOV star show
periods on short timescales comparable
to the Cen variables, while the other
types of pulsator oscillate on longer timescales of one to several hours. None of
the field star variables falls in the ~48000
52000K range where the Cen pulsators are found. Conversely, we did not
identify any counterparts to the field star
variables in Cen, although this is quite
possibly due to the limitations of our
dataset both in terms of quality and sample size.

'DlRC!5
GLDC)DEEDQX

The empirical Cen instability strip

derived from our survey is shown in Figure4. Here we have included only the
26targets for which we have both high
quality light curves from photometry as
well as reliable atmospheric parameters
from spectroscopy. Apart from a clear
bias towards the pulsators, which were
specifically included in the spectroscopic
sample, there should be no additional
selection effects beyond the colour
magnitude cuts described earlier. From
the plot it is not entirely clear whether
all stars falling within the instability strip
pulsate, or whether non-variable stars
co-exist in the same region of Teff log g
space, as is the case for the sdBVr pulsators in the field.

RC.5
6NTCSDS@K

RC!5Q
*HKJDMMXDS@K

KNF3DEE*

Pulsation calculations
Keeping in mind their very similar pulsation properties and relative proximity in
atmospheric parameter space, it seemed
likely that the pulsation driving mechanism active in the Cen variables is the
same opacity mechanism that explains
the sdBVr instability strip so well. Therefore, we extended our existing stellar
envelope models to higher temperatures,
encompassing the entire range where
hot subdwarfs are found. These so-called
Montral second-generation models
incorporate traces of iron that are levitating
KNFF

3DEE*
Figure 4. Observed hot

subdwarf instability
strip in Cen. The five
pulsators are marked
by large open circles.
Red symbols indicate
H-rich stars, blue ones
refer to He-rich targets.
25
Randall S. K. et al., Pulsating Hot Subdwarfs in Omega Centauri
to be improved, particularly with regards

to other metals besides iron being
included in the diffusion calculations.
Nevertheless, we are confident that we
have identified the basic driving mechanism in the Cen pulsators to be the
same opacity mechanism that is at work
in the sdBVr stars.

KNFF
The plot thickens

3DEE*

Figure 5. Theoretical instability strip for rapid pressure-mode pulsations in hot subdwarfs. Each red
point indicates a model where pulsations are driven.
The blue circles show the location of observed
sdBVr stars in the field, while the black cross represents the one sdO field pulsator known. The Cen
pulsators are depicted as black open circles with

error bars; the dotted extension to higher log g values for one of them indicates that the spectroscopic
log g is likely underestimated due to contamination
from nearby stars.
in a pure hydrogen envelope under the

assumption that an equilibrium has
been reached between radiative levitation
and gravitational settling (Charpinet et
al., 1996; 1997). The iron abundance profile as a function of stellar depth is then
necessarily a function of log g and Teff.
Since the presence of an opacity bump
associated with an overabundance of iron
in the driving region allows pulsations to
be excited, it follows that this will happen
only for models with certain log g/Teff
combinations.
els around 30 000 K, but also at higher

temperatures above ~ 50 000 K. This can
be seen nicely in Figure 5, where we
show the theoretical instability regions for
the entire log g Teff space where hot
subdwarfs are found.
From our non-adiabatic oscillation cal

culations, we find that pressure-mode
pulsations (corresponding to short periods up to a few hundred seconds) are
indeed driven not only for sdB star mod-
26
Unfortunately, while the cooler instability

strip matches the location of the sdBVr
stars perfectly, the hotter instability
region does not quite reach the Cen
pulsators, which lie beyond the predicted
red edge. This mismatch may be partially
due to the effective temperatures from
our optical spectra being underestimated
for these very hot stars, however the
periods of the oscillations predicted are
also significantly shorter than those
observed. It is clear that the models need
Far from giving definite answers, the

results of our survey have instead raised
more questions. But that is the beauty
of observational astronomy: you rarely
find what you expect, and the deeper you
look the more complicated the picture
becomes, opening up new avenues of
research. In this case, the unexpected
discovery of a hitherto unknown type of
pulsator in Cen triggered an intense
search for direct counterparts among the
field population, which however came
back negative (Johnson et al., 2014). On
the other hand, a space-based search
for rapid hot subdwarf pulsators in the
globular cluster NGC 2808 (Brown et al.,
2013) revealed six rapid pulsators with
periods on timescales similar to the sdOV
and sdBVr stars discussed here. Given
the low quality of the available spectroscopy, it is completely unclear whether
any of these pulsators correspond to the
Cen variables, or indeed any of the
known field hot subdwarf pulsators. Further investigations and more detailed
observations are clearly warranted.
References
Ahmad, A. & Jeffery, C. S. 2005, A&A, 437, 51
Brown, T. M. et al. 2013, ApJ, 777, L22
Castellani, V. et al. 2007, ApJ, 663, 1021
Charpinet, S. et al. 1996, ApJ, 471, L103
Charpinet, S., Fontaine, G. & Brassard, P. 1997,
ApJ, 483, L123
Fontaine, G. et al. 2012, A&A, 539, A12
Green, E. M. et al. 2003, ApJ, 583, 31
Johnson, C. et al. 2014, ASP Conf. Ser., 481, 153
Kilkenny, D. et al. 1997, MNRAS, 285, 640
Randall, S. K. et al. 2011, ApJ, 737, 27
Randall, S. K. et al. 2016, A&A, 589, 1
Schuh, S. et al. 2005, ASPC, 334, 530
Woudt, P. A. et al. 2006, MNRAS, 371, 149
First Results from the XXL Survey and Associated

Multi-wavelength Programmes
Christophe Adami 1
Marguerite Pierre 2
Nikola Baran 3
Dominique Eckert 4
Sotiria Fotopoulou 4
Paul A. Giles 5
Elias Koulouridis 2
Chris Lidman 6
Maggie Lieu 7
Adam B. Mantz 8
Florian Pacaud 9
Emanuela Pompei 10
Vernesa Smoli 3
Felicia Ziparo 8
and the XXL Team 11
1
Aix Marseille Universit, CNRS, LAM

UMR 7326, Marseille, France
2
Laboratoire AIM, CEA/DSM/IRFU/SAp,
CEA Saclay, Gif-sur-Yvette, France
3
Department of Physics, University of
Zagreb, Croatia
4
Department of Astronomy, University
of Geneva, Versoix, Switzerland
5
H. H. Wills Physics Laboratory,
University of Bristol, United Kingdom
6
Australian Astronomical Observatory,
North Ryde, Australia
7
School of Physics & Astronomy, University of Birmingham, United Kingdom
8
Department of Astronomy and Astrophysics, University of Chicago, USA
9
Argelander Institut fr Astronomie,
Universitt Bonn, Germany
10
ESO
11
XXL Team: http://irfu.cea.fr/xxl/members
The XXL survey has mapped two extragalactic regions of 25 square degrees,
using 10 ks XMM observations down to
a point source sensitivity of ~5 10 15
erg s 1 cm 2 in the 0.52 keV band. It
is the largest XMM project approved to
date (> 6 Ms in total). The two fields
have been or will be observed by several ground- and space-based facilities
from ultra-violet to radio wavelengths.
Besides the imaging, the spectroscopic
follow-up is of special interest and ESO
has contributed through Large Programmes and dedicated allocations. As
of December 2015, ~ 450 new galaxy
clusters are detected to z ~ 1.52, as
well as more than 22000 active galactic
nuclei (AGN) to z ~ 4. The main goal of
the project is to constrain the dark
energy equation of state using clusters
of galaxies. This survey will have lasting

legacy value for cluster scaling laws and
studies of AGNs and the X-ray background. The first XXL scientific results
are summarised.
Introduction
The distribution of matter in the Universe
reveals its history: small density pertur
bations present at the time of recombi
nation (just after the Cosmic Microwave
Background [CMB] was emitted) grew
under the competing actions of gravity
and expansion of spacetime. This led to
the progressive formation of a remarkable
network of filaments, sheets and voids
that was first revealed from the galaxy
distribution some 30 years ago and is
now well rendered by numerical simulations. Basically as time proceeds, overdensities become denser and low density
regions emptier as matter flows along
the filaments. The rate at which structure
forms and how it will ultimately develop
is described by general relativity, depending on a set of cosmological parameters
and on the nature of dark matter and
dark energy (the latter being responsible
for the acceleration of the expansion).
Tracking the evolution of cosmic structure
in the quest for the cosmological parameters that describe the Universe is especially interesting, in particular to corroborate the cosmological constraints inferred
from the CMB (z ~ 1000) and from the
matter distribution at much later times
(z~04), since they originate from very
different physical processes. A traditional
approach follows the distribution of galaxies in visible or infrared light. However
galaxies account only for a few percent
of the total matter content and are highly
nonlinear objects, which makes establishing the connection with the initial density spectrum non-trivial. At the upper
hierarchical level, one finds clusters of
galaxies, the most massive objects in the
Universe. They are located at the nodes
of the cosmic web and are huge reservoirs of hot gas that fill the space between
the galaxies. The gas has a temperature
of a few tens of millions of degrees and
emits in the X-ray.
The cosmological analysis of cluster surveys basically relies on number counts
(dn/dz or dn/dM/dz) and on the spatial

analysis (3D correlation function). In order
to establish these functions, it is necessary to have a good knowledge of: (i) the
scaling relations that link the X-ray observables, like temperature and luminosity, to
the mass; and (ii) the selection function,
i.e., the probability of detecting a cluster
of given mass as a function of redshift.
In this context, the XXL project (XXL
Paper I) has undertaken a large survey
of the X-ray sky with the European
Space Agency XMM-Newton observatory. The primary goal of the XXL survey
is to detect and confirm a few hundred
clusters back to a time when the age of
Universe was about half of its present
value, i.e., z ~ 1. From this data collection,
we can infer the evolutionary properties
and spatial distribution of the deepest
potential wells of the Universe and, subsequently, test various cosmological
scenarios. In addition, XXL will deliver
several tens of thousands of active galactic nuclei (AGNs) out to a redshift of at
least four, allowing statistical studies of
the AGN population as a function of environment to be performed.
The XXL Survey is an XMM Very Large
Programme that covers two extragalactic
areas of 25 square degrees, located
around (RA, Dec) of (2h 20m, 4.5)
and (23h 30m, 55). Over 500 XMM XXL
observations, typically of 10 ks each,
were performed between 2011 and 2013.
Two years after the completion of the
X-ray observations, we have detected
~450 clusters and 22 000 AGNs in the
0.52.0 keV energy band. The XXL survey samples the low end of the cluster
mass function around z ~ 0.30.5. This
group population is especially interesting,
since it contains the building blocks of
the local massive clusters. Moreover, it
is sensitive to non-gravitational physics,
like the energy input from AGNs, feedback from star formation or cooling processes. Such groups thus present useful
environments in which to study the formation of cosmic structures and provide
a unique input to test physical processes
via numerical simulations. The XXL consortium gathers some 100 scientists
including observers, theoreticians and
simulators with a wide range of expertise
encompassing the entire electromagnetic
spectrum.
27
Adami C. et al., First Results from the XXL Survey and M

ulti-wavelength Programmes
l
#DBKHM@SHNM
l
l
l
l
l
.SGDQRODBSQNRBNOX
$2.RODBSQNRBNOX
"NMjQLDCBKTRSDQR

Associated multi-wavelength programmes

and XXL samples
In order to characterise the properties
of the detected clusters and AGNs, the
XXL survey is accompanied by a comprehensive follow-up programme. Observations are coordinated with the largest
ground-based and space observatories,
from the X-ray to the radio (see, for example, XXL paper X for the K-band luminosityweak-lensing mass relation for galaxy
structures). Of these coordinated observations, the ESO Large Programme and
principal investigator allocations (ESO
191.A-0268, 089.A-0666, and 60.A-9302),
as well as galaxy surveys in the XXL
fields, like the VIMOS Public Extragalactic
Survey (VIPERS; for example, Guzzo et
al., 2014) and the Galaxy And Mass
Assembly survey (GAMA; for example,
Driver et al., 2016), are of special relevance as they allow galaxy redshifts to
be measured, hence accurately locating
the galaxy clusters in 3D space (Figures1
and 2). In the framework of the ESO
Large Programme, we observed clusters
out to a redshift of ~0.5 at the New
Technology Telescope (NTT) with the
ESO Faint Object Spectrograph and
Camera 2 (EFOSC2) and the more distant
ones (0.5 < z < 1.2) with the FOcal
Reducer/low dispersion Spectrograph2
(FORS2) instrument on the Very Large
Telescope (VLT). An example from the
latter is the detection of a z ~1.9 cluster
(see XXL Paper V).
28

1HFGS RBDMRHNM

Cluster redshifts constitute a critical input

to cosmological analysis, allowing the
computation of the clustercluster correlation function and further topological
studies. Detailed investigations of the
dynamics and galaxy properties of individual clusters are also undertaken at the
William Herschel Telescope (WHT). The
redshift measurement of the AGN is systematically performed with the AAOmega
instrument on the Australian Astronomical Telescope (XXL XIV). Radio follow-up
of the X-ray AGNs is undertaken with
Giant Meterwave Radio Telescope (GMRT),
the Jansky Very Large Array (VLA) and
the Australia Telescope Compact Array
(ATCA). In parallel, four numerical simulation programmes provide the necessary
material to test in situ the various physical
mechanisms that drive cluster and AGN
evolution and their respective interactions.
Simulations also allow accurate deter
mination of the cluster selection function,
which is an essential ingredient for the
cosmological analysis.
In a first series of 13 articles published
in the December 2015 Astronomy &
Astrophysics special issue, plus one in
Publications of the Astronomical Society
of Australia, we released the catalogues
of the brightest 100 clusters (XXL-100-GC)
and 1000 AGNs (XXL-1000-AGN) as well
as the reduced XMM observations. We
present a number of scientific results
for the, to date, little-explored low-mass
range characterising the XXL cluster
sample.

Figure 1. Sky distribution

of the available spectro
scopy and confirmed
X-ray clusters of galaxies
in the northern XXL field.
The chessboard pattern
of red points corresponds to the mosaic of
VIMOS fields from the
VIPERS survey (Guzzo
et al., 2014), while the
darker spot at RA = 36.5
degrees shows the
deeper data from the
VIMOS-VLT Deep Survey
(VVDS; Le Fvre et al.,
2005). The isolated concentrations of red points
are from the XXL ESO
Large Programme
(191.A-0268) and represent ~ 50 % of the redshifts inside all the confirmed clusters at z > 0.5.
Scaling relations
Scaling relations are key elements for
parameterising cluster evolution and the
XXL project aims at a self-consistent
determination. We therefore investigated
the luminositytemperature (LT) relation of
the XXL-100-GC sample (XXL III, see also
XXL IV for the masstemperature relation).
The first step was to measure the cluster
X-ray luminosity and temperature from
the XMM survey data. The sample spans
a wide range of redshift (0.05 < z < 1.05),
temperature (0.6 < kT < 7.0 keV), and
luminosity (9 1041 < L < 5 1043 erg s 1)
and is equivalent to a flux-limited sample
(3 10 14 erg s1 cm2).
Our methods to determine the LT relation
fully take into account the selection effects
of the survey. We measure the evolution of
the LT relation internally using the broad
redshift range of the sample and find a
slope of the bolometric LT relation that is
steeper than the self-similar expectation.
Regarding evolution, our best fit is fully
consistent with strong self-similar evolution where clusters scale self-similarly with
both mass and redshift: clusters in the
past would appear as scaled replicates of
the local ones, as a function of mass.
However, this result is marginally more significant than the weak self-similar evolution solution, where clusters scale with
redshift alone. We further investigated the
sensitivity of our results to the assumptions made in our fitting model, finding
that the use of an external LT relation as
a low-z baseline can have a profound

effect on the measured evolution. The present calculations were performed assuming a given cosmology. The next step,
involving the complete XXL sample, will be
to implement the cosmological dependency of the luminosity in the determination
of the evolution rate of the LT relation.
We further used the XXL-100-GC sample

to infer preliminary cosmological constraints (XXL II). The measured luminosities and temperatures allowed us to model
the predicted cluster number counts for
the presently favoured set of cosmo
logical parameters. The number density
of clusters is found to be less than predicted by the Planck CMB cosmology,
once the XXL selection effects are duly
accounted for; this tension is similar
to that found for the Planck Sunyaev
Zeldovich (SZ) detected cluster sample,
but pertains to a cluster mass range about
one order of magnitude lower. However,
the alternate Nine Year Wilkinson Microwave Anisotropy Probe (WMAP9) model
is not perfect either and still overpredicts
the number of faint clusters. Basic
attempts to assess the cosmological
leverage of the sample revealed that, even
with the current subsample, the error
budget is already dominated by uncertainties in the cluster mass calibration, which
therefore will be one of the priorities of
the XXL cluster science programme over
the next few years. In particular, we are
starting a collaboration with the Subaru
Hyper-SuprimeCam team in order to
obtain reliable lensing masses for most of
the XXL clusters.
Cluster baryon content
Beside counts and topology, another
potential cosmological probe is the
baryon fraction (fgas) in clusters (XXL XIII).
This test assumes that fgas is constant
as a function of redshift for the high-mass
clusters, because they are supposed
to have retained all the material accreted
since their formation epoch. Gas mass
and total mass measurements do not
scale in the same way as a function
of cosmology; therefore, requiring that
their ratio is constant ought to put con-
23:50
23:45
23:40
23:35
23:30
23:25
23:20
23:15
23:10
53:00
Declination ()
Cosmological constraints
Right Ascension (h,m)

23:55
52:00
54:00
55:00
56:00
7.978 5.176
Figure 2. Upper: Spectroscopically

confirmed AGNs (blue) and cluster
g alaxies (red) in the southern XXL field
from the AAOmega follow-up (XXL
Paper XIV). Left: Corresponding X-ray
mosaic (units of colour bar are counts
s 1 degree 1). The 25-square-degree
field is covered by some 220 XMM
pointings. The XMM field of view has
a size comparable to that of the Moon
(30 arcminutes diameter, shown at
the upper left). More than 12 000 AGNs
have been detected in this image.
The red circles show the clusters of
galaxies.
XMM-XXL South
0.59 0.88 1.40 2.00 2.90 3.90 5.20 6.60 8.20
straints on the cosmological parameters.

We measured the number of baryons
(X-ray gas and stars) in the XXL-100-GC
sample. Halo masses were estimated
from a masstemperature relation that
was directly calibrated using weak-lensing
measurements (from the Canada France
Hawaii Telescope Legacy Survey, CFHTLS)
for a subset of XXL clusters. We find that
the weak-lensing calibrated gas fraction
of XXL-100-GC clusters is lower than was
found in previous studies using hydrostatic masses.
We compared our fgasM relation with the
predictions of cosmological simulations
using different gas physics. It was found
that our results favour extreme AGN feedback schemes in which a large fraction
of the baryons are expelled from the
potential well of dark matter halos. Such
models are, however, in tension with
X-ray-only proxies, such as the gas density and entropy profiles (Le Brun et al.,
2014) and are not able to reproduce the
relation between gas mass and temperature of XXL clusters.
Therefore, the results presented are
challenging for current numerical simu
lations, and reconciling the observed
gas fraction with the predictions would
require that our weak-lensing masses be
systematically overestimated. A mass
29

Figure 3. Upper: Distribution of the bright XXL

cluster sample in the
M500 vs. redshift plane.
The dashed line shows
the 50 % completeness
limit in the WMAP9 cosmology. Middle: Redshift distribution of the
XXL-100-GC sample
(filled histogram) compared with different
model expectations
(CMB measurements
from WMAP + H0 + BAO
presented in Hinshaw
et al. [2013], and latest
CMB measurements
from the Planck satellite,
but without additional
constraints from Planck
Collaboration 2015).
Lower: Differential luminosity function of the
bright XXL cluster sample in three redshift
bins. The lower plot
shows the residuals with
respect to the low-
redshift WMAP9 prediction (XXL Paper II).
,,3,

2ODBSQNRBNOHBQDCRGHES
/GNSNLDSQHBQDCRGHES
CDSDBSHNMOQNA@AHKHSX

1DCRGHES

6, /lXD@Q
/K@MBJ@KNMD
77+ll&"

"KTRSDQRODQAHM

1DRHCT@KR
CMC+:G ,OB l DQFR l l<
l
l
l
l
YHM l
YHM l
YHM l
6, /DWODBS@SHNMR
l
l

l

30

+ 77+
:G DQFR <

bias 1 b = 0.58 0.04, which is required

to reconcile Planck cluster counts with
the primary CMB, would further exacerbate the tension between the baryon
fraction and the cosmological value, challenging our understanding of cluster
physics. Therefore, a satisfactory solution
to the tension between CMB and cluster
counts must also simultaneously explain
the low baryon fraction measured for the
XXL halos.
Discovery of five superclusters
Given its depth and sky coverage, XXL
is particularly suited to systematically
unveiling the clustering of X-ray clusters
and to identifying superstructures in a
homogeneous X-ray sample down to the
typical mass scale of a local massive
cluster. Although a few isolated, very high
redshift superclusters are known, our
work is the first attempt to systematically
unveil superstructures up to z ~0.5 in a
homogeneous X-ray sample. In this
framework, we discovered five low-mass
superclusters consisting of three to seven
galaxy structures.
One of them is particularly interesting and
is presented in more detail (XXL VII and
Figure4). It is composed of six clusters of
galaxies and the structure is very compact with all the clusters residing in one
XMM pointing. Our subsequent spectroscopic follow-up with WHT and NTT
provided a median redshift of z ~ 0.43.
An estimate of the X-ray mass and luminosity of this supercluster returns values
of 1.7 1015 MA and of 1.68 1044 erg
s1, respectively, and a total gas mass of
Mgas = 9.3 1013 MA . Instead of having
a massive central cluster with infalling
filaments and smaller structures, it has
almost two equal-sized objects, making
it qualitatively different from the network
around an already formed massive cluster. This supercluster is currently the most
massive and most distant found in XXL.
If we put together the relatively small
crossing time, the common X-ray emission
of three members, and the measured
gas fraction and mass, we can speculate
that we are observing an unrelaxed structure with an ongoing merging between
at least three of the member clusters. If
nothing else intervened to alter the system, and assuming that the estimated
@QBLHM
#DBKHM@SHNM
l l l
JOBY
7+22"
7+22"
7+22"
7+22"
7+22"

l
7+22"
+HMDlNElRHFGSCHQDBSHNM,OB
l

7+22"
7+22"
7+22"

l
l
7+22"
7+22"
l
7+22"
# D
BK
H M
@S
HN
M
C H l

QD

B S l
B

HN
l
M , O
M l l l
HQ D B S HN
C
,
M
R HN
OB
RBDM

1 HF G S
1HFGS RBDMRHNM
Figure 4. Top left: CFHTLS MegaCam mosaic image

in the i-band with the XXL contours superimposed
in green; X-ray clusters are red circles. The positions
of the BCG in each cluster are shown and the black
crosses highlight the point sources excluded from the
X-ray analysis. Top right: Three-dimensional configuration of the z = 0.43 supercluster (XXL Paper VII).
Right: CFHTLS-wide colour u*r:z: image (8 by 8 arcminutes) of XLSS-C 82-83-84 inner complex.
crossing time is a good estimate of the

merging time, it is likely that by the present time the supercluster would already
have completely merged ~2 Gyr before
and resemble the most massive known
clusters.
First spectral characterisation of diffuse
light in a cluster
Within a cluster, gravitational effects can
lead to the removal of stars from their
parent galaxies and subsequent dispersal into the intracluster medium (ICL).
Gas hydrodynamical effects can additionally strip gas and dust from galaxies. The
properties of the ICL can therefore help
constrain the physical processes at work
in clusters by serving as a fossil record
of the interaction history. We detected in
the XXL survey a very peculiar intermediate-mass cluster: XLSSC-116, with a very
high level of such diffuse light, by applying a wavelet-based method to CFHT
Megacam and Wide-field InfraRed Camera (WIRCAM) images (XXL VIII). The
amount of diffuse light is equivalent to two
brightest cluster galaxies (BCGs) in the
i:-band. To the best of our knowledge,
this is the first detected cluster with such
a large amount of diffuse light.
The source of the diffuse light was then

spectroscopically characterised with
ESO Multi-Unit Spectroscopic Explorer
(MUSE) spectroscopic data (proposal
60.A-9302). MUSE data were also used
to compute redshifts of 24 cluster galaxies and search for cluster substructures.
The cluster consists of a main component with a velocity dispersion of the
order of 600 km s1 and an infalling lowmass group with a velocity dispersion

of 170km s1. Part of the detected diffuse
light has a very weak optical stellar component and apparently consists mainly
of gas emission, while other diffuse light
sources are clearly dominated by old
stars. Furthermore, emission lines were
detected in several places coinciding with
31

77+ &".2,.2
$&(2
+'
"#%2
"#%-
7,2
XXL-N
XXL-S
COSMOS
ATLAS
XMM-CDFS

10 3

10 2
N (> S)/deg 2
HS@
D1
NR

10 0
+

101
77
+ JD5 DQFR

Figure 5. Left: Luminosityredshift plane for different X-ray fields in the

210 keV energy band.
Right: log Nlog S for
the XXL-1000-AGN in
the XXL-N field (blue circles) and in the XXL-S
field (orange squares) as
compared to other literature surveys (XXL
Paper VI).
104
10 1

l

l

1DCRGHES
the diffuse light. Our spectral analysis

shows that this emission likely originates
from gas with a low-excitation parameter.
Globally, the stellar contribution to the
ICL is about 2.3 109 years old even
though the ICL is not currently forming a
large number of stars. On the other hand,
the contribution of the gas emission to
the ICL in the optical is much greater than
the stellar contribution in some regions,
but the gas density is likely too low to
form stars. These observations favour
ram-pressure stripping, turbulent viscous
stripping, or supernovae winds as the
origin of the large amount of intracluster
light. Since the cluster appears not to
be in a major merging phase, we conclude that ram-pressure stripping is the
most plausible process that generates
the observed ICL sources.
AGNs in the XXL survey
X-ray extragalactic surveys are also ideal
laboratories for the study of the evolution
and clustering of AGN. Optimally, a combination of deep and wide surveys is
necessary to create a complete picture of
the population. Deep X-ray surveys provide the faint population at high redshift,
while wide surveys provide the rare bright
sources and the large-scale structure
information. Nevertheless, very wide area
surveys often lack the ancillary information available for modern deep surveys.
32
10 2
10 16
10 15
10 14
10 13
S210keV /erg s 1 cm 2
The XXL survey occupies the parameter

space that lies between deep surveys
and very wide area surveys, constituting
a stepping stone between current deep
fields and planned wide area surveys
(Figure 5).
XXL-1000-AGN is our first point source
catalogue. The sample is selected in the
210 keV energy band with the goal of
providing an initial sizable sample useful
for AGN studies (XXL Papers VI, XI, XII).
The limiting flux is F210 keV = 4.8 10 14 erg
s1 cm2. We use both public and proprietary datasets in the optical and nearinfrared to identify the counterparts of the
X-ray point-like sources by means of a
likelihood ratio test, reaching 97.2 % identification rate. Using multi-wavelength
data from the far ultra-violet to the mid
infrared, we derive the median spectral
energy distribution of AGN as a function
of X-ray absorption. We improve upon
the photometric redshift determination
for AGN by applying a Random Forest
classification trained to identify, for each
object, the optimal photometric redshift
category (passive, star-forming, starburst,
AGN, quasar). In addition, we assign a
probability to each source that indicates
whether it might be a star or an outlier.
The photometric redshift accuracy is
0.095 for the full XMM-XXL with 28 %
catastrophic outliers estimated on a sample of 339 sources. We apply Bayesian
analysis to model the X-ray spectra
assuming a power-law model with the
10 12
10 11
presence of an absorbing medium, finding no trend of photon index or hydrogen

column density with redshift and a fraction of 26 % absorbed sources, consistent with the literature on bright sources.
The XXL-1000-AGN sample number
counts extend the COSMOS log Nlog S
to higher fluxes (Figure 5); they are fully
consistent with the Euclidean expectation
and agree with previous deep (Chandra
Deep Field South [CDFS] and COSMOS)
and wide (Herschel-ATLAS) XMM-Newton
surveys. We constrain the intrinsic luminosity function of AGN in the 210keV
energy band where the unabsorbed X-ray
flux is estimated from the X-ray spectral
fit up to z = 3. An application of the
friends-of-friends algorithm, at 10 h1
Mpc and 20 h1 Mpc percolation radii,
shows significant structures with 23
members and identifies a superclustersized structure at redshift 0.14.
In future publications we will expand the
analysis to the full XXL catalogue, containing an unprecedented number of
~104 (X-ray) point-like 210 keV detected
sources, and analysing the fully combined XMM pointings to reach maximum
depth. A unique advantage of the XXL
large statistical samples is the combined
study of X-ray point-like sources and
X-ray detected galaxy clusters, which will
allow the study of AGN with respect to
their environment and, vice versa, the
study of the impact of AGN on clusters.
1
Arcminutes
Arcminutes
Figure 6. Left: Simulated X-ray emissivity map

extracted from a 25 square degree by z = 03 light
cone (OverWhelmingly Large Simulations [OWLS],
Le Brun et al., 2014). Right: Same portion of the virtual sky transformed into an XMM image by the XXL
simulator. All XMM characteristics are taken into
account: point spread function (and its variation with
energy), quantum efficiency, various background
components and vignetting.
References
0
1
Arcminutes
Prospects
The second phase of the XXL project is
expected to end in 2018, with the publication of the complete cluster catalogue,
including all data available from the
multi-wavelength follow-up along with the
selection functions. These will incorporate a thorough analysis of the impact of
cluster shapes and AGN activity on the
detection rates, which we will measure
by means of N-body simulations performed under various physical hypotheses (Figure 6).
ESO/H. Dahle
The final cosmological analysis of the full

XXL cluster sample will involve a completely self-consistent treatment of the
cluster evolution, selection function and
cosmology. With some five times as
many clusters (already some 450 clusters
0
1
Arcminutes
have been identified), we expect to provide a useful and stand-alone set of

cosmological constraints. In addition to
the traditional study of the dn/dz and dn/
dM/dz counts, we have developed a
method based on the modelling of the
X-ray observable parameter space, that
globally by-passes the intermediate
phase of the (cosmologically dependent)
cluster mass determination (Clerc et al.,
2012; Pierre et al., 2016). We will also
release the XXL multi-wavelength AGN
catalogue, that is some 22 000 objects.
The dataset, including spectroscopic
follow-up, has a unique legacy value
for cosmological and extragalactic environmental studies and will constitute
a calibration resource for future surveys,
like eRosita (all-sky, but at significantly
lower sensitivity and resolution) and
Euclid.
Clerc, N. et al. 2012, MNRAS, 423, 3561

Clerc, N. et al. 2012, MNRAS, 423, 3545
Driver, S. P. et al. 2016, MNRAS, 455, 3911
Hinshaw, G. et al. 2013, ApJS, 208, 19
Le Brun, V. et al. 2014, MNRAS, 441, 1270
Le Fvre, O. et al. 2005, A&A, 439, 845
Guzzo, L. et al. 2014, A&A, 566, 108
Hinshaw, G. et al. 2013, ApJS, 208, 19
XXL Survey: I. Pierre, M. et al. 2016, A&A, 592, A1
XXL Survey: II. Pacaud, F. et al. 2016, A&A, 592, A2
XXL Survey: III. Giles, P. A. et al. 2016, A&A, 592, A3
XXL Survey: IV. Lieu, M. et al. 2016, A&A, 592, A4
XXL Survey: V. Mantz, A. B. et al. 2016, ApJ,
794, 157
XXL Survey: VI. Fotopoulou, S. et al. 2016, A&A,
592, A5
XXL Survey: VII. Pompei, E. et al. 2016, A&A, 592, A6
XXL Survey: VIII. Adami, C. et al. 2016, A&A, 592, A7
XXL Survey: IX. Baran, N. et al. 2016, A&A, 592, A8
XXL Survey: X. Ziparo, F. et al. 2016, A&A, 592, A9
XXL Survey: XI. Smoli, V. et al. 2016, A&A, 592, A10
XXL Survey: XII. Koulouridis, E. et al. 2016, A&A,
592, A11
XXL Survey. XIII. Eckert, D. et al. 2016, A&A, 592, A12
XXL Survey XIV. Lidman, C. et al. 2016, PASA, 33, 1
Night-time observations with the New Technology

Telescope at the La Silla Observatory.
Following two pages: VLT Survey Telescope (VST)

composite u-, g- and i-band, large-field image
(120 by 84 arcminutes) of the core of the nearby
Fornax Galaxy Cluster. At a distance of about
19Mpc, the Fornax cluster is a rich cluster (class 0)
with more than 340 known members. The central
brightest cluster galaxy, NGC 1399 can be seen to
the north east, with the prominent barred spiral
NGC1365 to south west. See Release eso1612 for
more details.
33
ESO. Acknowledgement: Aniello Grado and Luca Limatola
The LEGA-C Survey: The Physics of Galaxies 7 Gyr Ago
Arjen van der Wel 1

Kai Noeske 1
Rachel Bezanson 2
Camilla Pacifici 3
Anna Gallazzi 4
Marijn Franx 5
Juan-Carlos Muoz-Mateos 6
Eric F. Bell 7
Gabriel Brammer 8
Stephane Charlot 9
Priscilla Chauk 1
Ivo Labb 5
Michael V. Maseda 5
Adam Muzzin 10
Hans-Walter Rix 1
David Sobral 11, 5
Jesse van de Sande 12
Pieter G. van Dokkum 13
Vivienne Wild 14
Chris Wolf 15
1
Max-Planck-Institut fr Astronomie,
Heidelberg, Germany
2
Steward Observatory, University of
Arizona, Tucson, USA
3
Goddard Space Flight Center,
Greenbelt, USA
4
INAFOsservatorio Astrofisico di Arcetri,
Firenze, Italy
5
Leiden Observatory, Leiden University,
the Netherlands
6
ESO
7
Department of Astronomy, University
of Michigan, Ann Arbor, USA
8
Space Telescope Science Institute,
Baltimore, USA
9
Institut dAstrophysique de Paris,
France
10
Institute of Astronomy, University of
Cambridge, UK
11
Department of Physics, Lancaster
University, UK
12
Sydney Institute for Astronomy, University of Sydney, Australia
13
Department of Astronomy, Yale University, New Haven, USA
14
School of Physics and Astronomy,
University of St. Andrews, UK
15
Research School of Astronomy and
Astrophysics, Australian National
University, Canberra, Australia
The LEGA-C (Large Early Galaxy Census) survey is made possible by the
refurbishment of the Very Large Telescope VIsible and Multi Object Spectrograph (VIMOS) instrument and the
36
implementation by ESO of a new gen

eration of large spectroscopic surveys.
The goal is to obtain high-quality
continuum spectra of thousands of
galaxies with redshifts up to z = 1, with
which key physical parameters that
were previously inaccessible can be
measured. These include star formation
histories and dynamical masses,
which greatly improve our insight into
how galaxies form and evolve. This
article coincides with the first public
data release of fully reduced and
calibrated spectra.
Our knowledge of stellar populations
tells us about the formation and evolution
of galaxies. High-quality (continuum)
spectroscopy of galaxies reveals the stellar absorption features that trace star
formation histories and chemical content.
Such data have been available for galaxies in the present-day Universe for some
decades and have brought into clear focus
the multi-variate correlations between
stellar population properties and mass,
structure, size, stellar velocity dispersion,
nuclear activity and environment. This
information has greatly illuminated the
processes that drive star formation and
the ongoing assembly of present-day
galaxies.
The main limitation when examining present-day galaxies for the purpose of
reconstructing their formation history is,
however, that most of the star formation
occurred in the distant past: the mean
stellar ages of typical galaxies are typically well over 5 Gyr (Gallazi et al., 2005)
and it is difficult to resolve star formation
histories from spectra: it is nearly impossible to distinguish stellar populations
with ages of, for example, 5, 7 and 9 Gyr
from integrated spectra. For this reason
the community has put much effort into
lookback studies, aimed at directly ob
serving galaxies at earlier cosmic times.
On the one hand, redshift surveys, such
as the VIMOS Very Deep Survey (VVDS;
Le Fvre et al., 2005) and the COSMOS
spectroscopic survey (zCOSMOS; Lilly et
al., 2008) have created large samples of
galaxies with spectroscopic redshifts,
tracing the evolution of the number density and the luminosity function of galaxies. On the other hand, photometric sur-
veys (for example, UltraVISTA; McCracken

et al., 2013) have collected multi-wavelength datasets used to derive photo
metric redshifts, which have gradually
improved to the point that spectroscopic
redshift surveys are no longer needed for
the purpose of quantifying galaxy evolution (with the exception of the effect of
environment). In addition, the photometric
surveys provide estimates of integrated
galaxy properties such as stellar mass,
star formation rate and restframe colours.
Adding Hubble Space Telescope (HST)
imaging to the mix enables us to reveal
the internal structure of distant galaxies
(for example, van der Wel et al., 2014a).
The results of these very significant
efforts is that we now understand that
the galaxy population at large lookback
times is in many ways similar to the present-day galaxy population: mass, structure and star formation activity are correlated in the same manner (for example,
Franx et al., 2008). At the same time,
there are many differences: at higher redshift, star formation rates were much
higher (Madau et al., 1996), morphologies
less regular (for example, Conselice et al.,
2008), and galaxies of the same stellar
mass are smaller in size (for example, van
der Wel et al., 2014b).
The main limitation of the lookback studies is the challenge of connecting progenitors and descendants: despite our
exquisite knowledge of the evolution of
the population of galaxies as an ensemble, the evolutionary history of individual
galaxies has remained hidden from view.
In our Universe, with hierarchical structure growth that is largely stochastic in
nature, we should expect that galaxies
that are similar today were probably very
different in the past, and, analogously,
galaxies that were similar in the past will
be very different today.
To summarise, our current insight into
galaxy evolution is limited by two factors:
1) the archaeological approach of obtaining spectroscopy of present-day galaxies
lacks the power to reveal the bulk of star
formation history, because galaxies are
too old; 2) the lookback approach only
reveals the evolution of the population, not
of individual galaxies. The solution is both
obvious and very challenging: to obtain
high-quality spectroscopy of galaxies at
0T@KHSXNERODBSQ@

/@QDMSR@LOKD
%TKKRTQUDX

RSXD@Q#1

*A@MCKHLHS

1DCRGHES
1DRSEQ@LD45

,

BG@Q L@RR

2SDKK@QL@RR
Figure 1. Rest-frame UV colour vs. stellar mass of

the K-band selected primary galaxy sample of the
LEGA-C survey at redshift 0.6 < z < 1.0. Light grey
points refer to the full UltraVISTA sample; black
points: primary galaxy candidates; and red points:
primary galaxies included in the LEGA-C survey.
large lookback times. This is the aim of the

LEGA-C survey (van der Wel et al., 2016).
The LEGA-C survey
The challenge lies in obtaining continuum
spectra of sufficient resolution and depth
for faint targets. The design of the LEGA-C
survey is constrained by several practical
factors:
1) the resolution should be at least
R~2000 to distinguish individual features and constrain the kinematic
properties of the targets;
2) t he signal-to-noise ratio (S/N) should
be at least 10 per resolution element,
and preferably ~ 20;
3) the maximum redshift is z ~ 1, otherwise targets become too faint and the
diagnostic features shift into the nearinfrared, where ground-based nearinfrared spectrographs are still a factor
~ 100 slower in survey speed for the
purpose of continuum spectroscopy;
4) the sample size should be in the
1000s, otherwise the population is
either undersampled or biased toward

2-
particular types of galaxies, either of

which would preclude the general
goals of constraining the evolutionary
history of galaxies in general.
Given these constraints, we started the
LEGA-C survey in December 2014 with
VIMOS in multi-object (MOS) mode.
The survey is led and coordinated from
the Max Planck Institute for Astronomy
in Heidelberg, Germany, and has coinvestigators across Europe and indeed
the globe.
Upon completion, likely in 2018, we will
have collected more than 3000 spectra of
galaxies in the redshift range 0.6 < z < 1.0
in the COSMOS field, at Right Ascension
10 hr and Declination + 2. The sample is
selected based on K-band magnitude
in order to approximate a selection by
stellar mass and avoid biases due to
extinction from the publicly available
UltraVISTA catalogue by Muzzin et al.
(2013). The selection is independent of
any other parameter and the sample,
shown in Figure 1, therefore spans the
full range of galaxy properties in terms
of morphology, star formation activity
and dust attenuation across the galaxy
population with stellar mass in excess of
about 1010 MA . This primary sample is
complemented by ~900 fillers lowermass galaxies and higher-redshift galaxies. With the high-resolution red grating
Figure 2. Redshift and signal-to-noise (S/N) distribution of the L

EGA-C primary sample. The parent
s ample from which targets are selected is shown in
black; the survey design includes objects shown in
red; while the observed sample is shown in yellow/
black.
(HR_red) and integration times of 20

hours per target, the required resolution
and S/N can be achieved. The typical
wavelength range of 63008800 samples essential features such as the
Balmer/4000 break, all Balmer lines
except H, the G-band, and multiple Fe,
Ca and Mg features.
VIMOS slits are assigned to the primary
targets first, prioritised by apparent
K-band magnitude as far as slit collisions
permit. Then one or more blue stars for
telluric absorption correction and several
alignment stars are included in the slit
mask design. The remaining slit real
estate is used for fillers: higher-redshift
objects with bright K-band magnitudes,
fainter targets in the 0.6 < z < 1.0 redshift
range, and other fainter sources, respectively. With slit lengths of ~ 10 arcseconds, VIMOS can simultaneously observe
~ 130 objects, bringing the on-sky survey
execution time to 640 hours. In Visitor
Mode 128 nights were allocated to achieve
this goal, spread out over 200 actual
nights (due to fractional-night scheduling). This allocation makes LEGA-C the
37
Figure 3. 1D extracted
restframe spectra of
654 p
rimary-sample galaxies observed in the
first year of LEGA-C.
Each row shows one
spectrum, where the
galaxies are sorted from
high star formation
activity (at the top) to
low star formation activity (at the bottom).
van der Wel A. et al., The LEGA-C Survey
:.((<
'DK ' '
* '

most expensive extragalactic spectroscopic survey to date on an 8-metreclass telescope. The redshift and ex
pected S/N distribution of the resulting
primary galaxy sample is illustrated in
Figure 2.
The spectra
The data are reduced using a combination of the ESO pipeline and our own
pipeline based on customised algorithms
for sky subtraction, object extraction
and co-addition. In Figure 3 we show
extracted 1D spectra of 654 primary targets observed in the first year of obser
vations. The galaxies are sorted by their
specific star formation rate (star formation
rate per unit stellar mass), with the most
actively star-forming galaxies at the top.
The high star formation rate galaxies
show nebular emission lines, Balmer lines
in absorption and emission, but also
metal lines. The more passive systems
show stronger metal features and across
this sample up to 50 unique absorption
features are readily visible, illustrating the
superb depth of the spectra. Kinematic
information is revealed thanks to the high
spectral resolution: Ca and Fe features
appear more Doppler broadened for the
spectra near the bottom, as those galaxies are more massive.
38
'b
'a
"-
"@

&
'` :.(((< :.(((<
%D
"@
%D
%D

1DRSEQ@LDV@UDKDMFSGML
Figures 4 and 5 show typical examples

of 1D spectra in more detail. In Figure4
we show ten galaxies ordered by their
basic morphology as traced by the Srsic
index: galaxies with high Srsic indices,
that is, more concentrated light profiles
usually associated with early-type morphologies, are at the top; galaxies with
low Srsic indices, that is, disc-like morphologies, are at the bottom. The correspondence with morphology is clearly
seen in the HST image cut-outs from the
COSMOS survey (Scoville et al., 2007).
LEGA-C reveals that for the first time,
beyond the present-day Universe, a clear
correlation exists between morphology
and spectral properties: early-type galaxy
spectra are characterised by strong metal
absorption line features and a lack of
strong Balmer absorption and nebular
emission lines; while for late-type galaxies this is reversed.
In Figure 5 we show five late-type gal
axies ordered by inclination. Absorption
lines are detected regardless of viewing
angle. The example spectra illustrate
a unique aspect of the LEGA-C survey:
even the dustiest and the bluest galaxies
show absorption line features with high
fidelity, despite the challenges presented
by the low surface brightness and bright,
blue stellar populations with strong continuum features.
%D

,F
%D

Towards understanding the physics of

galaxy formation
The high-quality spectra shown in Figures 35 allow us to measure with good
accuracy and precision a large range of
physical parameters that were previously
inaccessible for galaxies at large lookback times. These parameters fall into
two broad categories: stellar populations
and kinematics. The basic stellar population measurements are mean stellar age
and metallicity, and eventually a more
general description of the star formation
history. Now that LEGA-C has overcome
the challenge of creating spectra of galaxies at significant lookback times, we
actually have a clear advantage over present-day galaxy studies when it comes
to reconstructing the full star formation
history. At the redshifts studied, stars will
typically be younger than 3 Gyr in essentially all LEGA-C galaxies, which is an age
range over which different generations
of stars can be resolved in age. This will
inform us directly what the stellar masses
of progenitors at z > 1 must be. Our
knowledge of those potential progenitors
is very substantial indeed, such that we
can, for the first time, follow the evolution
of individual galaxies across a significant
part of cosmic time, right into the epoch
when star formation activity was highest
in a cosmically average sense.
30
20
10
Luminosity density (L/10 9 L)
15
CaK CaH
Ca
Fe
Ca
Fe
Fe
Mg
ID : 211409 ; z = 0.714 ; M = 1.3 10 11 M ; SFR = 5.9 M/yr

*
n = 4.4
ID : 210210 ; z = 0.654 ; M = 2.4 10 11 M ; SFR = 3.2 M/yr

*
n = 3.7
ID : 133163 ; z = 0.697 ; M = 3.7 10 11 M ; SFR = 0.6 M/yr

*
n = 3.4
ID : 135418 ; z = 0.740 ; M = 1.1 10 11 M ; SFR = 18 M/yr

*
n = 2.9
10
5
20
15
10
5
6
4
2
15
10
5
4000
8
6
4
2
n = 7.7
ID : 206573 ; z = 0.702 ; M = 2.5 1011 M ; SFR = 3.6 M/yr

*
10
4500
Restframe wavelength ()
Figure 4. Spectra of ten

typical LEGA-C targets
ordered by morphology
(using the Srsic index).
Flux density uncertainties are indicated in grey.
HST image cut-outs
are shown on the right.
Labels (left) record
object ID, redshift, stellar
mass and star formation
rate; with Srsic index
at right.
5000
ID: 209487 ; z = 0.656 ; M = 2.2 10 10 M ; SFR = 11 M/yr

*
n = 2.1
ID: 210003 ; z = 0.739 ; M = 3.9 10 10 M ; SFR = 10 M/yr

*
n = 2.0
ID: 131951 ; z = 0.687 ; M = 1.6 1011 M ; SFR = 24 M/yr

*
n = 1.7
ID: 206763 ; z = 0.688 ; M = 6.0 1010 M ; SFR = 25 M/yr

* H
H H
n = 1.0
20
10
15
10
5
10
5
ID: 128528 ; z = 0.658 ; M = 1.3 1011 M ; SFR = 11 M/yr

*
[O II]
[Ne III]
n = 1.0
[O III] [O III]
4000
4500
The primary kinematic measurement is

the stellar velocity dispersion. Collisionless star particles are excellent tracers of
the gravitational potential and thus the
total mass, which opens new avenues of
exploration. A practical application is to
test whether the much-maligned stellar
mass estimates are robust at large lookback times for all galaxy types. At these
times velocity dispersions and central
mass density are thought to be key fac-
5000
tors in determining the star formation

activity of a galaxy. Now that we have
direct measurements for distant galaxies
that support this notion (see Figure 6)
we can ask the question of how some
galaxies cease to form stars and evolve
sedately afterward. Furthermore, the
evolution of velocity dispersions in combination with other properties (for example, size) informs us about the rate of
growth of passive galaxies through merg-
ing. A secondary kinematic measurement

is the gas velocity dispersion. Gas motions
are affected by numerous other forces
besides gravity and, while gas kinematics
is widely used as a tracer of mass, it
remains to be tested how large are the
uncertainties. A comparison between gas
and stellar velocity dispersion now allows
us to make this test.
39
20
15
10
5
6
4
2
6
4
2
10
van der Wel A. et al., The LEGA-C Survey
ID : 206894 ; z = 0.914 ; M = 5.8 1010 M ; SFR = 45 M/yr

*
b/a = 0.88
ID : 130902 ; z = 0.759 ; M = 1.9 10 10 M ; SFR = 12 M/yr

*
b/a = 0.64
ID : 132535 ; z = 0.758 ; M = 5.9 1010 M ; SFR = 32 M/yr

*
b/a = 0.47
ID : 210392 ; z = 0.679 ; M = 6.6 1010 M ; SFR = 10 M/yr

*
b/a = 0.41
ID : 140059 ; z = 0.680 ; M = 4.8 10 10 M ; SFR = 8.5 M/yr

*
b/a = 0.17
5
4
3
2
1
4000
4500
2SDKK@QENQL@SHNMQ@SD,XQ l

5000
Figure 6. Star formation rate

vs. stellar mass for the primary
LEGA-C sample observed in
the first year. The colour coding
represents the stellar velocity
dispersion.
2SDKK@QUDK CHRO JLR l

be completed by 2018 and, excitingly,

the data quality is precisely as good as
anticipated, confirming the competitiveness of VIMOS.
A second data release will follow by the
end of 2016. This will double the sample
and expand to higher-level data products
in the form of derived physical parameters, such as velocity dispersions and
line indices. There will be updated data
releases in subsequent years. For more
technical information on the sample
selection, observations, data reduction
and modelling techniques, please refer
to the LEGA-C survey paper (van der Wel
et al., 2016).

2SDKK@QL@RR,
Finally, despite the fact that the LEGA-C

spectra are seeing limited, we have spatial
information on both the stellar population
characteristics and the kinematics. While
the primary goal of the LEGA-C survey is
to obtain integrated spectral properties,
exploring this aspect will undoubtedly
prove to be highly interesting. Stellar rotation curves and age gradients provide
strong constraints on the physics of galaxy formation and the connection to the
dark matter halo hosts.
40
Figure 5. Spectra of five

LEGA-C disc galaxies
ordered by inclination
(axis ratio b/a). Details
as Figure 4.
References
Timeline
At this point we have collected approximately 45 % of the dataset based on the
first two years of observation. This article
coincides with the first data release1 that
includes fully calibrated 1D and 2D spectra of the data taken in the first year: 925
galaxies (22 % of the full survey). Figure2
shows the redshift and S/N distribution
of the data obtained in the first year: as
can be seen, LEGA-C is on schedule to
Conselice, C. et al. 2011, MNRAS, 417, 2770

Franx, M. et al. 2008, ApJ, 688, 770
Gallazzi, A. et al. 2005, MNRAS, 362, 41
Le Fvre, O. et al. 2005, A&A, 439, 845
Lilly, S. et al. 2007, ApJS, 172, 70
Madau, P. et al. 1996, MNRAS, 283, 1388
McCracken, H. J. et al. 2012, A&A, 544, 156
Scoville, N. J. et al. 2007, ApJS, 172, 1
van der Wel, A. et al. 2014a, ApJ, 792, 6L
van der Wel, A. et al. 2014b, ApJ, 788, 28
van der Wel, A. et al. 2016, ApJS, 223, 29
Links
1
ccess to Phase 3 LEGA-C data: http://archive.

A
eso.org/wdb/wdb/adp/phase3_spectral/form
ALMACAL: Exploiting ALMA Calibrator Scans to Carry

Out a Deep and Wide (Sub)millimetre Survey, Free of
Cosmic Variance
Ivn Oteo 1, 2
Martin Zwaan 2
Rob Ivison 2,1
Ian Smail 3, 4
Andrew Biggs 2
1
Institute for Astronomy, University of

Edinburgh, United Kingdom
2
ESO
3
Centre for Extragalactic Astronomy,
Department of Physics, Durham University, United Kingdom
4
Institute for Computational Cosmology,
Department of Physics, Durham University, United Kingdom
We present the latest results from
ALMACAL, a novel, wide and deep
(sub-)millimetre survey that exploits
ALMA calibration data that comes for
free during science observations. Combining compatible data acquired during
multiple visits to many ALMA calibrators, sufficiently low noise levels can
be reached to detect faint dusty starforming galaxies (DSFGs). As of April
2016, we have analysed data for more
than 240 calibrators, reaching noise
levels as low as ~ 10 Jy beam1, at
sub-arcsecond spatial resolution. We
have found 15 DSFGs, some less luminous than the sources detected by
the deepest far-infrared surveys. Future
analyses will deliver larger samples,
free of cosmic variance, with redshifts
determined via detection of multiple
(sub-)millimetre lines, and dust emission imaged at milliarcsecond spatial
resolution. The combination of area
and depth reached by ALMACAL is
unlikely to be surpassed by any other
ALMA (sub-)millimetre survey.
A fraction of the restframe ultra-violet
light emitted by star-forming galaxies
(SFGs) is internally absorbed by dust and
re-emitted at far-infrared wavelengths
(Lutz et al., 2014). Detecting dust emission using far-infrared surveys is then
essential to determine accurate star formation rates (SFR) and levels of dust
attenuation for SFGs and, therefore, to
understand the star formation history of
the Universe. Indeed, this has been one
of the main goals of many surveys carried
out with past and current far-infrared/
submillimetre facilities (for example,

Spitzer, Herschel, James Clark Maxwell
Telescope or the Atacama Pathfinder
EXplorer). However, due to their limited
sensitivity, the detection of dust emission
has been restricted to the most extreme
sources at each r edshift, hindering
broader studies of galaxy formation and
evolution.
Currently, the Atacama Large Millimeter/
submillimeter Array (ALMA) is the only
facility able to surpass these limits and
select galaxies more than ten times
fainter, which are responsible for much
of the SFR density at redshift z > 1 and
the main contributors to the extragalactic
background light the integrated unresolved emission from extragalactic
sources. Furthermore, this faint far-infrared
population might be the link between
the far-infrared-bright galaxies extensively
studied in the past submillimetre galaxies or SMGs and the less extreme
SFGs selected in optical/near-infrared
surveys (and typically represented by Ly
emitters or Lyman-break galaxies) for
which far-infrared detections are found
for only ~ 5 % of the total population
(Oteo et al., 2013).
ALMACAL: A wide and deep (sub-)millimetre survey
Using ALMA calibration observations,
we are carrying out a novel, wide and
deep (sub-)millimetre survey: ALMACAL
(Oteo et al., 2016). Every ALMA science
project includes calibration observations
of very bright, compact sources to set
the flux density scale, to measure the
bandpass response, and to calibrate
the amplitude and phase of the visibilities
of the science targets (Fomalont et al.,
2014). Observations of such calibrators
are essential and represent a significant
fraction of each observing block (OB).
Each calibrator will typically be observed
several times, often many times, on different dates, in several different ALMA
bands, as part of one or several OBs
corresponding to one or several ALMA
science projects. Therefore, by com
bining compatible data for a high number
of calibrators, it is possible to reach
sensitivities and areas that allow the
detection of dusty star-forming galaxies.
There are several, key advantages of

using ALMA calibrators to search for and
analyse high-redshift DSFGs. First, the
sensitivities that can be reached after
combining compatible data for different
calibrators allow us to detect DSFGs
fainter than those detected previously
with the ESA Herschel satellite or other
submillimetre, ground-based single dish
observations. This opens a window to
study dust emission in normal SFGs, with
the data coming for free. Perhaps more
importantly, a number of calibrators will
be observed at extremely high spatial
resolution, if this is amongst the requirements of the science project within
which they are observed. The simultaneous presence in the map of one or more
DSFGs and a bright ALMA calibrator
lends itself perfectly to self-calibration (for
example, Pearson & Readhead, 1984),
which permits near-perfect imaging, even
with the longest available interferometric
baselines and highest frequencies. This
advantage enables us to analyse the
morphological properties in exquisite
detail for any DSFG that is fortuitously
located in a calibrator field.
The fact that each calibrator is often
observed in several different ALMA frequency bands allows us to ensure that
the faint detections are genuine with high
confidence. Furthermore, such multiband data allows the study of the spectral indices of DSFGs at matched spatial
resolution, and determination of their
redshifts via detections of (sub-)millimetre
spectral lines (for example, Weiss et al.,
2009). Furthermore, the number counts
obtained for the selected galaxies, and
the properties derived for those galaxies,
come from sparse sampling of the astronomical sky and are thus relatively free
of cosmic variance. ALMACAL is not
however restricted to the detection of
continuum sources. Several additional
science goals include: blind searches for
emission-line galaxies in the ALMACAL
cubes to constrain the CO and [C II] luminosity functions (Decarli et al., 2014;
Popping et al., 2016) and the cosmic H2
mass density, to compare with models
(for example, Lagos et al., 2011); studies
of the physical mechanisms involved in the
emission of powerful jets from quasars;
and searches for common interstellar
species in the lines of sight towards bright
quasars (for example, M
uller et al., 2014).
41
-
Figure 1. ALMA Band 6 (1.2 mm) image of the spectacular jet emanating from the calibrator J1229+0203.
The multi-frequency observations provided by
ALMACAL enable the nature of these jets to be studied in exquisite detail.
Oteo I. et al., ALMACAL
-
J2206 0031
J2223 3137
-
Source selection
We have extracted from the ALMA
archive all calibration data between the
beginning of ALMA Cycle 2 and April
2016. No restriction on ALMA bands was
applied, since different bands are useful
for different science cases. First, the data
for each calibrator in each OB are calibrated, following standard procedures.
Then, for each observation, the calibrator
is subtracted from the data in the Fourier
uv plane using a point-source model,
and the visibilities and clean maps are
visually inspected, discarding all those
datasets which show evidence of poor
calibration. Next, all data for each calibrator in each band are combined after rescaling the visibility weights. In order to
detect DSFGs via their dust emission,
we focus the source extraction on ALMA
Band 6 (B6, 1.2 mm) and Band 7 (B7,
870m). Imaging in Bands 3 and 4 are
used to identify jets emanating from the
calibrators, which are quasars. Some
jets are very obvious once the central
bright calibrator is subtracted from the
maps (one of the most spectacular cases
found so far is shown in Figure 1). However, others may appear as unresolved
blobs and are easily confused with
DSFGs. Whenever possible, we compute
the B3 or B4 to B6 or B7 flux density
ratios for the B6-/B7-detected sources.
42
Ratios higher than unity indicate jets,

while lower ratios, or a lack of detection
in B3 and/or B4, are compatible with the
far-infrared spectral energy distribution
of a high-redshift DSFG. At the present
phase of the survey, where data for more
than 240 calibrators have been analysed,
we have found 15 robust ALMACAL
DSFGs (see Figure 2 for a small selection).
Their far-infrared spectral indices (B7 to
B6 flux density ratios) are compatible with
them being at z ~ 23.
The importance of ALMACAL DSFGs
The discovery of a population of DSFGs
with flux densities 510 times fainter
than the faintest SMGs from the SCUBA2
survey (Koprowski et al., 2015) allows
us to study the sub-mJy population that
might overlap with normal SFGs, such
as Ly emitters, star-forming galaxies
selected by their BzK colours (sBzK), or
Lyman-break galaxies, which all lack
Figure 2. ALMA images (Band 6, 1.2 mm) of four

c alibrator fields in each of which at least one DSFG
has been detected. Jets emanating from calibrators
(identified via 3 mm imaging) are represented by
the blue arrows, while DSFGs are indicated by white
open squares. The calibrators, which have been
subtracted in the uv plane using point-source
models, are located in the centre of the images
(green crosses).
dust emission detections in 95 % of

cases. Uniquely detecting dust emission
and measuring the total SFR of normal
SFGs, without the uncertainties of ultraviolet-based dust correction factors, is
feasible without having to rely on analysis
via stacking (Bthermin et al., 2012).
Additionally, faint ALMACAL DSFGs are
among the very few high-redshift populations, alongside bright and extreme
SMGs and gravitationally amplified systems (where interpretation is complicated
by lens modelling), for which superb
high-resolution imaging will be available
from self-calibration by the bright calibrator in the centre of the map.

LL

-2CDF

3'(26.1*
'@SRTJ@CDDS@K
%TIHLNSNDS@K
'@SRTJ@CDDS@K
.MNDS@K
*@QHLDS@K
2HLORNMDS@K

Figure 3. Cumulative
number counts of
DSFGs derived from
ALMACAL at 1.2 mm,
along with previous
results (Hatsukade et
al. 2013; Karim et al.
2013; Ono et al. 2014;
Fujimoto et al. 2015;
Simpson et al. 2015).
Number counts derived
in previous works have
been converted to
1.2mm by using the
average far-infrared
spectral energy distribution of SMGs at z ~ 2.3
(Swinbank et al., 2014).
The grey curve is the
fit obtained in Simpson
et al. (2015), extrapolated toward the flux
densities covered in the
present ALMACAL
phase.
The future
ALMACAL is continuously in progress,
adding more and more calibration data
on a daily basis. We will keep gathering
and combining observations while ALMA
keeps observing. This will likely result in
the largest ALMA (sub-)millimetre map
down to a depth of several Jy, a unique
dataset to study the nature of the faint
submillimetre population. Apart from
compiling more and more calibrator data,
future work includes the characterisation
of the ALMACAL detected sources via
multi-wavelength observations to determine their redshift (for the sources without multiple millimetre line detections),
stellar mass or obscuration. This step is
needed to explore the relation between
ALMACAL DSFGs and the normal SFG
population typically represented by Ly
emitters, Lyman-break galaxies or sBzK
galaxies selected in ultra-violet/optical/
near-infrared-based surveys.

2 LLL)X
The (sub-)millimetre number counts

The cumulative number counts (i.e., the
number of galaxies above a given flux
density) is an observable that any credible model of galaxy formation and evolution must match. Most recent models
can reproduce the bright end of the number counts (S1.2 mm > 1 mJy), but there are
disagreements at fainter flux density levels. For example, model predictions
based on the redshift evolution of the
mass function of SFGs (Bthermin et al.,
2011) give a number density of sources
about five times lower than those based
on the spectral energy distribution of
the galaxies (da Cunha et al., 2013) and
about 2.5 times smaller than predictions
based on epoch-dependent luminosity
functions (Cai et al., 2013). Therefore,
firmer constraints on the faint end of the
(sub-)millimetre number counts are
needed to discriminate between the
plausible models. At present, ALMACAL
has derived number counts down to
S1.2mm ~ 0.2 mJy, resolving about 50 %
of the extragalactic background light,
and those counts are found to be lower
than previously reported (Figure 3).
References
One of the main differences between

ALMACAL and previous work is the
source extraction technique. Whilst in
ALMACAL only sources detected at
>5 are included in the sample, other
works employ significantly lower thresholds to compensate for the lower area
and/or sensitivity and boost the number
of sources. This leads to contamination
by spurious detections. Although statis
tical corrections are normally applied,
these are not always accurate. Recently,
Hatsukade et al. (2016) also obtained
slightly lower number counts, considering
only >5 detections. One of the unique
aspects of ALMACAL is that most of the
galaxies are observed in multiple bands,
so we can be confident that all ALMACAL
DSFGs are real. Furthermore, and very
importantly, since ALMA calibrators are
distributed over the whole southern and
equatorial sky, the ALMACAL (sub-)millimetre number counts are relatively free
from cosmic variance. The apparent flattening we observe in the (sub-)millimetre
number counts towards low flux densities
has strong implications for our understanding of the dust emission from (and
attenuation in) SFGs. ALMACAL will be
the key to a fuller understanding.
Bthermin, M. et al. 2011, A&A, 529, 4

Cai, Z.-Y. et al. 2013, ApJ, 768, 21
Casey, C. M. et al. 2014, PhR, 541, 45
da Cunha, E. et al. 2013, ApJ, 765, 9D
Decarli, R. et al. 2014, ApJ, 782, 78
Fomalont, E. et al. 2014, The Messenger, 155, 19
Fujimoto, S. et al. 2016, ApJS, 222, 1
Hatsukade, B. et al. 2013, ApJL, 769, L27
Hatsukade, B. et al. 2016, PASJ, 34
Karim, A. et al. 2013, MNRAS, 432, 2
Lagos, C. del P. et al. 2011, MNRAS, 418, 1649
Muller, S. et al. 2014, A&A, 566, 112
Ono, Y. et al. 2014, ApJ, 795, 5
Oteo, I. et al. 2013, A&A, 554, 3
Oteo, I. et al. 2016, ApJ, 822, 36
Pearson, T. J. & Readhead, A. C. S. 1984, ARA&A,
22, 97
Popping, G. et al. 2016, arXiv:1602.02761P
Simpson, J. M. et al. 2015, ApJ, 807, 128
Smail, I., Ivison, R. & Blain, A. W. 1997, ApJ, 490, L5
Swinbank, A. M. et al. 2014, MNRAS, 438, 1267
Weiss, A. et al. 2009, ApJ, 705, 45
43
Astronomical News
Installation of the star-roof for the ESO Supernova

Planetarium & Visitor Centre on 12 April 2016.
The roof consists of glass panels set in a metal
framework with representations of some southern
constellations.
Astronomical News
Light Phenomena over the ESO Observatories III:

Zodiacal Light
1
2
Y. Beletsky/ESO
Petr Horlek 1
Lars Lindberg Christensen 1
David Nesvorn 2
Rebecca Davies 1
ESO
Dept. of Space Studies, Southwest
Research Institute, Boulder, USA
The zodiacal light is often seen at the

ESO observatories in the hours after
sunset and before sunrise. The origin
of the zodiacal light is described and
recent research briefly summarised.
Some fine images of the zodiacal light
from Paranal and La Silla, including the
full extent of the night sky are presented.
Even after the brightest part of the zodiacal light is below the horizon, faint traces
of it are still present. During the night it
takes the shape of an extremely faint
wispy bridge that brightens again in the
early morning, before sunrise. Figures3
and 4 show sophisticated panorama
images of the zodiacal light from sunset
to sunrise, clearly demonstrating its full
extent across the sky.
The nature of the zodiacal light
The origin of the zodiacal light is to be
found in the inner Solar System. Infrared
observations from the Infrared Astronom-
Figure 1. (Above) Column of zodiacal light (right)

photographed from the platform at Paranal Obser
vatory. The bright source within the zodiacal light
cone is Venus.
Figure 2. (Below) The zodiacal light photographed

soon after sunset from La Silla Observatory.
ical Satellite (IRAS1; Low et al., 1984) and

Cosmic Background Explorer (COBE2;
Reach et al., 1995) revealed emission
from small grains composed of dust and
ice surrounding the Sun. Most of the ob
served particles have sizes in the range
1 to 100 m. The PoyntingRobertson
effect (Guess, 1962) forces particles
inward through the absorption of solar

radiation and isotropic emission, reducing
their angular momentum, while the dominant force for micrometre-sized particles
is the solar radiation pressure that accelerates them away from the Sun. The
zodiacal dust hence needs to be constantly replenished. This replenishment is
Y. Beletsky/ESO
On a moonless night from Paranal Observatory in the Atacama Desert in Chile, it

is possible to see several different largescale sky phenomena in addition to the
faint light from the Milky Way. These
include atmospheric phenomena, such
as airglow natural radiation arising
in the Earths atmosphere (Christensen
et al., 2016) and occasional flashes
called sprites that can appear over the
Andes and are connected with electrical
storms (Horlek et al., 2016). Another
light phenomenon that can be seen on
every moonless night at a very dark site
is the zodiacal light (Figure 1). Most evident in the hours after dusk or before
dawn, zodiacal light appears as a faint,
diffuse column of light in the sky, just
above the horizon and extending towards
the zenith (Figure 2).
45
Astronomical News
Horlek P. et al., Light Phenomena over the ESO Observatories III: Zodiacal Light
primarily served by crumbling icy comets,

but also by colliding asteroids and possibly interstellar dust (Rowan-Robinson &
May, 2013). The zodiacal light itself is
sunlight that is forward-scattered from
these particles in the direction of Earth.
period comets, with orbital periods less

than 20 years, that take their name
from the fact that their current orbits are
determined primarily by the gravitational
influence of Jupiter. The Jupiter family
of comets contains 67P/Churyumov
Gerasimenko, now extremely well known
as a result of the in situ research carried
out by ESAs Rosetta satellite in August
2014. The Philae probe landed on the
nucleus of this comet in November 2014.
Figure 3. (Upper) All-night 360-degree panorama of

the zodiacal light from Mauna Kea in April 2011,
showing the structure of the zodiacal light almost in
its entirety. On the left and right, the columns of light
are visible after dusk (right) and before dawn (left)
respectively, brighter closer to the Sun (which is
under the left and right horizons). In the middle of the
band, the Gegenschein appears at the antisolar point.
Figure 4. (Lower) 270-degree panorama of the
zodiacal light, photographed from La Silla in April
2016, showing almost the full structure. On the right,
the column of light is visible before dawn, brighter
closer to the Sun (which is below the right horizon).
To the left, the Gegenschein appears at the antisolar
point. Unique hints of substructure are visible around
the Gegenschein. Extensive image processing was
performed by Petr Horlek and Miloslav Druckmller
to enhance the structures.
Sun
Venus
330
The dust grains that give rise to zodiacal light are distributed in a disc in the
plane of the ecliptic (Nesvorn et al.,
2003). When viewed from Earth, this disc
appears as a band across the sky passing through the constellations of the
zodiac, hence the name zodiacal light.
As the scattering of sunlight is most
effective at smaller angular distances
from the Sun, the band of light along the
ecliptic gets fainter and narrower further
away from the Sun (Figure 3). Along the
ecliptic, at the point in the sky opposite
the Sun (the antisolar point), coherent
backscattering from dust particles further out in the Solar System beyond the
Earths orbit leads to the oval patch of
270
300
240
210
Saturn
180
150
Observations of the zodiacal light

The zodiacal light was first investigated
in the late 1600s, by the Italian astronomer Giovanni Cassini and the Swiss
mathematician Nicolas Fatio de Duillier.
Of course, back in the 17th century there
was very little light pollution, so it was
relatively easy to observe this phenomenon, even from cities. Observations of
zodiacal light in more recent times have
mostly focused on particular structures of
the zodiacal cloud. Interesting zodiacal
dust bands were discovered by IRAS in
1984 (Low et al., 1984; Dermott et al.,
1984). They are produced by collisions of
asteroids in the Main Belt between Mars
and Jupiter within the last few million
years (Nesvorn et al., 2003). Hints of
these dust bands are noticeable on the
full-night panoramic imaging of the zodiacal light taken by commercial cameras
after post-processing, including colour
120
90
60
Ecliptic longitude
30
Sun
30 N
30 N
Ecliptic latitude
30 S
Miloslav Druckmller & Shadia Habbal
Recent studies show that over 85 % of

the cometary material in the zodiacal
cloud comes from dust from comets in
the Jupiter family (Nesvorn et al., 2010;
Schulz et al., 2015). These are short-
light known as the Gegenschein (Figures3, 4 and 5). The name was given
by the German explorer Alexander von
Humboldt (17691859).
30 S
30 W
60 W
90 W
120 W
150 W
180
150 E
120 E
90 E
60 E
30 E
Elongation
ESO/P. Horlek & M. Druckmller
46
Astronomical News
Y. Beletsky/ESO
Figure 5. Image of the Gegenschein obtained in

October 2007 above Paranal Observatory.
correction of the background (see Figures3 and4).

Subsequent observations with IRAS and
more sensitive photometric observations
from NASAs STEREO satellites3 have
revealed a circumsolar dust concentration in a ring along the orbit of Venus.
This was expected and is caused by the
influence of the planets gravitational
resonance on the particles (Dermott et
al., 1994; Jones et al., 2013). A brightening of the zodiacal cloud by ~ 10 % near
the apex of the Earths orbital motion has
also been found (Dermott et al., 1994).
Recent observations of nearby stars have
demonstrated that the Solar System may
not be the only one to exhibit zodiacal
light. Data from the Very Large Telescope
Interferometer have revealed that other
planetary systems are also surrounded
by interplanetary dust leading to zodiacal
light, but much brighter than in the Solar
System (for example, Lebreton et al., 2013;
Marion et al., 2014; Ertel et al., 2015).
Acknowledgements
The contribution of David Nesvorn was supported
by the Southwest Research Institute. We are grateful
to Prof. Miloslav Druckmller (Inst. of Mathematics,
Brno University of Technology, Czech Republic) for

significant support of the image processing of Figure 4,
and to Bob Fosbury for illuminating discussions.
References
Christensen, L. L., Noll, S. & Horlek, P. 2016,
The Messenger, 163, 38
Dermott, S. F. et al. 1994, Nature, 369, 719
Dermott, S. F. et al. 1984, Nature, 312, 505
Ertel, S. et al. 2015, The Messenger, 159, 24
Guess, A. W. 1962, AJ, 135, 855
Horlek, P. et al. 2016, The Messenger, 163, 41
Jones, M. H., Bewsher, D. & Brown, D. S. 2013,
Science, 342, 960
Lebreton, J. et al. 2013, A&A, 555, A146
Low, F. J. et al. 1984, AJ, 278, L15

Marion, L. et al. 2014, A&A, 570, A127
Nesvorn, D. et al. 2003, ApJ, 591, 486
Nesvorn, D. et al. 2010, ApJ, 713, 816
Reach, W. T. et al. 1995, Nature, 374, 521
Rowan-Robinson, M. & May, B. 2013, MNRAS,
429, 2894
Schulz, R. et al. 2015, Nature, 515, 216
Links
1
IRAS: http://irsa.ipac.caltech.edu/Missions/iras.html
C OBE satellite: http://lambda.gsfc.nasa.gov/
product/cobe/
3
N ASA STEREO satellites: http://stereo.gsfc.nasa.
gov/mission/mission.shtml
2
The First NEON School in La Silla
Michel Dennefeld 1
Claudio Melo 2
Fernando Selman 2
1
Institut dAstrophysique de Paris,

CNRS, and Universit P. et M. Curie,
Paris, France
2
ESO
The NEON Observing Schools have long

provided PhD students with practical
experience in the preparation, e
xecution
and reduction of astronomical observations, primarily at northern observatories. The NEON School was held in Chile
for the first time, with observations
being conducted at LaSilla. The school
was attended by 20students, all from
South America, and observations were
performed with two telescopes, including the New Technology Telescope.

A brief description of the school is presented and the observing projects and
their results are described.
After many years of discussions and
preparations, the NEON Observing
School could finally take place at La Silla
47
Astronomical News
Dennefeld M. et al., The First NEON School in La Silla
for the first time, from 22 February to

4March 2016, with joint sponsorship
from ESO and OPTICON1 and strong
support from the ESO Director General,
the Director for Science and the Heads
of the Office for Science.
As is well known by now, the NEON
Schools are intended to provide PhD students with practical training on how to
prepare and execute an observing programme. The schools have taken place
in several active observatories, comprising the Network of European Observa
tories in the North (NEON; including
Asiago, Calar Alto, Haute-Provence and
LaPalma), with which ESO is also associated. Devoting a small fraction of the time
at 4- and 2-metre-class telescopes to
training purposes can thus be regarded
as a useful investment for the overall efficiency of all the facilities, developing the
observing skills of the next generation of
astronomers.
The format of the schools is similar:
beginning with writing a proposal followed
by the preparation of observations
(instrument set-up, calibrations, etc.) and
the observations themselves, then completing the data reduction and finally presenting the results. Work is carried out in
small groups of four students, with each
group under the supervision of an experienced tutor, who also provides the science project to be executed. This kind of
training is much needed, as instruments
are getting more and more complex, and
many observatories (and ESO not least...)
offer a large fraction of their observing
time at the large telescopes in Service
Mode, so that practical experience can
be lacking for many young astronomers.
Figure 1. Two of the

student project groups
around the EFOSC2
Nasmyth focus of the
NTT during the NEON
School at La Silla.
training for future use of larger (VLT) facilities. But it is still a bit too automatic to
understand all the details needed to
operate less advanced facilities. Thus,
since according to the local Chileans,
NTT also means No Tocar al Telescopio
(Do not touch the telescope), we also
used the Danish 1.54-metre telescope,
made available during Czech time
thanks to Petr Pravec and with the support of Jan Janik (Mazaryk University),
who came over specially to support our
run. The various science programmes
used, as far as possible, a combination of
both telescopes and instruments.
The observations were prepared in
Santiago, prefaced by a set of lectures
on fundamental observing techniques,
instrument and detector properties and
data reduction. The full programme of the
school can be found on the dedicated
website2. The tutors leading the observations in groups were all current, or former,
ESO staff in Chile, and thus perfectly
aware of all the requirements and instrument performance.
The La Silla school
Group projects
So, how did it work in La Silla? Thanks

to a generous allocation from the Directors Discretionnary Time, the New Technology Telescope (NTT) was made available for three nights for the school, close
to full Moon. As the operations there are
quite similar to the Very large Telescope
(VLT), with an operator for the telescope,
the same set of observing preparation
tools and operations principles (observing blocks to prepare, pipelines for
reduction, etc.), this provided an excellent
The group led by Amelia Bayo (University

of Valparaiso) and Koraljka Muzic (ESO,
and University of Diego Portales) studied
the widest binary stars in young stellar
groups from the Search for Associations
Containing Young stars (SACY) sample,
to find out whether or not they belonged
to the same group. The H and LiI6708
lines were used as age indicators and
observed with the ESO Faint Object
Spectrograph and Camera (EFOSC),
and the K I + Na I lines were observed
48
with Son of ISAAC (SOFI). This approach

allowed four out of the eight stars ob
served to be confirmed as young stars of
the same age.
The group led by Fernando Selman and
Evelyn Johnston (ESO) tried to disentangle the counter-rotating discs in the lenticular galaxy IC 719. Long-slit spectroscopy with EFOSC for velocities, and B,
V, R, H photometry with the Danish
1.54-metre (the H filter had inexplicably
disappeared...) revealed the different
structures, and that more than one stellar
component (identified through Lick indices) was present from the radial velocities. Evidence was also found for a young
stellar population. This is clearly a case
where a full 3D map with the Multi-Unit
Spectroscopic Explorer (MUSE) would
reveal the full power of this technique.
The group led by Giacomo Beccari (ESO)
tried to find protoplanetary discs around
young stars, from ultra-violet, H or infrared excesses. Multi-band photometry
of candidate regions in Carina was done
at the Danish 1.54-metre telescope. As
H imaging was not possible, Brackett
imaging was done instead with SOFI at
the NTT. Finally, six new disc candidates
were identified, whose confirmation is
pending with spectroscopy.
Linda Schmidtobreick (ESO) observed a
candidate cataclysmic variable detected
in the Sloan Digital Sky Survey (SDSS)
with her group, in order to measure/confirm its orbital period both by photometry,
and by radial velocity measurements.
Long series of imaging (with the Danish
1.54-metre telescope) and spectroscopy
(with EFOSC at the NTT) data were thus
and La Silla, including nice lunches in the

ESO gardens in Vitacura.
Prospects
Figure 2. All NEON School students and tutors pose

in the dome of the ESO 3.6-metre telescope for a
group portrait.
obtained, the expected periods being

about 80 minutes. The students, after
being confronted by all sorts of drifting
and aliasing problems, could finally show
that the real period was in fact closer to
34 hours for one target, and 67 hours
for the other, but longer series of data
are still needed to confirm these values.
Another, unknown, variable star was also
found in one of the fields, the nature of
which remains to be clarified.
Finally, Bruno Dias (ESO) conducted an
ambitious, and difficult (at least for beginners) multi-object spectroscopy programme at the NTT with his group, to
determine the nature of the globular cluster NGC 3201: whether it is a member
of the Milky Way or an absorbed dwarf
galaxy. This first required pre-imaging,
then preparation of the masks, and finally
observing during the last night of the run.
After all the hard work, those students
presented superb results, showing a
trend of Fe abundance versus heliocentric velocity (demonstrating that the
selected stars did indeed belong to the
cluster). The CH versus CN indices
showed that two different abundance
groups were present, but both belonging
to the Milky Way.
In addition to their main tasks, two of
the groups were triggered to observe a
Target of Opportunity (a situation that
many observers have to face nowadays),
led by Michel Dennefeld (IAP, Paris). The
targets were provided by the European
Space Agencys Gaia satellite, which has
very recently started to provide dozens of
alerts per day. One of them, Gaia16agf,
proved to be a young Type Ia supernova,

caught about one week before maximum,
while the other, Gaia16afz, was classified
as a dwarf nova in eruption. Both results
were immediately published as Astronomers Telegrams, Nos. 8754 and 8766,
respectively. Many of the other results
from the student groups will also lead to
publications, but over a slightly longer
timescale.
Back to Santiago
The last days in Santiago were mainly
devoted to data reduction, before the scientific results were presented on the last
day of the school. In addition, lectures
on other topics of great interest for the
future were presented: on the Atacama
Large Millimeter/submilllimeter Array
(ALMA), on adaptive optics, the European
Extremely Large Telescope, etc. There
were also lively discussions about career
prospects, how to make a good presentation or how to write a good telescope
observing proposal.
Overall, the 20 students, all from South
America, selected from over 140 applicants in total, were very enthusiastic,
worked hard during the two weeks, both
in Santiago and on the mountain, and
presented excellent results at the end.
The groups of four were very dynamic,
covering a wide range of levels (from first
year PhD to postdoc) and origins,
although the majority came from those
countries with a larger astronomical
population, namely Argentina, Brazil and
Chile, and with a good gender balance.
The ambiance was very friendly, as it
should be for a summer school, favoured
by nice weather and excellent environment at the ESO premises in Santiago
One may wonder about the motivation to

organise such a school at La Silla, as it
is not especially intended for European
students, who have ample opportunities
to attend the regular NEON Schools in
Europe, except for those students doing
their PhDs in Chile. In fact, there is a high
demand, as shown by the huge number
of applications, and it is in ESOs interests
(and the European community at large) to
ensure the most efficient use of its facilities, which are open to everybody, since
the allocation of observing time is based
on scientific merit only. In addition, the
operation of all the facilities in Chile
requires a large number of dedicated and
skilled astronomers and engineers, who
have to be trained somewhere. In this
sense, such a school also reinforces the
European links with the local commu
nities and emphasises the various job
opportunities for the future. The success
of the present event is already generating
suggestions for future similar schools, a
point well taken at ESO, and enthusiastically supported by the ESO tutors. It is a
revival of past successes as the first
observing school was organised jointly by
ESO and Centre national de la recherche
scientifique (CNRS) at the Haute-Provence
Observatory in 1988 (Chalabaev &
DOdorico, 1988).
Acknowledgements
This NEON School would not have been so successful without the efficient support of many people:
Paulina Jiron and Mara Eugenia Gmez (Office for
Science in Vitacura); Ivo Saviane, Javier Duk and
the whole La Silla Logistics Team; Esteban Osorio
and the IT team in Santiago, and many others, all of
whom we warmly thank.
References
Chalabaev, A. & DOdorico, S. 1988, The Messenger,
53, 11
Links
1
2
OPTICON: http://www.astro-opticon.org
School website: http://www.eso.org/sci/meetings/2016/lasilla_school2016.html
49
Astronomical News
Report on the
ESO Data Simulation Workshop

held at ESO Headquarters, Garching, Germany, 1415 April 2016
Kieran Leschinski/Univ. Vienna
Pascal Ballester 1
ESO
The role of simulated data is increasing

rapidly across all phases of instrumentation projects, from design to scientific
exploitation. The many commonalities
among ESO instruments, their reduction
software and archival products, makes
it especially worthwhile to exchange
knowledge between their instrument
teams. The data simulation workshop
was the first of its kind to bring together
the ESO instrument simulator community and a brief overview of the workshop is presented. The participants
expressed strong interest in continuing
to exchange knowledge in this area.
Workshop themes
End-to-end detector array simulations
of observation and calibration data are
applied in many phases of instrument
projects, from the assessment of instrument performance, optimisation of instrument design, validation of science cases,
preparation of the exposure-time calculators and observing tools to evaluation
and testing of data reduction algorithms,
optimisation of observing and calibration
strategies, as well as for the design of
physical-model driven calibrations.
The purpose of the data simulation workshop was to provide a forum for specialists from the instrument consortia and
from ESO engineering departments, to
exchange information, methods, and
experience about their respective simu
lation approaches and results. Around
50participants from about ten different
institutes and ESO attended, and a video
conference connection allowed participants at four other institutes, in Marseilles,
Oxford, Groningen and Innsbruck, to
contribute to the workshop The workshop was structured over two half days,
with 11 talks and two lively discussion
sessions. On Thursday evening, dinner
at an Indian restaurant in Garching provided a relaxed setting for continuing
discussions.
50
The first half-day included talks by the

instrument consortia, presenting the
intended usage, status and current
results of their simulation projects. The
projects represented included the European Extremely Large Telescope (E-ELT)
instruments HARMONI (first light integral
field spectrograph), MICADO (first light
adaptive optics imager) and METIS (first
light mid-infrared imager and spectrograph), as well as the spectroscopy survey facilities 4MOST (4-metre MultiObject Spectroscopic Telescope) and
MOONS (Multi-Object Optical and Nearinfrared Spectrograph). While some
projects are already completed, like the
Virtual MOONS simulation (Li Causi et
al., 2014), a number of first results were
shown for the MICADO (Leschinski et al.,
2016) and METIS (Schmalz et al., 2012)
instruments; see Figures 1 and 2 respectively. Many teams are building upon the
experience acquired in previous simulation
projects: for instance the Multi-Unit Spectroscopic Explorer (MUSE) instrument
numerical model (Jarno, 2012) is being
used in the development of HARMONI.
Time was also reserved for discussion
sessions. During the first day, the topics
of discussion covered configuration control of the projects, sharing of common
models and data, and the validation of
models. The first topic of discussion was
Figure 1. A simulated MICADO observation of a

~2000 MA open cluster in the Large Magellanic
Cloud produced with SimCADO. The left image
shows the raw readout from the central H4RG infrared detector chip in the MICADO detector array.
The SimCADO package simulates all aspects of the
optical train including the atmosphere, telescope
optics, various adaptive optics correction modes
and the MICADO optical design as well as the detector characteristics.
the integration of simulations within and

between the projects, in terms of change
and configuration control. Shared repositories of simulation results for the common parts, such as telescope wavefront
errors or adaptive optics simulations,
were also discussed. Since all the simulation projects presented make use of sky
radiance and transmission data from the
Austrian in-kind sky radiance model (Noll
et al., 2012), a programmatic interface to
the model, as a complement to the web
application1 Skycalc, was addressed. The
validation of the simulations was also a
topic of interest. Different approaches like
cross-instrument modelling were proposed. The respective advantages of
faster, but simpler simulations versus
advanced models based on Fourier
optics were also discussed.
During the second day, presentations
were provided by ESO engineers on
aspects of E-ELT modelling, including
telescope wavefront errors, telescope
vibrations, and adaptive optics modelling
Jeff Meisner/Leiden Observatory
Figure 2. Using a model of a star surrounded by

cooler circumstellar material, the METIS simulator
predicts detector images using four different camera
filters (from left to right) L:-band (3.8 m), M-band
(5m), N-band (10.7 m), and Q-band (17.7 m). At
shorter wavelengths the hotter star dominates; at
longer wavelengths the nebulous component is
brighter and the instrument sensitivity decreases.
and point spread function simulation.

The main philosophy for modelling and
simulation of the E-ELT was discussed
and presented. The project d
eliberately
avoids complex end-to-end modelling.
Instead, aspects of the temporal and
Figure 3. Sensitivity response of the amplitude of the
wavefront error (WFE) with the vibration frequency
from the telescope pier.
spatial frequencies for the dynamics

and control loops and the perturbations
acting on the telescope are explored.
The analysis and simulation environments
are split into three different simulation
toolkits: 1) active optics and phasing
toolkit; 2) telescope dynamical and control toolkit; 3) adaptive optics toolkit
(Octopus). This approach provides more
flexibility to adjust the models to dedicated purposes and also reduces the
computational effort. A model-based
vibration sensitivity analysis and budgeting using the dynamical and control
toolkit was presented (Figure 3).
A session on physical-model driven calibration presented results obtained in the
Babak Sedgh/ESO
6@UDEQNMSDQQNQ@ETMBSHNMNEEQDPTDMBX@MCDWBHS@SHNM@LOKHSTCD
area of data quality control and science

reduction for the CRyogenic Infra-Red
Echelle Spectrometer (CRIRES), the
Ultra-violet Visible Echelle Spectrograph
(UVES) and the X-shooter instruments
(Bristow, 2010). The Austrian in-kind sky
radiance model, used by all ongoing
instrument simulation projects, was also
presented, and a specific discussion on
modelling of infrared detectors concluded
the workshop.
The slides of all talks are provided on
the workshop web page2. The workshop was a very fertile platform for the
exchange of practical and theoretical
information among the participants. The
participants expressed their interest in
a continuation of this type of meeting and
further sharing of paper references,
methods and tools useful for data simulations.

Acknowledgements
LOKHSTCDMLML

Special thanks go to Martine Peltzer for help with

organising the meeting and to the ESO IT team for
their efficient support.
References

l
Bristow, P. et al. 2010, Proc. SPIE, 7735, 271

Jarno, A. et al. 2012, Proc. SPIE, 8449, 84490A
Leschinski, K. et al. 2016, Proc. ASP Conf. Ser.
ADASS XXV, in press
Li Causi, G. et al. 2014, Proc. SPIE, 9147, 64-1
Noll, S. et al. 2012, A&A, 543, A92
Schmalzl, E. et al. 2012, Proc. SPIE, 8449, 1P
l
Links
l
l

%QDPTDMBX'Y

kycalc web interface to the Austrian in-kind sky

S
radiance model: https://www.eso.org/observing/
etc/skycalc/
2
Workshop programme, list of participants and
presentations: http://www.eso.org/sci/meetings/2016/simu2016.html
51
Astronomical News
Retirement of Lothar Noethe
Jason Spyromilio 1
Ronald Holzlhner 1
1
ESO
Lothar Noethe retired in April after 33

years at ESO. An appreciation of his
contributions is presented, in particular
his pioneering work in active optics and
his role in the commissioning of the
NTT and VLT.
Lothar Noethe joined ESO in January
1983 to work in the Telescope Group led
by Ray Wilson. Lothar had studied physics at Bochum University and theoretical
physics at Liverpool University and
worked for Siemens. In this period Ray
was developing active optics and a thin
(19 mm) 1-metre test mirror with 78 actuators was the experimental testbed
(Noethe et al., 1986; Madsen, 2012, p. 121).
Active optics
Controlling the shape of telescope mirrors and their relative position has been a
challenge for telescope builders since the
earliest times. Passive control (for example, by mechanical astatic supports and
Serrurier trusses), as well as semi-active
operations (for example, by pneumatic or
vacuum mirror supports) had been in use
at telescopes for many years before the
advent of what we now call active optics.
The breakthrough for active optics was
to build a system that would operate in
closed loop on short enough timescales
that the engineering of the telescope
could be simplified, making for cheaper
and better telescopes. Active optics, in
contrast to adaptive optics, corrects the
fine alignment and optical figure of a telescope under temperature and gravity
vector variations, typically on a timescale
of minutes (Noethe, 2002). The enabling
technology was the digital detectors that
could be used to sense the wavefront;
the enabling thinking was the partnership
between Ray Wilson and Lothar Noethe.
In todays environment it is difficult to
contemplate an era when digital imaging
was at the cutting edge of telescope
technology. A charge-coupled device
52
(CCD) was first deployed at La Silla on

the Danish 1.54-metre telescope in 1982.
The CCDs were tiny, had poor quantum
efficiency and high readout noise. Computing resources were limited and efficiency in the algorithms was paramount.
The New Technology Telescope (NTT)
project was audaciously launched at
approximately the same time. It is worth
noting that as late as the early 1990s
there were highly respectable voices in
the field who held that the thin meniscus
mirror would never be able to deliver
images of the same quality as stiffer mirrors. Lothars keen mathematical mind
excelled in this environment and he solved
the problem of quickly processing the
data from the wavefront sensors into mirror positioning and shaping commands.
While optical engineers had long worked
in Zernike space when dealing with aberrations, Lothar revolutionised the field by
shifting the control of the optics to natural
modes. These, in effect, are the mecha
nical modes in which the mirror bends
most naturally, whose deformations are
significantly simpler to realise and require
less force and therefore are easier to
control. As Ray commented, this is one
of the great contributions of Lothar
Noethe (Wilson, 2003). The shift in thinking that this brought about is profound,
not only in the control of the optics and
the cost of the mirror cell and actuators
through the use of lightweight substrates,
but also, very critically, in the polishing
requirements for the mirrors.
NTT to VLT
Lothar was a key player in the commissioning of the NTT, and later for the Very
Large Telescope (VLT), and present for
all first stars (the first time the telescope
sees a star). Lothars work was critical in
making it appear to the outside world
that a smooth and elegant transition from
the assembly of the telescopes to their
first lights was taking place. Without
Lothar it would have been neither smooth
nor elegant, but there was also an element of luck involved. The initial success
of the NTT, with the spectacular appearance of astronomical images with an
image quality of 0.33 arcseconds during
the commissioning night of 2223 March
1989 (Wilson, 1989), was astounding; a
Simulated
Real
Figure 1. Simulated and real letters written in the

sky, produced by creating aberrations with the VLT
active primary mirror.
value difficult to believe at the time. This

was of course more than a year before
the launch of the Hubble Space Telescope. As observations with the Differential Image Motion Monitor (DIMM)
showed, this was a night with exceptionally good seeing (Sarazin, 1989). But not
much later, long-term atmospheric variations substantially increased the mean
seeing at La Silla for a period of several
years the great success of the NTT
first light would have had less impact
under less favourable seeing conditions.
Lothar continued as a pioneer of active
optics, which was also the foundation
for the primary mirror of the Very Large
Telescope (VLT). First star for Unit Telescope 1 (UT1) was achieved three weeks
prior to the official first light (25 May 1998).
Using just the guide probe wavefront
sensor Lothar, supported by Stephane
Guisard and Roberto Abuter, was able
to work with the active optics system to
give the commissioning team the confidence to communicate within ESO that
the telescope was likely to work. Semicryptic public releases of information
advised those in the know that all was
going well. Some first stars were truly
ugly and Lothar would compute manually
the commands to transform stars that
were 30 arcseconds in size and stretched
out like strings of beads, into small
roundish images that the active optics
could deal with.
New envelopes
Figure 2. Lothar Noethe pictured in one of the optics

laboratories at ESO Headquarters. The apparatus
behind Lothar is a laser tracker, capable of measuring distances on the micrometre level.
A couple of anecdotes deserve recalling.

After the first few million active optics
corrections on UT1 (a couple of months
into the commissioning of the telescope)
the logging system that is an integral part
of the VLT operational system could be
used to provide beautiful statistics of the
various aberrations. By this stage the
telescope was providing exquisite imaging down to 0.26 arcseconds in the
visible. The statistics were sent to Lothar
as a recognition of his excellent work.
Lothar immediately reacted, noting that
the standard deviation of the focus term
was too high and that something was not
quite right. Some probing by him in the
software resulted in a change in sign in
the force setting for the outer ring of
actuators for the primary mirror, and the
almost perfect image quality became that
little bit better.
One standard test of a telescopes performance is to make it write using starlight, by moving the telescope in particular patterns on the sky while exposing
the detector. The NTT was made to write
its own name. On the VLT Lothar decided
that he could beat this: he created a mathematical description of the aberrations,
converted them to forces on the primary
mirror and movements of the secondary
mirror as needed so that the stars would
have the shapes of the letters V, L and
T (see Figure 1). This was a strong demonstration that active optics had not only
arrived, but was firmly under control.
After the commissioning of all the VLT

UTs and the beginning of discussions
about the next step to larger telescopes,
Lothar became involved with control of
segmented mirrors. In 2002, he went to
the University of California at Irvine for
six months to work with Gary Chanan
and developed a long and fruitful collaboration on the challenges of active optics
with segmented telescopes. Lothar also
obtained a Habilitation (German higher
degree) in 2001 at the Berlin Technische
Universitt, again on active optics. He
was appointed head of the Optical Systems Department in 2003. When ESO
embarked on the E-ELT project, Lothar
was critical in pushing for novel systems
engineering solutions (Karban et al., 2008)
and for ESO to maintain its position as an
innovation centre for telescopes. Lothar
also had a clear vision for the needs of
complex projects to be managed through
requirements and structured thinking
and contributed towards the adoption of
a proper systems engineering culture at
ESO.
Most recently Lothar looked at applying
closed-loop active optics to wide-field telescopes. The correction procedure is more
demanding in a wide-field telescope
because of the tight alignment tolerances in
fast optics. The field dependence of aberrations in misaligned wide-field systems
also places tighter constraints on various
system degrees of freedom than is the
case with more conventional, narrow fieldof-view telescopes, such as the VLT. On
account of the higher sensitivity to aberrations and the large field of view, the science
images can be used to derive the aberrations. This realisation laid the foundation
for a novel method of active correction.
In the two wide-field telescopes operated by ESO, namely the 4-metre Visible
Infrared Survey Telescope for Astronomy
(VISTA) and the 2.6-metre VLT Survey
Telescope (VST), curvature wavefront
sensors are employed for the active
optics control, in contrast to the more
commonly used ShackHartmann (in
the case of the NTT and VLT) or pyramid
wavefront sensors. In a generalisation
of curvature wavefront sensing, Lothar
initiated an active optics control method
based on an analytical model of small
point spread function aberrations across

the science image. The scheme has been
implemented in Mathematica and successfully operated with OmegaCAM on
the VST during several technical nights in
2015 (Holzlhner et al., 2014). Development is still required to simplify the processing before this can become a standard technique for improving the image
quality of the VST.
Lothar has advised many projects outside
ESO and has selflessly supported the
work of many junior engineers to advance
the field. As Ray put it in his article
(Wilson, 2003), [Lothars] application for
the ESO job was one of the greatest
pieces of good fortune in our whole
active optics development. Telescope
design and operation, and ground-based
observational astronomy in general, have
been profoundly affected by Lothars
work.
Retirement
Lothar retired at the end of April 2016 and
a farewell lunch for the Optics Department was held on 29 April 2015 at ESO
Headquarters (he had asked not to have
a large farewell party organised). Besides
excelling in optical engineering, Lothar
has always been active in various sports,
in particular table tennis, which he plays
several times a week and coaches young
players at the local sports club. Never
ceasing to be curious and eager to learn,
he has now enrolled in an advanced algebra course at the Technische Universitt
Mnchen. We wish him a very active
retirement and look forward to more contributions in the field of telescope optics.
Acknowledgements
We gratefully thank Dietrich Baade and Katjuscha
Lockhart for comments.
References
Holzlhner, R. et al. 2014, SPIE, 9151, 2I
Karban, R. et al. 2008, Proc. SPIE, 7017, #1I
Madsen, C. 2012, The Jewel on the Mountaintop,
(London: Wiley-VCH)
Noethe, L. et al. 1986, SPIE, 628, 285N
Noethe, L. 2002, Progress in Optics, 43, 1
Sarazin, M. 1989, The Messenger, 56, 8
Wilson, R. 1989, The Messenger, 56, 1
Wilson, R. 2003, The Messenger, 113, 2
53
Astronomical News
Fellows at ESO
Katharina Immer
Katharina Immer
I do not exactly know when I started to

be interested in astronomy. For as long
as I can remember, I have watched the
night sky and the beautiful coloured
images of astronomical objects from the
Hubble Space Telescope and other telescopes. When I was a teenager, I got
a little reflecting telescope with which I
observed stars and the Moon.
In high school, I was particularly interested in physics and mathematics and
thus it was a natural step to decide to
study physics at the University of Bonn.
Since I already knew at that time that I
wanted to be an astrophysicist at the
end of my studies, I chose astronomy
as my minor subject. The introduction to
Galactic astronomy, as well as lectures
about microquasars and the interstellar
medium, were particularly interesting.
During my studies, I had several oppor
tunities to visit the 100-metre radio telescope in Effelsberg. Standing in front
of this huge telescope was very impressive. Since this telescope operates in
the radio wavelength regime, it can even
observe through heavy clouds and rain.
In 2008, I started my Diploma thesis at
the Max Planck Institute for Radio Astronomy (MPIfR) in Bonn, one of the hubs
of radio astronomy in Germany. I worked
with Karl Menten on Very Large Array
(VLA) observations of the dust ridge
clouds in the Galactic Centre in the search
for H II regions and thus ongoing star
formation in these clouds. Since we did
not find any radio sources in the clouds,
we could exclude the existence of stars
with spectral types earlier than B0.5.
I also worked on data from the infrared
spectrograph onboard the Spitzer Space
Telescope with Frederic Schuller. From
the slope of the spectra and the detected
emission lines, we sorted the observed
sources into young stellar objects and
late-type evolved stars. By the end, we
could derive a star formation rate for the
Galactic Centre.
For my PhD, I decided to stay in Karl
Mentens group in Bonn. In 2010, I went
for a year to the HarvardSmithsonian
Center for Astrophysics (CfA) in Cambridge, Massachusetts, as a predoctoral
54
fellow to work with Mark Reid on VLA

observations of DR21 and Submillimeter
Array observations of W33. In addition,
I am part of the Bar and Spiral Structure
Legacy Survey, led by Mark Reid, which
determines the trigonometric parallaxes
of star-forming regions in the Milky Way.
That year was one of the best of my life,
learning a lot about interferometry, submillimetre and radio observations, as
well as high-mass star formation, meeting
some of my closest collaborators and
friends, as well as my boyfriend, travelling
along the East coast of the US and having a wonderful time.
receivers at both APEX and ALMA as my

duties. This is a very interesting project,
giving me insights into the workings of
a new instrument and the different problems and quirks that come with it at the
beginning. The Science Verification of
these receivers brought me into contact
with cutting-edge science in very different astronomical fields, from star formation to evolved stars to extragalactic projects, and I learnt a lot about many topics
outside of my own field. I am now also
involved in the ongoing commissioning
and Science Verification for the Band9
receiver at APEX.
After I moved back to Bonn, I became

one of the visiting observers for the
Atacama Pathfinder Experiment (APEX)
telescope in Chile. During my time at the
MPIfR, I went to APEX for four runs. Here,
I learned how to observe with a single
dish telescope and gained insights into all
kinds of science projects that were conducted with this telescope. The experience that I gained helped me to improve
my observing proposals significantly.
For my science, I am still mostly focused

on high-mass star formation, trying to
understand the chemical evolution of the
molecular clouds during the star formation process as well as the kinematics in
the clouds. While I have started several
projects with colleagues within the star
formation group, the multi-science environment at ESO also sparked new ideas
and collaborations outside of my field
of expertise, which is both challenging
but rewarding.
After finishing my PhD in 2013 and a

short-term postdoc at the MPIfR, I
started at ESO in Chile in 2014. I decided
to come to ESO because I liked the duty
and science combination of the fellowship. Since I wanted to stay in contact
with APEX, but also get to know the
Atacama Large Millimeter/submillimeter
Array (ALMA), I chose the commissioning
and Science Verification of the Band5
I am looking forward to continuing with

research on high-mass star formation
and the work with submillimetre and radio
telescopes of this and future generations
like APEX, ALMA, the Jansky Very Large
Array (VLA), or the Square Kilometer
Array (SKA), during my remaining time at
ESO as well as in future positions during
my career.
Evelyn Johnston
Evelyn Johnston
I was always interested in science as
a child, but growing up in Belfast in
Northern Ireland, with the light pollution
and regular wet and cloudy weather,
meant that I was lucky to see any stars
at night. But when we went on holiday
and found clear, dark skies, I would often
be amazed at the number of stars that
would appear. I loved trying to identify
new constellations, looking at the planets
through binoculars, and wondering what
else was out there that I couldnt see.
As early as nursery school, my favourite
song was Twinkle twinkle little star,
although it was only later, when a teacher
told us that the Sun is a star, that I started
to really wonder about the Universe.
I never really considered astronomy
as a potential career until I started university, but I did know as a teenager that
I wanted to go to university and become
a scientist. At college I met very enthusiastic physics and maths teachers who
picked up on my interest and helped me
widen my knowledge beyond simply
what I needed to know to pass my exams.
Maybe it helped that, during those years,
I skipped classes to watch two partial
solar eclipses with my teachers, and
ended up explaining what was happening
to the crowds of interested students who
joined us.
I moved to Sheffield in England to do a
degree in Physics and Astronomy. During
those years, my interest in astronomy
grew, and for my Masters year I was

lucky enough to be offered the chance
to carry out my research project while
working as a student support astronomer
at the Isaac Newton Telescope (INT) in
La Palma. This position gave me my
first opportunity to experience life as an
astronomer at an observatory, and while
the job was hard, I thoroughly enjoyed
the experience. The INT is a rather handson telescope, where visitors have to do
everything themselves, including refilling
the cryostats, pointing the telescope,
opening and closing the dome, etc. As a
result, it gave me a real feel for how
astronomical observations are really carried out, the preparation that has to go
into planning the night, and a true appreciation of the work by the whole obser
vatory team to build and maintain the
precision of such impressive machinery.
My experience in La Palma convinced me
to do a PhD in astronomy, so I returned
to the UK for my PhD at the University
of Nottingham on how the star formation
in spiral galaxies shuts off to transform
them into lenticulars. I quickly became
interested in understanding the star formation histories within different components, such as bulges and discs, and
the role those structures play in the evolution of the galaxy as a whole. As part
of my PhD, I developed a technique to
apply bulgedisc decomposition to longslit spectra along the major axis of a
galaxy in order to separate the spectra
of these two components. From their
independent spectra, I was able to study
their individual stellar populations and

star formation histories.
I found that the lenticular galaxies in the
Virgo and Fornax clusters experienced
a final episode of star formation within the
bulge region once the discs had already
started fading. These young stellar populations within the bulge region suggest
that at some point during the transformation, small amounts of gas were funnelled
into the inner regions of the galaxy, where
the gas eventually became dense enough
to trigger a final episode of star formation.
I am now working to further develop my
spectroscopic bulgedisc decomposition
technique to apply it to integral field unit
(IFU) datacubes to make full use of the
spectral and spatial information. I hope to
be able to determine whether this final
central episode of star formation activity
occurs throughout the bulge or only at
the centre of the disc, and which processes are most likely to trigger it and the
dependence on environment.
During my PhD I only had one opportunity to go observing, since all the data
I needed had already been collected.
During that time, I found that I missed the
observing side of astronomy and wanted
the opportunity to experience a new
culture again. So, after my PhD, I moved
to ESO in Chile, and now have duties
on Unit Telescopes 1 and 4, particularly
working with the Multi-Unit Spectroscopic
Explorer (MUSE). Carrying out obser
vations with IFU spectrographs, such as
SINFONI, KMOS and MUSE, helps me
better understand the limitations of my
own plans with such data. I am enjoying
contributing to the running of such a
large observatory and learning more
about the instrumentation at Paranal from
the various specialists I work with there.
I also love learning my way around the
southern skies when I get a few free minutes to stargaze, and I hope that by the
end of my fellowship, my Spanish will be
good enough to talk people through the
constellations at local outreach events.
Wolfgang Kerzendorf
I have always been fascinated with
science. As a kid in elementary school
I was very interested in biology. From
experiments with growing tadpoles to
55
Astronomical News
catching various critters in the Medi

terranean Sea, I always wanted to know
how things work and live. Astronomy
was also an early passion of mine. I
remember that I was fascinated by the
planetarium at the Deutsches Museum
in Munich. This must have inspired my
parents to Xerox a star map (laser printers didnt exist then) to A3 format at the
local library. I cut self-adhesive phos
phorescent tape to the different sizes
of stars, representing the different
magnitudes, and this would illuminate
my bedroom at night.
During high school I started to become
more interested in physics. In particular,
I remember one project where our physics teacher built air-pressure powered
rockets with plastic soda bottles and
we would try to gauge how high things
flew. In the later years of high school, I
became very interested in computing and
helped to set up the computer labs. My
Facharbeit (end of high school thesis)
was a Java program that would calculate
and visualise the electric field of a radiating dipole (created with the help of a Professor of Physics in Wrzburg).
Despite my strong interest in computing,
I wanted to study physics and selected
a university in Germany that did active
research in astrophysics. Thus, I started
studying physics at Heidelberg University in the autumn 2002. My first serious
contact with astronomy came with my
advanced laboratory course in physics
that involved obtaining photometry of a
globular cluster and subsequently fitting
isochrones. My lab partner and I were
one of the few who were actually able to
obtain the required photometry on a cold
December night. After many nights configuring my computer to run Linux and
installing IRAF (a difficult endeavour even
in 2004), we were finally able to obtain
the photometry with DAOPhot and fitted
an isochrone. Despite the rather difficult
and arduous nature of this lab, I was
excited: finding out the age and distance
of a star cluster was definitely a highlight
of my undergraduate studies.
After two and a half years of study, I
started to be interested in doing a semester abroad. My tutor Rainer Wehrse
had his colleague Mike Bessel visiting
him and he encouraged me to meet him.
56
Wolfgang Kerzendorf
Mike suggested contacting a young

professor named Brian Schmidt at his
home institute, Mount Stromlo Obser
vatory. Brian was very enthusiastic
about my applying for a summer scholarship and working with him over three
months at the end of 2005. I departed
for Australia in December 2005, not
knowing that would be the start of a
ten-year journey outside Germany.
My main research task was to find sur

viving companion stars of these objects
in ancient supernova remnants. Over
the course of my PhD I tried and failed to
locate these companions in three different ancient remnants (SN 1006, SN 1572
and SN 1604), thus contributing to an
important result Type Ia supernovae
likely work differently than described in
current astronomy text books.
As my first task, Brian instructed me to

work on a tool that would help the
observing astronomers at the 2.3-metre
telescope in Siding Spring Observatory
select the right instrument set-up when
a new gamma-ray burst (GRB) alert triggered a Target of Opportunity observation. With my training in physics, I also
started attending astronomy courses for
the first time and learned about the rather
strange units that are commonplace in
astrophysics.
For my second project, I tried to auto

matically fit the spectrum of Type Ia
supernovae with a sophisticated radiative
transfer code. This technique would
allow us to reconstruct the structure of a
supernova by analysing subsequent
spectra. This meant optimising a function
in a high-dimensional and complex
search space. After some initial failures
using the techniques I was taught in my
physics classes, I started working with
a friend of mine in computational neurobiology who taught me several natureinspired optimisation algorithms that are
ideally suited for such problems, but
rarely used in astrophysics. This collaboration introduced me to the power of
interdisciplinary work, that to this day is
one the core themes of my research.
After three months, I was still excited

about working in astronomy and decided
to extend my stay for a further three
months. During this time, the then PhD
student Anna Frebel told me about an
opportunity to switch from a Masters in
physics to a PhD in astrophysics at Mt
Stromlo Observatory. With the help of
Brian, Rainer Wehrse and Penny Sackett
(then director of Mt Stromlo Observatory),
I applied for a PhD at the Australian
National University (the parent university
of Mt Stromlo). I started this PhD in
February 2007 with Brian as my primary
supervisor on Type Ia supernovae.
After finishing my PhD in the middle of

2011, I started as a postdoc working with
Marten van Kerkwijk at the University of
Toronto. Marten was not simply a boss,
but a mentor who guided my independent research. During this time, I started
working on my own radiative transfer
code for Type Ia supernovae (having used
Astronomical News
someone elses during my PhD). This

was the start of the TARDIS collaboration1 that I am now leading. TARDIS is
built to be a modular code that allows us
to experiment with different physical
approximations, built on modern software
development techniques and uniting
researchers with very different backgrounds (computer science and statistics),
to work on an astrophysical problem.
Part of this success was through a programme called the Google Summer
of Code that allowed me to recruit smart
and enthusiastic undergraduates from
several fields of study to work with us for

three months.
At the beginning of 2014, I heard the
great news that I had been selected for
an ESO Fellowship. This would not only
allow me to build on my current research,
but also start close collaborations with
the supernova researchers at ESO. I
started my fellowship at the end of 2014
and was very quickly introduced to my
duties: testing new algorithms to obtain
better results in data processing and
analysis.
Now almost one and a half years into my

fellowship, I have expanded on my
research of analysing Type Ia supernova
spectra with several results to be published soon. The ESO Fellowship is an
exciting step forward for me in my career
and Im keen to see what the second half
of it will bring.
Links
1
ARDIS spectrum synthesis code: http://tardis.

T
readthedocs.org
Personnel Movements
Arrivals (1 April30 June 2016)
Departures (1 April30 June 2016)
Europe
Europe
Bhardwaj, Anupam (IN)

Student
Cheffot, Anne-Laure (FR)
Student
Palla, Federica (IT) Assistant to the Data Reduction
Manager
Puglisi, Annagrazia (IT)
Student
Grunhut, Jason Harley (CA)

Instrument Scientist
Karabal, Muhammet Emin (TR)
Student
Noethe, Lothar (DE) Head of the Optical Engineering
Department
Chile
Chile
Acua, Andrea (CL)

Bartlett, Elizabeth (UK)
Iglesias, Daniela (CL)
Prez Snchez, Andrs Felipe (CO)
Sbordone, Luca (IT)
Executive Bilingual Secretary

Fellow
Student
Fellow
Operation Staff Astronomer
Acua, Margarita (CL)

Administrative Technician
Gonzalez, Oscar (CL)
Fellow
Marsset, Michal (FR)
Student
Muzic, Koraljka (HR)
Fellow
Rodriguez, Paula Valentina (CL) Team Leader Public Outreach
Vitacura
Image on page 59: Artists rendering of the design of

the telescope and dome of the European Extremely
Large Telescope (E-ELT). The ACe Consortium,
c onsisting of Astaldi, Cimolai and the nominated
sub-contractor EIE Group, signed a contract with
ESO for the construction of the dome and telescope
structure on 25 May 2016. See Release eso1617
for more information.
57
Astronomical News
ESO
European Organisation
for Astronomical
Research in the
Southern Hemisphere
ESO Fellowship Programme 2016/2017

The European Organisation for Astronomical Research in the Southern
H emisphere awards several postdoctoral fellowships each year. The goal
of these fellowships is to offer outstanding early-career scientists the opportunity to further develop their independent research programmes. ESO
Fellows return to the scientific community with the insights and experience
gained from working at one of the worlds foremost observatories.
The programme is open to applicants who will have achieved their PhD in
astronomy, physics or a related discipline before 1 November 2017. Earlycareer scientists from all astrophysical fields are welcome to apply. While
scientific excellence is the primary selection criterion for all fellowships,
c andidates should also explain (in their motivation letter) how ESOs facilities
and their work at ESO would facilitate their scientific development.
ESO is the leading intergovernmental astronomy organisation in Europe. Its

approximately 110 staff astronomers, 40 fellows and 40 PhD students conduct
frontline research in fields ranging from exoplanets to cosmology, offering one
of the most vibrant and stimulating scientific settings anywhere in the world.
We offer an attractive remuneration package including a competitive salary

and allowances (tax-free), comprehensive social benefits, and we provide
financial support for relocating families.
Fellowships are available both at ESOs Headquarters in Garching near

Munich, Germany, and at ESOs astronomy centre in Santiago, Chile.
ESO Headquarters is situated in one of the most active research centres in
Europe, boasting one of the highest concentrations of astronomers. ESOs
offices are adjacent to the Max Planck Institutes for Astrophysics and for
Extraterrestrial Physics and close to the observatory of Munichs LudwigMaximilian University. Additionally, ESO participates in the Excellence Cluster
Universe at the Garching campus, which brings together nearly 200 scientists
to explore the origin and structure of the Universe. Consequently, ESO Fellows
in Garching have many opportunities to interact and collaborate with astronomers at neighbouring institutes.
In Chile, fellows have the opportunity to collaborate with the rapidly growing
Chilean astronomical community as well as with astronomers at other international observatories located in Chile. The ALMA building next to ESOs
Santiago offices and the many astronomers and fellows working on the ALMA
project further enhance the stimulating scientific environment available to
ESO Chile Fellows.
At both sites, ESO Fellows are expected to actively participate in ESOs scientific life, for instance, by proposing and getting involved in the organisation
of scientific workshops, co-supervising PhD students, coordinating thematic
research groups, joining scientific committees, organising seminars, etc.
The fellowships in Garching start with an initial contract of one year followed
by a two-year extension (three years total). In addition to developing their
independent research programmes, ESO Garching Fellows will be expected
to engage in some functional work, for up to 25 % of their time, related to, for
example, instrumentation, the VLT/I, ALMA, APEX, E-ELT, science operations
support either in Garching or at one of ESOs observatories in Chile, software
development, or public outreach. This provides the fellows with the opportunity to get involved with ESO projects or operations, and to gather valuable
insights and experience not available in any other setting.
The fellowships in Chile are granted for one year initially, with annual extensions
for three additional years (four years total). During the first three years, the fellows are assigned to one of the science operations groups of Paranal, ALMA or
APEX, where they will contribute to the operations at a level of 80 nights per
year. For ALMA fellows a fraction of their duties can alternatively b
e spent on
data processing, participation in the ALMA review process as technical ex
perts, software testing and optimisation and extension of the array capabilities.
During the fourth year of Chile fellowships several options are provided. The
fellow may choose to spend the fourth year either at ESOs astronomy centre
in Santiago, or at ESO Headquarters in Garching or at any astronomy/astrophysics institute in an ESO Member State. There are no functional duties during the fourth year, except when the fourth year is spent at ESO Chile, where
fellows are expected to carry out functional work for up to 25 % of their time.
Under certain conditions, the fellow may also be hosted by a Chilean institution where she/he will be eligible to apply for time on all telescopes in Chile
through competition for Chilean observing time.
58
If you are interested in enhancing your early career through an ESO Fellowship, then please apply by completing the web application form available at:
http://jobs.eso.org
Please include the following documents in your application:
a cover/motivation letter;
a curriculum vitae with a list of publications;
a proposed research plan (maximum of two pages);
a brief outline of your technical/observational experience (maximum of one
page);
the names and contact details of three persons familiar with your scientific
work and willing to provide a recommendation letter. Referees will be automatically invited to submit a recommendation letter. However, applicants
are strongly advised to trigger these invitations (using the web application
form) well in advance of the application deadline.
The closing date for applications is 15 October 2016. Review of the application documents, including the recommendation letters, will begin immediately. Incomplete or late applications will not be considered.
Candidates will be notified of the results of the selection process between
December 2016 and February 2017. Fellowships will begin in the second half
of 2017.
Further information
For more information about the fellowship programme and ESOs astronomical research activities, please see:
http://www.eso.org/sci/activities/FeSt-overview/ESOfellowship.html
Details on the application procedure and the working conditions are
explained in the FAQs.
For a list of current ESO staff and fellows, and their research interests please
see: http://www.eso.org/sci/activities/personnel.html
Details of the Terms of Service for fellows including details of remuneration
are available at: http://www.eso.org/public/jobs/conditions/fellows/
For any additional questions please contact:
For Garching: Eric Emsellem, Tel. +49 89 3200 6914,
email: eric.emsellem@eso.org
For Chile: Claudio De Figueiredo Melo, Tel. +56 2 463 3032,
email: cmelo@eso.org
Although recruitment preference will be given to nationals of ESO Member
States (members are: Austria, Belgium, Brazil, the Czech Republic, Denmark,
Finland, France, Germany, Italy, the Netherlands, Poland, Portugal, Spain,
Sweden, Switzerland and United Kingdom) no nationality is in principle
excluded.
The post is equally open to suitably qualified female and male applicants.
ESO/L. Calada/ACe Consortium
ESO, the European Southern Observatory, is the foremost intergovernmental

astronomy organisation in Europe. It
is supported by 16 countries: Austria,
Belgium, Brazil, the Czech Republic,
Denmark, France, Finland, Germany,
Italy, the Netherlands, Poland, Portugal,
Spain, Sweden, Switzerland and the
United Kingdom. ESOs programme is
focused on the design, construction
and operation of powerful groundbased observing facilities. ESO operates three observatories in Chile: at
La Silla, at Paranal, site of the Very
Large Telescope, and at Llano de
Chajnantor. ESO is the European partner
in the Atacama Large Millimeter/submillimeter Array (ALMA).Currently ESO
is engaged in the construction of the
European Extremely Large T
elescope.
The Messenger is published, in hardcopy and electronic form, four times
a year: in March, June, September and
December. ESO produces and distributes a wide variety of media c
onnected
to its activities. For further information,
including postal subscription to The
Messenger, contact the ESO education
and Public Outreach Department at:
ESO Headquarters
Karl-Schwarzschild-Strae 2
85748 Garching bei Mnchen, Germany
Phone +49 89 320 06-0
information@eso.org
Contents
Arsenault R. et al. Adaptive Optics Facility Status Report:
When First Light Is Produced Rather Than Captured
Fourniol N. et al. A Fruitful Collaboration between ESO and the
Max Planck Computing and Data Facility
Martayan C. et al. Solar Activity-driven Variability of
Instrumental Data Quality
Asmus D. et al. Science Verification for the VISIR Upgrade
Kamann S. et al. A Stellar Census in NGC 6397 with MUSE
Randall S. K. et al. Pulsating Hot Subdwarfs in Omega Centauri
Adami C. et al. First Results from the XXL Survey and Associated
Multi-wavelength Programmes
van der Wel A. et al. The LEGA-C Survey: The Physics of Galaxies
7 Gyr Ago
Oteo I. et al. ALMACAL: Exploiting ALMA Calibrator Scans to
Carry Out a Deep and Wide (Sub)millimetre Survey,
Free of Cosmic Variance
Astronomical News
Horlek P. et al. Light Phenomena over the ESO Observatories III:
Zodiacal Light
Dennefeld M. et al. The First NEON School in La Silla
Ballester P. ESO Data Simulation Workshop
Spyromilio J., Holzlhner R. Retirement of Lothar Noethe
Fellows at ESO K. Immer, E. Johnston, W. Kerzendorf
Personnel Movements
ESO Fellowship Programme 2016/2017
The Messenger:
Editor: Jeremy R. Walsh;
Design, Production: Jutta Boxheimer;
L ayout, Typesetting: Mafalda Martins;
Graphics: J
oanna Law.
www.eso.org/messenger/
Printed by FIBO Druck und
Verlags GmbH, Fichtenstrae 8,
82061Neuried, Germany
Unless otherwise indicated, all images
in The Messenger are courtesy of ESO,
except authored contributions which
are courtesy of the respective authors.
ESO 2016
ISSN 0722-6691
Front cover: First light for the 4Laser Guide Star System
(4LGSF) at the Very Large Telescope occurred on 26 April
2016. The image shows the four 22 watt Raman laser
beams (at the frequency of the Na D 2 line 5896 ) propagated from Unit Telescope 4 into the sky in the plane
of the Milky Way. The 4LGSF is an integral part of the
Adaptive Optics Facility (AOF) that is being gradually commissioned. Further details can in found in the article by
A rsenault et al. (p. 2) and in Release eso1613.
Credit: ESO/F. Kamphues
2
8
10
14
18
23
27
36
41
45
47
50
52
54
57
58

The Messenger 164

Uploaded by

Document Informationclick to expand document informationThe Messenger is published by the European Southern Observatory (ESO), the foremost intergovernmental astronomy organisation in Europe, in hardcopy and electronic form, four times a year: in March, June, September and December.

Document Informationclick to expand document information

Copyright:

Available Formats

The Messenger 164

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

The Messenger 164

Uploaded by

Copyright:

Available Formats

Solar activity and instrument data quality

VISIR Upgrade Science Verification

Telescopes and Instrumentation

Adaptive Optics Facility Status Report: When First Light

First light for the 4 Laser Guide Star

ESO Project Team: