Professional Documents
Culture Documents
© ESO 2023
ISSN 0722-6691
Front Cover: Chile has hosted ESO’s telescopes
since the 1960s, in observatories based at La Silla,
The Messenger and all articles are Paranal, and Chajnantor Plateau. Shown here is the
published open access under a Creative Visible and Infrared Survey Telescope for Astronomy
Commons Attribution 4.0 International (VISTA), situated at Paranal Observatory. Perched
License. atop a mountain adjacent to Cerro Paranal, the
home of the flagship Very Large Telescope (VLT),
VISTA is the largest telescope in the world designed
to survey the sky in near-infrared light. Credit:
ESO/B. Tafreshi (twanight.org)
Vincenzo Mainieri 1 high multiplex capability, with 1624 fibres (PSP) composed of scientific experts
Bruno Leibundgut 1 feeding two low-resolution spectrographs from the ESO Community. A subset of
Roelof S. de Jong 2 (R = l/Dl ~ 6500), and 812 fibres trans- the proposing teams was invited to sub-
J. Miguel Mas-Hesse 3 ferring light to the high-resolution spec- mit full proposals by December 2020,
Jochen Liske 4 trograph (R ~ 20 000). It was decided which were again evaluated by the PSP.
Karin Lind 5 from the start to use the instrument in a Finally, considering the scientific merit of
Jon Loveday 6 survey mode only. The 4MOST consor- the proposals and the technical boundary
tium prepared ten survey proposals (see conditions of 4MOST, the PSP recom-
the papers in issue 175 of The Messen- mended the implementation of 15 Com-
1
ESO ger) spanning the whole sky accessible munity Surveys. The selection was
2
Leibniz Institute for Astrophysics, from Paranal and addressing many sci- endorsed by the Observing Programmes
Potsdam, Germany ence goals. The surveys were presented Committee at its meeting in November
3
Department of Astrophysics, Centre for at a dedicated workshop entitled Prepar- 2021.
Astrobiology (CSIC–INTA), Torrejón de ing for 4MOST: A community workshop
Ardoz, Spain introducing ESO’s next generation spec- The selected Community Surveys have
4
Hamburg Observatory, University of troscopy survey facility, held in May 20191. now joined the ten Consortium Surveys
Hamburg, Germany to form the final 4MOST survey pro-
5
Department of Physics and Astronomy, The 4MOST consortium will receive 70% gramme for the first five years of operation.
Uppsala University, Sweden of the available fibre-hours in the first five This joint set of surveys covers a very
6
University of Sussex, Brighton, UK years of operation in return for building wide range of Galactic and extragalactic
the facility and supporting ESO with its science cases. This issue of the Messenger
operation. The remaining 30% of the contains a detailed description of the
The 4-metre Multi-Object Spectroscopic observing time in the first five years is selected Community Surveys, which com-
Telescope (4MOST) is a new high- available to conduct surveys led by the plements the description of the Consortium
multiplex, wide-field spectroscopic survey ESO Community. The selection of these Surveys in Messenger 175.
facility under construction for the four- Community Surveys went through the
metre-class Visible and Infrared Survey two-step process for ESO Public Surveys.
Telescope for Astronomy (VISTA) at A Call for Letters of Intent was issued in Links
Paranal. Its key specifications are: a large 2019. The submitted science cases were 1
Preparing for 4MOST workshop:
field of view of 4.2 square degrees and a reviewed by the Public Survey Panel https://www.eso.org/sci/meetings/2019/4MOST.html
P. Horálek/ESO
Since construction of the ESO Residencia was com- Mars-like Atacama Desert, just a few kilometres
pleted in 2002, it has been a home from home for from Cerro Paranal — the mountain that hosts ESO’s
the astronomers, engineers and technicians working Very Large Telescope (VLT) and Visible and Infrared
at ESO’s Paranal Observatory in Chile. The sleek Survey Telescope for Astronomy (VISTA).
building sits 2400 metres above sea level in the
Odette Toloza 1 17
Leibniz Institute for Astrophysics, lead to a common-envelope phase and to
Alberto Rebassa-Mansergas 2 Potsdam, Germany a dramatic shrinkage of the orbit (Ivanova
Roberto Raddi 2 18
University of Washington, USA et al., 2013). These close WD+MS binaries
Nicole Reindl 3 19
University of Amsterdam, the have orbital periods ranging from a few
Boris Gaensicke 4 Netherlands hours to several days and the companions
Nicola Gentile Fusillo 5 20
University of Valparaíso, Chile sample a wide range of spectral types
Simone Scaringi 6 (A, F, G, K, M). Hence, they hold the poten-
Diogo Belloni 1 tial to give rise to a wide range of impor-
Elme Breedt 7 Binary systems containing one or two tant outcomes, for example type Ia super-
Maria Camisassa 2 white dwarfs are important for studying novae (SN Ia), gravitational wave sources,
Tim Cunningham 4 stellar evolution and interactions under a and WD pulsars, or themselves being out-
Domitila de Martino 8 wide range of astrophysical conditions. comes of fascinating phenomena, such as
Alessandro Ederoclite 9 Thanks to the Gaia mission we have planetary nebulae. Indeed, binaries are
Stephan Geier 3 identified the first statistically significant overrepresented in the central star plane-
Matthew Green 10 and unbiased sample of ~ 170 000 white tary nebula population (Jones & Boffin,
Keith Inight 4 dwarf binary candidates. It comprises 2017). In particular, the bipolar structures
Thomas Kupfer 11 both systems that never interacted that in planetary nebulae cannot be explained
Jesús Maldonado 12 are part of common proper motion pairs by single-star evolution. The WD binary
Tom Marsh 4 and systems that evolved through mass survey aims to resolve a wide variety of
Anna Francesca Pala 5 transfer episodes resulting in close important open problems which will ulti-
Steven Parsons 13 white dwarf binaries. White dwarf bina- mately help to understand the evolution
Ingrid Pelisoli 4 ries hold the potential to observationally of close and wide WD binaries.
Juanjuan Ren 14 constrain a wide variety of key ingredi-
Pablo Rodriguez-Gil 15 ents that currently limit the validity of
Snehalata Sahu 4 theoretical models. Specific scientific goals
Linda Schmidtobreick 16
Matthias Schreiber 1 The 4MOST WD binary survey is divided
Axel Schwope 17 Scientific context into three sub-surveys, as outlined below,
Danny Steeghs 4 each aiming for different goals.
Paula Szkody 18 Over 90 per cent of stars are born with
Silvia Toonen 19 masses < 8−10 M☉, with the certain fate The compact white dwarf binary
Pier-Emmanuel Tremblay 4 of ending up as white dwarfs (WDs). (CWDB) sub-survey. Within this sub-
Monica Zorotovic 20 Moreover, about half of the main sequence survey we aim to address important
(MS) stars in the Milky way are found in aspects of close WD evolution such as
binary systems. The binary separation the mechanisms of angular momentum
1
Federico Santa Maria Technical determines the evolution of these sys- loss, the energy sources during the
University, Valparaiso, Chile tems. If the binary components are sepa- common-envelope phase, understanding
2
Catalunya Polytechnic University, rated enough to avoid mass transfer the formation of magnetic WDs in close
Barcelona, Spain episodes (≥ 10 au), they can evolve as binaries, identifying the progenitors of AM
3
Potsdam University, Germany single stars. In these binaries, the more CVn stars, and determining the dominant
4
University of Warwick, Coventry, UK massive star evolves first through the fate of close WD binaries. With 4MOST,
5
ESA t ypical nuclear burning phases and ends the current samples of the distinct types
6
Durham University, UK its life as a WD (García-Berro, Ritossa & of binaries will be significantly enlarged,
7
University of Cambridge, UK Iben, 1997). The orbital separations of without bias. The 4MOST spectra will pro-
8
INAF–Capodimonte Astronomical the resulting detached WD+MS binaries vide the effective temperatures, surface
Observatory, Naples, Italy widen with time as a result of modest gravities, radii, and masses that will allow
9
University of Sao Paolo Institute of velocity kicks caused by asymmetric mass the characterisation of the stellar compo-
Astronomy, Brazil loss of the WD progenitors (El-Badry & nents, while the orbital periods will be
10
Tel Aviv University, Israel Rix, 2018). In these cases, the WDs can measured from the photometric data from
11
University of California at Santa be used as reliable cosmo-chronometers Vera C. Rubin Observatory. These param-
Barbara, USA to measure stellar ages, a difficult endeav- eters will ultimately provide observational
12
INAF–Palermo Astronomical our that is otherwise subject to substan- constraints to the evolutionary models.
Observatory, Italy tial uncertainties (Soderblom, 2010).
13
University of Sheffield, UK These wide WD+MS binaries are excel- The common proper motion pairs
14
National Astronomical Observatories – lent probes of the evolution of age-related (CPMP) sub-survey. This sub-survey
Chaoyang Beijing, China changes in activity, rotation, and metal aims at using WDs as ‘clocks’ to provide
15
Canary Islands Institute of Astrophysics, licity of the MS companions. Conversely, additional observational input for
Tenerife, Spain for smaller initial separations (< 10 au), improving our understanding of the
16
ESO mass-transfer interactions ensue, which chemical evolution of the Milky Way and
Number of targets
15°
Dec.
0° LRS
HRS
10 2
–15°
–30°
–45°
101
–60°
–75°
RA
10 0
0 500 1000 11 13 15 17 19
Number of targets per 2 × 2 degree2 G (mag.)
Figure 1. Left: Density map of the target catalogue Right: Distribution of the white dwarf binary targets number of targets per sub-survey for each
in equatorial coordinates; the orange line represents for each sub-survey as a function of Gaia G magni-
spectral resolution is shown in Table 1.
the galactic plane which lacks GALEX data. We aim tude (0.1 mag. bin). Black and red lines show the
to follow up at least 41% and 95% of our catalogues total number of targets requested to be observed
that will be observed with the low-resolution and with the high-resolution (HRS) and low-resolution
high-resolution spectrographs, respectively. (LRS) spectrograph, respectively. Target selection and survey area
SSC: S/N for LRS SSC: S/N for HRS Number of Number of FoM for LRS FoM for HRS
targets for LRS targets for HRS
CWDB CaT or Ha or BLUE > 13 (up to 100 for G = 15) Ha > 150 (up to 500 for G = 10) 142 000 22 000 40% 95%
CPMP Ha_wing > 30 BLUE_narrow > 30 2800 900 70% 90%
CSPN Ha_wing > 50 Ha_wing > 250 670 70 71% 95%
Table 1. Relevant parameters for the three sub- median of the S/N per Ångstrom calculated assuming CaT = MEDIAN(S/N 8350−8850 Å). The low-resolution
surveys, compact white dwarf binary (CWDB), com- grey conditions. The S/N is calculated within a wave- spectra of the CWDB sub-survey utilise the combi-
mon proper motion pairs (CPMP), and central star length range as follows: BLUE = MEDIAN(S/N nation of three wavelength ranges. The fraction of
planetary nebulae (CSPN), for the low-resolution 4000−4500 Å); BLUE_narrow = MEDIAN(S/N observed targets is calculated for a figure of merit
spectra (LRS) and high-resolution spectra (HRS). 4000−4300 Å); Ha = MEDIAN(S/N 6512−6612 Å); (FoM) of 0.5 for each sub-survey.
The spectral success criteria (SSC) correspond to the Ha_wing = MEDIAN(S/N 6400−6500 Å);
catalogue with non-degenerate stars in emission and continuous emission from sub-surveys such that the success rate
the Gaia DR3, by adopting the selection an accretion disc can also be observed. is 0.5, which is the minimum requirement
and quality cuts defined by El Badry et al. The minimum requirement is an optical such that the scientific goals can be
(2021), Rebassa-Mansergas et al. (2021), spectrum with sufficient S/N to recognise achieved. However, the overall success of
and Raddi et al. (2022). The contaminants key features for classification of the type the WDB survey is dictated by a combined
from open clusters (Cantat-Gaudin et al., of binary (for example, cataclysmic varia- FoM, which is defined as the weighted
2018) are removed. Finally, the targets for bles, AM CVn stars, double WDs, post- average of all the individual FoMs.
the central star of planetary nebulae cata- common-envelope binaries or symbiotic
logue were selected from the catalogues binaries). The spectra with higher S/N
of González-Santamaría et al. (2021) and will be fitted with templates to determine Acknowledgements
Chornay & Walton (2021). In general, the stellar parameters. In the common proper ARM acknowledges support from the Spanish
combined catalogue comprises approxi- motion sub-survey, we aim to determine MINECO grant PID2020-117252GB-I00 and from the
mately 170 000 targets with brightnesses precise abundances, radial velocities, AGAUR grant SGR-386/2021. OT was supported by
within G = 10 and G = 20 mag. (see right and stellar parameters, as well as an FONDECYT project 321038.
panel of Figure 1). Most of the targets of estimate of stellar activity by measuring
all the sub-surveys cover the 4MOST foot- the flux of chromosperic emission lines References
print, i.e., sky declinations –80 < dec < +5 such as Ha and Ca H&K or the IR triplet.
(see left panel in Figure 1). Thus, we will acquire high-resolution Cantat-Gaudin, T. et al. 2018, A&A, 618, A93
Chornay, N. & Walton, N. A. 2021, A&A, 656, A110
and low-resolution spectroscopy for the El-Badry, K. & Rix, H.-W. 2018, MNRAS, 480, 4884
F/G/K stars and M-type companions, El-Badry, K., Rix, H.-W. & Heintz, T. M. 2021,
Spectral success criterion and figure respectively. Finally, for the central star MNRAS, 506, 2269
of merit planetary nebula sub-survey it is impor- García-Berro, E., Ritossa, C. & Iben, I. 1997, ApJ,
485, 765
tant to avoid any (strong) nebular or photo- Gentile-Fusillo, N. P. et al. 2021, MNRAS, 508, 3877
In general, we utilise the signal to noise spheric lines when calculating the S/N González-Santamaría, I. et al. 2021, A&A, 656, A51
ratio (S/N) per Ångstrom as the spectral such that spectral analysis can be per- Ivanova, N. et al. 2013, A&ARv, 21, 59
success criterion. The spectral energy formed. Our S/N requirements for each Jones, D. & Boffin, H. M. J. 2017, Nature Astronomy,
1, 0117
distribution of close WD binaries results sub-survey are listed in Table1. Parsons, S. G. et al. 2016, MNRAS, 463, 2125
in a composite spectrum, which is either Raddi, R. et al. 2022, A&A, 658, A22
dominated by the WD or by the non-de- The figure of merit (FoM) measures the Rebassa-Mansergas, A. et al. 2021, MNRAS,
generate companion, depending on the scientific success of each sub-survey. 506, 5201
Soderblom, D. R. 2010, ARA&A, 48, 581
temperature and radius of either of them. Table 1 shows the fraction of targets that
In the case of interacting binaries, line need to be observed for each of the
Y. Beletsky (LCO)/ESO
Located deep in the Chilean Atacama Desert, far technicolour in this image. The striking radiant light
from the light pollution associated with human activ- visible in the sky here is a phenomenon called
ity, ESO’s Paranal Observatory enjoys some of the a irglow, which is created as atoms and molecules in
darkest skies on Earth. Paradoxically, it is this the atmosphere combine and emit radiation.
extreme darkness that allows the sky to light up in
G. Germano Sacco 1 7
ESO inhomogeneity in the local disc; vastly
Rob Jeffries 2 8
Department of Physics, Tor Vergata expand the number of identified young
Alex Binks 3 University of Rome, Italy stars for exoplanetary studies; and pro-
Laura Magrini 1 9
Department of Physics and Astronomy, vide huge coeval samples to improve
Francesco Damiani 4 University of Padua, Italy young stellar evolutionary models.
Nicholas Wright 2 10
INAF–Rome Astronomical Observatory,
Loredana Prisinzano 4 Italy
Eleonora Zari 5 11
Institute for Astronomy & Astrophysics, Scientific context
Christian Schneider 6 Eberhard Karls University, Tübingen,
Giacomo Beccari 7 Germany The distribution of young stars in space,
George Weaver 2 12
Andalucia Institute of Astrophysics kinematics, chemistry and age is of key
Valentina D’Orazi 8,9 (CSIC), Granada, Spain importance in understanding the pro-
Katia Biazzo 10 13
Department of Astrophysics, University cesses that drive star formation, the ori-
Brunella Nisini 10 of Vienna, Austria gins of the Galactic field population, and
Simone Antoniucci 10 14
Institute of Physics and Astronomy, the early evolution of stars, their discs
Elena Franciosini 1 Faculty of Science, University of and planetary systems. Most stars
Jan Robrade 6 Valparaíso, Chile (including the Sun; Adams, 2010) are
Beate Stelzer 11 15
Bordeaux Astrophysics Laboratory, born in clusters of various sizes in the
Emilio J. Alfaro 12 University of Bordeaux, CNRS, Pessac, Galactic disc, but most of these clusters
João Alves 13 France disperse into the field on timescales of
Amelia Bayo 7,14 16
Leiden Observatory, Leiden University, 10–100 Myr (Lada & Lada, 2003;
Henri Boffin 7 the Netherlands K rumholz, McKee & Bland-Hawthorn,
Rosaria Bonito 4 17
Enrico Fermi Department of Physics, 2019). Most young stars (with ages
Hervé Bouy 15 University of Pisa, Italy < 100 Myr) in the Galactic disc, including
Anthony Brown 16 18
Rio Tinto Alcan Planetarium, Space For the nearest ones, are not located in com-
Edvige Corbelli 1 Life, Montreal, Canada pact clusters and star-forming regions
Scilla Degl’Innocenti 17 19
Institute for Research on Exoplanets, (for example, Zari et al., 2018); they are
Davide Fedele 1 Department of Physics, University of instead part of the field population. Dec-
Jonathan Gagné 18,19 Montreal, Canada ades of work have revealed a population
Philip Galli 20 20
Centre for Theoretical Astrophysics, of about 1500 young stars within 100 pc
Josefa Großschedl 13 São Paulo City University, Brazil of the Sun, a volume that contains no sig-
Tereza Jerabkova 7 21
School of Physics and Astronomy, nificant clusters (for example, Gagné &
Joel Kastner 21 Rochester Institute of Technology, NY, Faherty, 2018). Many, but not all, are part
Daisuke Kawata 22 USA of kinematically coherent, roughly coeval
Stefan Meingast 13 22
Mullard Space Science Laboratory, ‘moving groups’ of large spatial extent
Núria Miret-Roig 13 University College London, UK (for example, Torres et al., 2006). Mem-
Estelle Moraux 23 23
University of Grenoble Alpes, CNRS, bers of this widespread population trace
Javier Olivares 24 IPAG, France the recent star formation history in the
Richard J. Parker 25 24
National University for Distance Milky Way disc on a variety of scales and,
Pier Giorgio Prada Moroni 17 Learning, Spain because of their proximity, have become
Timo Prusti 26 25
Department of Physics and Astronomy, exemplars for studying early stellar evolu-
Sofia Randich 1 University of Sheffield, UK tion and planet formation; yet their origins
Veronica Roccatagliata 1 26
European Space Agency, European and birth environments remain obscure.
Lorenzo Spina 9 Space Research and Technology While the census of young stars within
Centre (ESTEC), the Netherlands 100 pc is substantially complete, much
less is known about their chemistry or the
1
INAF–Arcetri Astronomical spatial and kinematic properties of young
Observatory, Florence, Italy Most nearby young stars (with ages stars outside this radius, beyond the
2
Astrophysics Group, Keele < 100 Myr) in the Galactic disc no longer biased perspective offered by high-den-
University, UK reside in their dense, clustered birth- sity compact clusters and star-forming
3
MIT Kavli Institute for Astrophysics and places; they are found all around us. The regions. The Gaia astrometry satellite1 is
Space Research, Massachusetts Institute 4MOST Survey of Young Stars will iden- revolutionising this field (for example,
of Technology, Cambridge, MA, USA tify a representative sample of about 105 Kounkel & Covey, 2019; Cantat-Gaudin,
4
INAF–Palermo Astronomical young, low-mass stars within 500 pc of 2022; Alfaro et al., 2022; Prisinzano et al.,
Observatory, Italy the Sun and will measure their chemistry, 2022). Precise astrometry and homoge-
5
Max Planck Institute for Astronomy, 3D kinematics and ages in order to: neous photometry can place stars in
Heidelberg, Germany trace the spatial and dynamical evolution Hertzsprung–Russell diagrams and pro-
6
Hamburg Observatory, University of of star-forming structures; quantify the vide 2D tangential motions. Early studies
Hamburg, Germany star formation rate and chemical are providing tantalising glimpses of rich
Number of stars
(Meingast, Alves & Rottensteiner, 2021;
Gagné et al., 2021) and finding evidence
for kinematically coherent streams MG
arranged in grand filamentary structures 10
101
on scales > 100 pc (for example, Beccari,
Boffin & Jerabkova, 2020). 120 Myr
15 40 Myr
These studies will continue to be limited
10 0
in several important ways, however, until 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5
they are supplemented by spectroscopy (G BP – G RP) 0
of large and representative samples: (1) 4SYS input catalogue
samples selected by astrometry and pho-
tometry alone are heavily contaminated
with older stars; (2) astrometry yields only
2D kinematics, but 3D velocities are
important for determining the history and
Bulge/inner halo show that dark sub-halos with masses cut, from which we aim to select targets
of 10 9 M⊙ cause large convergent veloc- as uniformly as possible over the full sky
The RR Lyrae in the inner few kpc of the ity flows of ~ 15 km s –1 in the stellar halo accessible to 4MOST. Assuming approxi-
Milky Way are an extremely old stellar on scales out to ~ 2 kpc. To detect this mately 50% completeness in targeting,
population (13.41 ± 0.54 Gyr; Savino et signal, it is necessary to have good dis- up to 100 000 RR Lyrae stars may be
al., 2020). They form a bulge spheroid, tance measurements, in order to pick out observed.
the nature of which is still unclear — it the populations at the distance of interest
might be a small classical bulge, the (and so suppress contamination from
result of a puffed up disc, or it could be stars at different distances). Again, the Acknowledgements
the inner region of the halo (or a combi- excellent RR Lyrae distances are essen- R. E., B. F., R. I., N. M, G. M., L. P. and A. S.
nation of these). Using the largest sample tial, and their 4MOST radial velocities, acknowledge funding from the European Research
of RR Lyrae with radial velocities in the together with Gaia proper motions, will Council (ERC) under the European Unions Horizon
inner Galaxy to date (2768 stars), Kunder provide the velocity signal of interest. 2020 research and innovation programme (grant
agreement No. 834148).
et al. (2020) find some evidence for differ- With this, we aim to construct a three-di-
ent spatial and kinematic populations of mensional map of the velocity coherence
RR Lyrae in the inner Galaxy. With a sam- (its local convergence and divergence) References
ple of inner-Galaxy RR Lyrae that is more throughout the halo, and compare this
Clementini, G. et al. 2022, arXiv:2206.06278
than an order of magnitude larger than result to numerical simulations in LCDM Deason, A., Belokurov, V. & Evans, N. W. 2011,
currently available (~ 50 000 in the bulge and with other prescriptions for gravity/ MNRAS, 416, 2903
region), 4GRoundS will be able to accu- dark matter. This will allow 4GRoundS to Garavito-Camargo, N. et al. 2019, ApJ, 884, 51
rately disentangle the halo and spheroidal quantify the degree of lumpiness of the Hansen, C. J. et al. 2011, A&A, 527, A65
Ibata, R. et al. 2021, ApJ, 914, 123
bulge for the oldest populations in the halo in the full 6D phase space. We will Iorio, G. & Belokurov, V. 2021, MNRAS, 502, 5686
inner galaxy, constraining the early for- make use of robust estimators like, for Kunder, A. et al. 2020, AJ, 159, 270
mation of the Milky Way. This will provide example, the two-point correlation func- Laporte, C. F. P. et al. 2018, MNRAS, 481, 286
a complementary perspective to that of tion calculated in this extended space Muraveva, T. et al. 2018, MNRAS, 481, 1195
Petersen, M. S. & Peñarrubia, J. 2021, Nature
4MOST Consortium Survey 3. and compare them with the same quanti- Astronomy, 5, 251
ties measured in cosmological simula- Posti, L. & Helmi, A. 2019, A&A, 621, A56
tions and in smooth models of the halo to Savino, A. et al. 2020, A&A, 641, 96
Spatial variations in kinematic probe the validity of the currently Sesar, B. 2012, AJ, 144, 114
Sesar, B. et al. 2014, ApJ, 793, 135
coherence through the halo favoured cosmological paradigm. Stringer, K. M. et al. 2021, ApJ, 911, 109
Thomas, G. F. et al. 2019, ApJ, 886, 10
ΛCDM posits that all galaxies are con Wegg, C., Gerhard, O. & Bieth, M. 2019, MNRAS,
stituted of a myriad of dark sub-halos. Target selection and survey area 485, 3296
There is the hope of detecting these
invisible sub-halos via their heating effect The 4GRoundS input catalogue is Links
on star streams. Here we propose an simply the full Gaia DR3 sample of
1
alternative means to detect these dark 271 779 RR Lyrae (Clementini et al., WEAVE instrument overview: https://www.ing.iac.
es/astronomy/instruments/weave/weaveinst.html
matter substructures. Our simulations 2022), without any apparent magnitude
ESO/VVVX survey
evolution of the MW and galaxies in gen- MW formation. Clusters can be accu- MCs, including a few newly identified
eral. These issues are set out below. rately dated and hence their study allows bulge GCs. Elemental abundances
us to place further constraints on Galactic probing all nucleosynthetic channels
Cluster formation and disruption. Stars formation, evolution, and structure, com- (Fe-peak, light, α, p- and n-capture) will
are thought to form in associations and plementing the study of the field popula- be obtained from high-resolution (HR)
clusters, the vast majority of which dis- tion. The study of the chemo-dynamical spectra for a few tens of stars per clus-
perse because of dynamical and stellar properties of stellar clusters provides ter (the actual number depending on
evolutionary effects, such as supernova important information on the merger distance, reddening and concentration)
feedback. Studying clusters is therefore a events involved in the assembly history of for ~ 5 000 stars with G < 16.5. As con-
key ingredient in the understanding of our galaxy. The age-metallicity relation of cerns the study of multiple populations,
both star formation and how they populate the MW disc, its radial, azimuthal and ver- we will obtain precise abundances of
the Galactic components through disrup- tical metallicity distributions, and their evo- O, Na, Mg and Al, i.e., of the elements
tive processes. It is currently still unclear lution with time can be studied through defining the classical (anti-)correlations
whether OCs are the dominant source of OCs. The chemo-dynamic study of GCs between light elements; this cannot be
stars in the disc, what the role of GCs in can answer several major open questions done using photometry, which traces
building up the Galactic bulge and halo about the Galactic halo and bulge, such essentially just N variations (Gratton et
is, and how the multiple populations, a as how they came into place, and can al., 2019). RVs, atmospheric parame-
phenomenon observed in virtually all help probe the evolution of the MCs and ters, and a few key elements (for exam-
Galactic GCs, form. Chemical information the role of interactions with the MW. ple α’s) will be determined from low-
and radial velocities (RVs) for a large sam- resolution (LR) spectra for ~ 35 000 stars
ple of clusters, along with Gaia astrome- with 16.5 < G < 20 (several tens per
try and precise and independent ages Specific scientific goals cluster) in 140 GCs. This will be by far
from asteroseismology, will be necessary the largest sample of homogeneously
to make headways on these matters. The 4MOST Stellar Clusters Survey aims derived detailed compositions and RVs
to obtain spectra of the largesta uniformly for GC stars.
Stellar evolution. Our understanding of selected and analysed sample of Galactic – Derive chemical abundances (at 0.05
stellar populations ultimately relies on clusters (and a few MC GCs) and very to 0.1 dex precision) in all the nucleo-
stellar models: stellar age dating is nec- young clusters (VYCs) in star-forming synthetic channels for the vast majority
essary to investigate galactic formation regions (SFRs), covering metallicities from of known OCs accessible to 4MOST.
and evolution and processes such as the [Fe/H] = −2.5 dex of GCs to super- We expect to study ~ 30 000 stars at
mass loss and feedback are crucial ingre- solar OCs and ages from a few Myr to HR (G < 14.5 for most clusters, with
dients in probing the chemical evolution 13.5 Gyr. We will derive chemical charac- G < 15.5 in a few tens of clusters,
of galaxies. However, the physics of many terisation, probe cluster internal kinemat- selected to study details of the compo-
processes (accretion, convection, rota- ics up to the outskirts, examine dissolving sition down to the main sequence) and
tion, magnetic field, atomic diffusion, mix- (and dissolved) clusters, and probe the ~ 75 000 at LR (14.5 < G < 19). Again,
ing, activity, stellar winds, mass loss) is presence and nature of halos and tidal this will be by far the largest sample
uncertain and is treated in different, sim- tails. Considering the expected perfor- of homogeneously derived chemical
plified ways in stellar models. The large, mance of 4MOST, in the time allocated to abundances and RVs for OCs.
homogeneously selected and analysed our survey we expect to reach the follow- – Derive the dynamics, study activity indi-
samples of clusters at different ages and ing goals: cators and probe the composition of
chemical compositions we will collect will ~ 80 VYCs in nearby SFRs, where stars
contribute to considerable advancements –D
erive homogeneous, full chemo- are still in the pre-main-sequence
in detailed stellar modeling, providing dynamical characterisation (at 0.05 to phase (PMS), and study low-mass PMS
empirical isochrones and calibration of 0.1 dex precision) for an unprecedented stars from already dissolved VYCs.
the associated physics. sample of ~ 120 GCs in the MW and
surveys, while ensuring the maximisation were selected to have 10 < G < 14.5, chemo-dynamical tagging in the age of large sur-
veys; PI S. Lucatello). This research has been partly
of the number of successfully observed with the exception of a few tens of OCs,
funded by the Spanish MICIN/AEI/10.13039/
targets. Estimates of the number of stars chosen to study the detailed chemistry 501100011033 and by ERDF A way of making
we will successfully observe, based on down to the main sequence, where we Europe by the European Union through grant
simulations performed within the 4MOST selected targets with 10 < G < 15.5. LR RTI2018-095076-B-C21 and PID2021-122842OB-
C21, and the Institute of Cosmos Sciences University
collaboration, are listed in Table 1, along targets span from the faint end of the HR
of Barcelona (ICCUB, Unidad de Excelencia ‘Maria de
with a few key facts about targets for the to G = 19. Stars bright enough for HR, Maeztu’) through grant CEX2019-000918-M.
six sub-surveys. More details of the criteria but with Gaia DR3 broadening parameter
for the different classes of objects in our vbroad > 30 km s –1, have been included in
References
sample are given below along with a visual the LR rather than HR sample, as for
representation of the input catalogue tar- objects of such rotational velocity HR Cantat-Gaudin, T. et al. 2020, A&A, 640, A1
get distribution in Figure 1. does not allow to derive additional infor- Dias, W. S. et al. 2002, A&A, 389, 871
mation over LR. The HR sample includes Gaia Collaboration et al. 2023, A&A, in
press, arXiv:2206.05534
GCs. Our aim is to target all visible GCs, red-clump stars out to distances of ~ 10
Gratton, R. et al. 2019, A&ARev., 27, 8
from centre to outskirts (comprising halos kpc, and the LR sample includes faint K Massari, D., Koppelman, H. H. & Helmi, A. 2019,
and tails). Targets are selected on Gaia stars within 6 kpc. A&A, 630, L4
DR3 astrometry (for example, Vasiliev & Pfeffer, J. et al. 2018, MNRAS, 475, 4309
Vasiliev, E. & Baumgardt, H. 2021, MNRAS, 505,
Baumgardt, 2021) and colour–magnitude VYCs. We selected targets from the cata-
5978
diagrams. To allow for an easier alloca- logues of Dias et al. (2002, and revised
tion, LR targets were selected to be out- online in 20151) and Cantat-Gaudin et al.
side a few core radii from the centre of (2020), including all VYCs in the 4MOST Links
the clusters, where the bulk of HR stars footprint for which a signal-to-noise ratio 1
Revised catalogue of optically visible open clusters
are located. Magnitude ranges for LR and of about 40 Å –1 could be reached in 1 hour (after Dias et al., 2002): http://cdsarc.u-strasbg.fr/
HR targets are reported in Table 1. All of exposure time in HR. This includes viz-bin/qcat?J/A+A/389/871
selected targets are of FGK spectral type; objects within 2.5 kpc of the Sun, with
(almost) all HR targets are giants, while reddening E(B-V) < 2.5, and ages 2–20 Myr.
Notes
both giants and dwarfs are observed in Magnitude ranges for LR and HR targets
LR mode. are reported in Table 1. Stars brighter a
Actually, the Gaia RVS instrument is producing the
than G = 15, but with Gaia DR3 broaden- largest database of intermediate resolution spectra,
OCs. We plan to observe all OCs acces- ing parameter v broad > 30 km s –1, have providing RVs, metallicity and a set of chemical
abundances for ten of millions stars; some of those
sible to 4MOST, sampling from the centre been included in the LR rather than the belong to stellar clusters (see Gaia Collaboration et
to the outskirts. The original selection HR sample. al., 2023, where the metallicity of about 600 open
based on Gaia DR2 membership (based, clusters has been derived from an average of three
for example, on Cantat-Gaudin et al., 2020) member stars). However, the magnitude limit for
Acknowledgements Gaia spectroscopy is bright (implying that GC stars
was refined with Gaia DR3. Targets were are rarely observed), only a limited set of abundances
selected to be of spectral type FGK (both This work was partially funded by the PRIN INAF is measured, and there is no controlled selection of
giants and dwarfs), with Av < 2. HR targets 2019 grant ObFu 1.05.01.85.14 (Building up the halo: cluster stars to be observed.
Acknowledgment: Cambridge Astronomical Survey Unit
ESO, M.-R. Cioni/VISTA Magellanic Cloud survey.
Michał Pawlak 1 high-mass binary candidates, the only ellipsoidal effect, induced by tidal forces
Tsevi Mazeh 2 confirmed stellar-mass BH in the Galaxy exerted by the BH on its optical compan-
Simchon Faigler 2 (see, for example, BlackCAT; Corral- ion. For a general discussion of the ellipsoi-
Tomer Shenar 3 Santana et al., 2016). Most of the stel- dal effect see, for example, Mazeh (2008).
Hugues Sana 4 lar-mass BH known so far have been dis-
Dolev Bashi 2 covered by their X-ray emission, which The goal of the survey is to discover tens
is due to either mass transfer from the of those unknown dormant BH. The new
low-mass (mostly F–K-type) companion discoveries will allow us to study the char-
1
Astronomical Observatory, Jagiellonian overflowing its Roche lobe (BH low-mass acteristics of the dormant objects in detail.
University, Kraków, Poland X-ray binaries, BH-LMXBs), or accretion We will be able to understand how fre-
2
School of Physics and Astronomy, from the stellar wind coming from the quent compact objects are, as well as the
Tel Aviv University, Israel high-mass (OB star) companion (BH range and distribution of their masses.
3
Anton Pannekoek Institute for high-mass X-ray binaries BH-HMXBs). This will lead to a much better under-
Astronomy, Amsterdam, The known masses of stellar X-ray BH standing of the astrophysics behind the
the Netherlands as a function of their orbital periods formation of dormant compact systems.
4
Institute of Astronomy, KU Leuven, (Tetarenko et al., 2016) are plotted in Fig-
Belgium ure 1. As can be seen, most of the known
BH binaries have relatively short periods, Specific scientific goals
with a median at about 0.5 days. The
The goal of the Spectroscopic Discovery X-ray emission of the BH-LMXBs is char- We plan to use ellipsoidal modulation
of Binaries with Dormant Black Holes acterised by luminous (Lx ~ 10 37 erg s –1) (Faigler & Mazeh, 2011; Faigler et al.,
survey is to spectroscopically follow up outbursts, lasting for about a month, fol- 2012) to discover tens of new unknown
stars that might have dormant compact lowed by decades-long periods of quies- dormant BH binaries. Since BH binaries
companions, either black holes or cence (Lx ~ 10 31–33 erg s –1) (see Remillard are rare, we have to search for these sys-
neutron stars. These stars have been and McClintock, 2006 for a review). tems in large samples, on the order of
identified as ellipsoidal binaries in the 10 6 or more stars. Astronomy is now at a
Magellanic Clouds by the Optical According to the commonly accepted unique point in its history, where enough
Gravitational Lensing Experiment model, outbursts from BH-LMXB systems stellar light curves with adequate preci-
(OGLE). A sample of more than 700 ellip- are due to accretion disc instabilities that sion to measure ellipsoidal modulation
soidals with periods shorter than 10 days modulate the accretion rates onto the are available for the first time.
will be observed to obtain multi-epoch BH. Between eruptions, these systems
radial-velocity measurements. 4MOST are barely detectable, because a substan- The key step is identifying stars with pho-
radial velocities in conjunction with tial part of the energy generated by the tometric periodic ellipsoidal modulations
OGLE photometry will allow the determi- small mass flow is not radiated but stored that indicate the existence of companions
nation of the secondary component as thermal energy in the accretion discs. more massive than the optical compo-
mass and hence the identification of Thus, many BH-LMXB remain undetected, nents. If the optical star is on the main
systems with compact companions. because they have been in their quies- sequence (MS), then the more massive
cent state since X-ray surveys began. unseen companion must be a compact
object. To confirm our detection of a com-
Scientific context A much larger fraction of BH with low- pact companion and identify false alarm
mass stellar companions are not yet cases, we will measure the radial velocity
Stellar black holes (BH) and neutron stars detected because their optical counter- (RV) of the candidates and find the sys-
(NS) are the end products of massive stars parts are within their Roche lobes. In tems displaying ellipsoidal modulations
that collapse after they have consumed these systems, mass is not transferred, with large amplitudes. The mass of the
their energy reservoirs. In NS, gravitational and X-rays are not generated, making companion will be derived from the com-
contraction breaks the protons of all these systems dormant BH. For example, bination of the RVs and the ellipsoidal
atoms and creates one big nucleus of 1057 Breivik, Chatterjee & Larson (2017) esti- modulation. The companion can be either
neutrons. BH are even more exotic objects mated that there should be 106 Galactic a white dwarf (very rare), or a NS, with a
— their entire mass is collapsed into one BH binary systems, most of which are still mass at around 1.4 M☉ or a BH (with a
singular point, such that nothing, includ- waiting to be discovered. One such sys- larger mass). We can discover only bina-
ing electromagnetic waves, has enough tem, Gaia-BH1, was discovered recently ries for which the orbital separation is not
energy to escape its event horizon. using astrometry (El-Badry et al., 2023). too large, otherwise the ellipsoidal effect
Because dormant BH do not emit any will be too small to detect. From the OGLE
electromagnetic radiation, they are very All of the 19 confirmed short-period light curves available, we were able to
difficult to discover. Only about 20 dynam- BH-LMXBs show periodic optical bright- detect tens of binaries with an ellipsoidal
ically confirmed stellar BH are known to ness modulations on the order of a few modulation of 1% or larger, corresponding
reside in X-ray binary systems with low- percent on the timescale of the orbital to binaries with separations which are
mass stellar companions. These are, period (for example, XTE J1118+48 and larger than the radius of the visible com-
along with Cyg X-1 and several other A0620-00). This is mostly due to the panion by up to a factor of 10.
Ása Skúladóttir 1,2 13
Lagrange Laboratory, Côte d’Azur Scientific context
Arthur Alencastro Puls 3 Observatory, Côte d’Azur University,
Anish M. Amarsi 4 Nice, France The Milky Way environment is rich in
Giuseppina Battaglia 5,6 14
Strasbourg Astronomical Observatory, satellite galaxies, stellar streams and
Sven Buder 7,8 CNRS, University of Strasbourg, France accreted systems, that trace galaxy evo-
Simon Campbell 8,9 15
Astrophysics Laboratory, Physics Insti- lution and chemical enrichment through-
Salvador Cardona-Barrero 5,6 tute, École Polytechnique Fédérale, out cosmic time. Because they are close
Norbert Christlieb 10 Lausanne, Switzerland by, these relics of early galaxy formation
Diane K. Feuillet 11 16
Leibniz Institute for Astrophysics, can be studied in extraordinary detail,
Viola Gelli 1,2 Potsdam, Germany star by star, giving invaluable insight into
Camilla J. Hansen 3,12 17
Computational Astronomy Institute, stellar evolution and nucleosynthesis, as
Vanessa Hill 13 Centre for Astronomy, Heidelberg well as the Milky Way’s hierarchical
Rodrigo Ibata 14 University, Germany growth. This provides a clear motivation
Pascale Jablonka 15 18
Department of Astronomy, Stockholm for the 4MOST survey of dwarf galaxies
Nikolay Kacharov 16 University, Sweden and their stellar streams (4DWARFS)
Amanda Karakas 8,9 19
GEPI, Paris Observatory, France which aims to study and map these indi-
Andreas J. Koch-Hansen 17 20
Konkoly Observatory, Research Centre vidual structures. 4DWARFS will target
Karin Lind 18 for Astronomy and Earth Sciences, all ~ 50 of the known Milky Way satellite
Linda Lombardo 19 Eötvös Loránd Research Network, dwarf galaxies in the southern hemi-
Romain E. R. Lucchesi 1 Budapest, Hungary sphere, covering over three orders of
Maria Lugaro 20,21,22,9 21
CSFK, MTA Centre of Excellence, magnitude in galaxy mass. Additionally,
Nicolas Martin 14,12 Budapest, Hungary 4DWARFS will observe dozens of already
Davide Massari 23 22
Institute of Physics, ELTE Eötvös identified stellar streams in the Galactic
Thomas Nordlander 7,8 Loránd University, Budapest, Hungary halo which are the remnants of currently
Moritz Reichert 24 23
INAF–Bologna Astrophysics and Space interacting or previously dissolved sys-
Martina Rossi 1,2 Science Observatory, Italy tems (Figure 1).
Ashley J. Ruiter 25 24
Department of Astronomy and
Stefania Salvadori 1,2 Astrophysics, University of València, The dwarf galaxy satellites of the Milky
Ivo R. Seitenzahl 25 Spain Way are intrinsically metal-poor, with a
Eline Tolstoy 26 25
School of Science, University of New dominant old stellar population, making
Theodora Xylakis-Dornbusch 10,27 South Wales, Sydney, Australia them ideal fossils with which to study the
Kristopher C. Youakim 18 26
Kapteyn Astronomical Institute, imprints of the first stars in the Universe.
University of Groningen, Their star formation was inefficient,
the Netherlands making the effects of delayed nucleo
1 27
Department of Physics and Astronomy, International Max Planck Research synthetic channels (type Ia supernovae
University of Florence, Italy School for Astronomy & Cosmic Physics [SN Ia], asymptotic giant branch [AGB]
2
INAF–Arcetri Astronomical Observatory, at the University of Heidelberg, Germany stars, neutron star mergers) especially
Florence, Italy prominent and straightforward to study,
3
Institute of Applied Physics, Goethe at various metallicities and star formation
University Frankfurt, Germany The present-day Milky Way is the histories (for example, Tolstoy, Hill & Tosi,
4
Theoretical Astrophysics, Department result of a long history of mergers 2009), providing unique insights into
of Physics and Astronomy, Uppsala and interactions with smaller galaxies. chemical evolution and nucleosynthesis.
University, Sweden The 4DWARFS survey will target the
5
Canary Islands Institute of Astrophysics, dwarf galaxies and stellar streams in These small systems are relics of hierar-
Tenerife, Spain the 4MOST footprint, and unveil their chical structure formation, opening up the
6
University of La Laguna, Tenerife, Spain chrono-chemo-kinematical properties. possibility of tracing mass assembly, and
7
Research School of Astronomy and The survey will provide radial velocities, both ex-situ and in-situ star formation in
Astrophysics, Australian National chemical abundances and stellar ages the Milky Way environment. Dwarf galax-
University, Canberra, Australia for 140 000 stars, and thus increase ies are dominated by dark matter, and
8
ARC Centre of Excellence for All Sky the number of stars with detailed infor- can be used as probes of cosmology and
Astrophysics in 3 Dimensions, Australia mation in such systems by several galaxy formation. Those that halted star
9
School of Physics and Astronomy, orders of magnitude. 4DWARFS will formation long ago are likely the best
Monash University, Australia provide a new, deeper view of the Milky examples of pristine dark-matter halos in
10
Astronomy Centre, Heidelberg Way environment, shedding light on the Universe, and they are the ideal
University, Germany. the first stars, chemical evolution, dark probes to understand the nature of dark
11
Department of Astronomy and Theoret- matter halos, and hierarchical galaxy matter through the study of their stellar
ical Physics, Lund Observatory, Sweden formation down to the smallest scales. 3D kinematics (for example, Massari et
12
Max Planck Institute for Astronomy, al., 2018). Furthermore, the stellar
Heidelberg, Germany streams observed in the Galactic halo
represent clear evidence of systems cur- What are the properties of the first stars? stars (s- and i-processes) and the kilono-
rently being accreted, allowing us to vae of neutron star mergers (r-process).
study their tidal disruption in incompara- The chemical yields of the first stars
ble detail. In particular, the Sagittarius (Pop III) are thought to be preserved on
dwarf spheroidal (dSph) is the prime the surfaces of low-metallicity, low-mass, What are the dynamical properties of
example of a tidally disrupted Milky Way long-lived stars (for example, Hansen et dwarf galaxies?
satellite galaxy, and it leaves a trail of al., 2020). 4DWARFS will identify the most
stars that wraps more than a full orbit metal-poor stars of dwarf galaxies, By providing an unprecedented sample
around the Galaxy (Figure 1). This makes increasing their numbers by more than an of precise radial velocities and coupling
it a major contributor to the Milky Way order of magnitude. Furthermore, we will those with Gaia proper motion measure-
stellar halo along with other past merg- quantify the properties of metal-poor ments, 4DWARFS will determine the dark
ers. Sagittarius and its stream therefore stars across different galaxies, for exam- matter profile of almost 50 galaxies,
provide one of the best opportunities to ple the fraction of carbon-enhanced thereby shedding light on the nature of
study a galaxy that is currently undergo- metal-poor stars which are thought to dark matter itself. We will drastically
ing disruption. be the descendants of faint Pop III super- increase observations in the outskirts of
novae (for example, Iwamoto et al., 2005). these galaxies, which will allow us to test
A careful study of dwarf galaxies and We will search for and identify rare fossils whether dwarf galaxies are in equilibrium,
their stellar streams is thus of fundamen- and descendants of the first stars (Rossi, as commonly assumed, or if they exhibit
tal importance to obtaining a complete Salvadori & Skúladóttir, 2021), such as more complex kinematics (for example,
picture of the evolution of our Galaxy. those formed in the ejecta of zero- Martin et al., 2016). These dynamical
Furthermore, this will enable us to under- metallicity pair-instability supernovae, studies will benefit hugely from the fact
stand in detail the properties of the dwarf and high-energy Pop III supernovae. This that 4DWARS will constrain the binary
galaxies themselves, which are the earli- will allow us to put constraints on the fraction and period distribution of stars in
est and most common type of galaxy in intrinsic properties of the first stellar popu- dwarf galaxies, which are currently
the Universe. lation, such as their mass distribution and essentially unknown. Understanding their
the energy distribution of their supernovae. binary properties will also provide fresh
insights into star formation and stellar
Specific scientific goals evolution at low metallicities.
How are the chemical elements created
The main goal of 4DWARFS is to char and distributed?
acterise dwarf galaxies and stellar What are the small-scale limits of
streams by mapping their chrono-chemo- Even after decades of study, the funda- hierarchical galaxy formation?
kinematical properties. We will derive mental physical conditions of major
radial velocities, chemical abundances nucleosynthetic sites remain obscure. In recent years the complexity in the Milky
and stellar ages (for example, Feuillet et Investigating different galaxies with a Way environment has been revealed, with
al., 2016) for stars in all known dwarf gal- variety of star formation and chemical dozens of identified structures. The pro-
axies (smaller than the Small Magellanic enrichment histories is key to breaking genitors of these systems are still being
Cloud) and stellar streams in the 4MOST degeneracies in chemical evolution (for investigated, but they are likely a mixed bag
footprint of the Milky Way halo. Using both example, Skúladóttir & Salvadori, 2020). of objects: a variety of dwarf galaxies and
high-resolution (HR) and low-resolution Dwarf galaxies have intrinsically low star disrupted globular clusters with different
(LR) spectra, our survey will provide formation rates compared to larger galax- ages and metallicities. 4DWARFS will char-
10–25 elemental abundances for 100 000 ies, and all long-timescale nucleosyn- acterise the accreted systems of the
stars in dwarf galaxies, and 30 000 stars thetic channels, such as SN Ia, AGB stars Galactic halo both kinematically and chemi-
in stellar streams (Figure 1). In addition, and neutron star mergers are thus more cally, uncovering their metallicity distribu-
we will push down to the fainter dwarf prominent relative to core-collapse SN tion, chemical evolution and star formation
galaxies and smaller stellar streams, (short timescales). Compared to the Milky history. This will allow us to understand the
retrieving abundances of around five ele- Way, dwarf galaxies have much simpler progenitors of these streams and the sys-
ments for a further 10 000 stars. The star formation histories, which makes tems that build up the Milky Way halo (for
homogeneity of this sample will be them easier to model, providing a more example, Ibata et al., 2021). In particular, we
unprecedented in comparison to the detailed understanding. 4DWARFS will will map and characterise the currently
present literature. constrain the fundamental physics of ongoing disruption of the Sagittarius dSph
nucleosynthetic sites: their rates, yields, in extraordinary detail. Finally, we will move
The 4DWARFS dataset is therefore incred- metallicity dependence and time-delay to even smaller scales of accretion, and for
ibly rich, and will impact several fields, distribution functions. In this way, the first time identify in a systematic way
ranging from cosmology to stellar physics. 4DWARFS will provide key insights into mergers that have happened within the
The key scientific questions that 4DWARFS the elusive progenitors of SN Ia (for exam- dwarf galaxies themselves (for example,
will attack are listed below. ple, Maoz, Mannucci & Nelemans, 2014), Cicuendez & Battaglia, 2018), thus quanti-
as well as the various production sites of fying hierarchical galaxy formation down to
neutron-capture elements, such as AGB the smallest scales.
–45º
combination of 4DWARFS and other
–60º
Galactic surveys within 4MOST will there-
–75º fore allow us to paint a comprehensive
R.A. picture of the formation and evolution of
the Milky Way environment, from cosmic
Sagittarius Fornax & Sculptor dawn until the present day.
Sagittarius stream Carina & Sextans
Stellar streams Ultra faint dwarfs
Acknowledgements
Angela Iovino 1 6
Centre for Astrobiology (CSIC-INTA), non-linear processes are at play (for
Amata Mercurio 2 Torrejón de Ardoz, Spain example, energetic feedback, both from
Anna Rita Gallazzi 3 7
INAF–Trieste Astronomical Observatory, active galactic nuclei [AGN] and star for-
Francesco La Barbera 2 Italy mation, and mergers). A detailed compar-
Marcella Longhetti 1 8
Turku University, Finland ison between observational data, simula-
Crescenzo Tortora 2 9
Kavli Institute for Cosmology, tions and advanced theoretical models
Stefano Zibetti 3 Cambridge, UK can shed light on the mechanisms that
Francesco Belfiore 3 10
Andalucia Institute of Astrophysics, regulate the connection between galaxies
Matteo Bianconi 4 Granada, Spain and their DM halos.
Giovanni Busarello 2 11
University of Atacama, Copiapò, Chile
Enrico M. Corsini 5 12
INAF–Padua Astronomical Observatory, Good quality (both in resolution and
Luca Costantin 6 Italy signal-to-noise ratio [SNR]) spectra pro-
Gabriella De Lucia 7 13
Heidelberg University, Germany vide information about the current physi-
Roberto De Propris 8 14
INAF–Bologna Astrophysics and Space cal conditions of the multiple components
Francesco D’Eugenio 9 Science Observatory, Italy in a galaxy (stars, gas and, indirectly,
Fabio Fontanot 7 15
Madrid Complutense University, Spain DM), while simultaneously unveiling the
Rubén García-Benito 10 16
Ghent University, Belgium archaeological treasure that is encoded
Michaela Hirschmann 7 17
Canary Islands Institute of Astrophysics, in its stellar populations. The whole past
Christopher Haines 11 Tenerife, Spain star formation and chemical-enrichment
Filippo Mannucci 3 18
ESO history of a galaxy are reflected in its pres-
Sean McGee 4 19
INAF–Institute of Space Astrophysics ent stellar population properties. The
Paola Merluzzi 2 and Cosmic Physics, Milan, Italy chemical abundance of gas and stars in a
Lorenzo Morelli 11 20
Space Telescope Science Institute, galaxy holds the memory of the baryonic
Alessia Moretti 12 Baltimore, USA cycle that regulates its star formation, by
Anna Pasquali 13 21
University of Groningen, balancing the inflow of pristine gas, the
Bianca Poggianti 12 the Netherlands outflow of metal-loaded gas blown out by
Lucia Pozzetti 14 stellar/AGN winds, and the re-accretion of
Giulia Rodighiero 5 this metal-enriched gas (Hunt et al., 2020).
Patricia Sánchez-Blázquez 15 Galaxy spectra encode in their contin-
Arjen van der Wel 16 uum and absorption/emission features a The study of galaxies in the local Universe
Alexandre Vazdekis 17 wealth of information on galaxy physics, has greatly enriched our understanding of
Benedetta Vulcani 12 mass assembly and chemical enrich- galaxy evolution. The spectra from the
Anita Zanella 12 ment history. The 4MOST-StePS survey Sloan Digital Sky Survey (SDSS; York et
Marianna Annunziatella 6 will collect high-quality spectra (with a al., 2000) have provided a robust z ~ 0
Alice Concas 18 median signal-to-noise ratio of about anchor to both theoretical and empirical
Letizia P. Cassarà 19 30 Å –1, and resolution R ~ 5000) for a approaches, proving to be an immensely
Giovanni Cresci 3 sample of about 3300 galaxies brighter valuable tool to understand the physics of
Mirko Curti 9 than IAB = 20.5 within the RA-Dec-z foot- galaxies. Unfortunately, the so-called
Adriana de Lorenzo-Cáceres 17 print of the WAVES-Deep survey. These archaeological reconstruction from z ~ 0
Anna Ferre Mateu 17 spectra will provide a precise empirical galaxies is limited by inherent physical
Rosa M. González Delgado 10 description of the evolutionary path of degeneracies between different parame-
Chiara Mancini 5 massive galaxies in the intermediate ters (for example age and metallicity) and
Camilla Pacifici 20 redshift range (0.3 < z < 0.7) between by the similarity of stellar population
Enrique Perez-Montero 10 the LEGA-C and SDSS surveys. The spectra with ages greater than 5 Gyr;
Alessandro Pizzella 5 locations of the galaxies within the cos- together these prevent an accurate
Pablo G. Perez-Gonzalez 6 mic web, unveiled by WAVES-Deep, will reconstruction of a galaxy’s ancient star
Scott C. Trager 21 disclose the connection between galaxy formation history. To circumvent this
Daniela Vergani 14 properties and environment, down to problem a direct look-back approach is
the scales of galaxy pairs. needed, i.e., observing galaxies at high
redshifts to make a direct census of dif-
1
INAF–Brera Astronomical Observatory, ferent populations at different remote
Italy Scientific context cosmic epochs. High-quality galaxy
2
INAF–Capodimonte Astronomical spectra, observed in different environ-
Observatory, Naples, Italy The complex physical processes that ments and across a continuous range of
3
INAF–Arcetri Astronomical Observatory, cause the formation and evolution of cosmic times, provide precise measure-
Florence, Italy luminous structures to deviate from the ments of a variety of physical parameters
4
University of Birmingham, UK assembly history of dark matter (DM) of galaxies, enabling the star formation
5
University of Padua, Italy halos are yet to be fully understood. On and chemical enrichment histories at
galactic scales (below ~ 1 Mpc), highly early times to be unveiled. This strategy in
systems (Gallazzi et al., 2005, 2014). The interplay between theoretical rendi- Our sample is positioned within the
This allows the evolution of different tion and 4MOST-StePS observed spectra Large-Scale Structure web, revealed in
scaling relations and their intrinsic scat- will enable a uniquely accurate two-way fine detail by WAVES-Deep (see Driver et
ter to be tracked. Unbiased values of comparison of theoretical predictions al., 2019). This accurate environmental
metallicity can be obtained with with observational data. This kind of information will enable us to explore the
detailed analysis of single element approach is needed for a quantitative connection between the different galaxy
abundance ratios (for example, [Mg/Fe], understanding of the processes driving properties and the environment where
[Ca/Fe], [C/Fe]), measured with an the evolution of massive galaxies. galaxies reside in an evident and advan-
accuracy of 0.1 to 0.25 dex (depending 4MOST-StePS data will lift the degenera- tageous synergy with WAVES-Deep.
on the element) for intermediate-age cies of the interpretation that WEAVE-
populations based on an SNR > 30 Å –1. StePS will not be able to do, given its lack 4MOST-StePS will push the 4MOST spec-
Gas-phase metallicity can also be accu- of detail. trograph to the limit of its performance,
rately measured using the faint auroral well beyond its redshift machine capabili-
lines (for example, Maiolino & Mannucci, Finally, our observational strategy will ties. Our survey will obtain high-quality
2019) that can be detected in these provide a tail of galaxies with excellent data, bridging the gap in our knowledge
high-SNR spectra, after stacking. These SNR (> 40 Å –1 in the I band; red points in in the intermediate redshift range, which
quantities can constrain the physical Figure 1), enabling innovative science. For still lacks such high-quality data, and
processes of chemical enrichment in this higher-SNR sample, we will derive will put to full use the environment charac-
galaxies, as well as the timescale of precise individual abundance ratios, con- terisation provided by WAVES-Deep. The
quenching processes and of the main strain the stellar IMF from key sensitive synergy with the science case of WAVES-
star formation event (Sánchez-Blázquez features, detect even tiny fractions (at the Deep is a win-win opportunity that our
et al., 2003; Carretero et al., 2004); 1% level) of residual young stars, detect project fully exploits, making 4MOST-
– Characterisation of the kinematics of faint auroral lines for direct gas metallicity StePS a low-cost survey with the poten-
the stellar and gaseous components estimates in star-forming galaxies and tial for a very high scientific return.
(velocity distribution, inflows/outflows). identify gas flows probed by absorption
An accurate estimate of dynamical and emission lines. The higher moments
masses can be obtained with a formal of the line-of-sight velocity dispersion will References
accuracy of ~ 25% (40%) using spectra also be retrievable, improving dynamical Bruzual, G. & Charlot, S. 2003, MNRAS, 344, 1000
with SNR ~ 30 (20) Å –1, by using Jeans mass estimates. The high-SNR sample Carretero, C. et al. 2004, ApJ, 609, L45
modelling of the 4MOST aperture will be used to test for systematics in our Chabrier, G. 2003, PASP, 115, 763
velocity dispersion of individual galax- estimates (for example the accurate Chauke, P. et al. 2018, ApJ, 861, 13
Concas, A. et al. 2017, MNRAS, 468, 1747
ies. Its comparison with stellar masses reconstruction of star formation histories Concas, A. et al. 2019, A&A, 622, A188
is crucial to linking the evolution of at lower SNR values). Driver, S. P. et al. 2019, The Messenger, 175, 46
galaxies to their DM halos (Tortora et Fontanot, F. et al. 2017, MNRAS, 464, 3812
al., 2018) and to providing dynamical Fontanot, F. et al. 2021, MNRAS, 504, 4481
Gallazzi, A. et al. 2005, MNRAS, 362, 41
constraints on one of the most disputed Target selection and survey area Gallazzi, A. et al. 2014, ApJ, 788, 72
ingredients in stellar populations, i.e., Hirschmann, M. et al. 2017, MNRAS, 472, 2468
the shape of the stellar initial mass We will pool our targets with those of Hunt, L. et al. 2020, A&A, 643, A180
function (IMF; for example, Tortora, Consortium Surveys in the ~ 66-deg2 Iovino, A. et al. 2022, submitted to A&A
Maiolino, R. & Mannucci, F. 2019, A&ARv, 27, 3
Romanowsky & Napolitano, 2013). The footprint of the WAVES-Deep survey (four Sánchez-Blázquez, P. et al. 2003, ApJ, 590, L91
excellent spectral quality will reveal the so-called Deep Drilling Fields and the Speagle, J. S. et al. 2014, ApJS, 214, 15
demographic of gas outflows/inflows GAMA23 Deep field). The choice of these Tortora, C., Romanowsky, A. J. & Napolitano, N. R.
traced by interstellar absorption and sky regions is driven by observational 2013, ApJ, 765, 8
Tortora, C. et al. 2018, MNRAS, 473, 969
broad high-velocity nebular emission as strategy advantages and significant sci- van der Wel, A. et al. 2021, ApJS, 256, 44
done at z ~ 0 with the SDSS galaxy entific benefits. All the WAVES-Deep Wu, P.-F. et al. 2021, AJ, 162, 201
spectra (Concas et al., 2017, 2019). areas possess a variety of ancillary York, D. G. et al. 2000, AJ, 120, 1579
photometric data that can be used to
All these quantities will be complemented complement the 4MOST high-quality
by the environment information provided spectroscopic information on a wider
by WAVES-Deep and compared with wavelength range. In these fields, we can
state-of-the-art large-scale cosmological take advantage of the repeated passes
simulations (Fontanot et al., 2017, 2021). needed to achieve the sampling rate
The 4MOST-StePS sample will be the ideal requested by WAVES-Deep and success-
observational counterpart to synthetic fully pursue our 30-hour exposure time
spectral energy distributions obtained from strategy, while using a minimal fraction of
galaxy formation simulations in a cosmo- fibre hours: less than 0.5% of the total
logical context — including the stellar amount available to Community surveys
continuum and the nebular emission lines in low-resolution mode.
of various origins (Hirschmann et al., 2017).
Kenneth Duncan 1 18
Space Telescope Science Institute, cosmic SFR density and the cosmic
Andrew Baker 2 USA molecular gas mass fraction (for example,
Philip Best 1 19
ESO Davé et al., 2020). However, these mod-
Sarah Blyth 3 els’ predictions of more modest redshift
Nina Hatch 4 evolution in the cosmic HI density can
Benne Holwerda 5 Galaxy evolution is regulated by the presently be compared only to absorption-
Matt Jarvis 6 continuous cycle of gas accretion, con- line studies, where the exact relationships
Natasha Maddox 7 sumption and feedback. Crucial in this between galaxies and absorbers remain
Daniel J. B. Smith 8 cycle is the availability of neutral atomic uncertain, and where possible relation-
Marina Arnaudova 8 (HI) and molecular hydrogen. Our cur- ships between HI content and star forma-
Laurent Chemin 9 rent inventory of HI, however, is very tion averaged over Gyr timescales cannot
Romeel Davé 1 limited beyond the local Universe be explored at all. Testing these models
James Dunlop 1 (z > 0.25), resulting in an incomplete and fully understanding the buildup of
Bradley Frank 10 picture. ORCHIDSS is designed to stellar mass over cosmic time requires
Eric Gawiser 2 address this critical challenge, using the directly observing all components of the
Anniek Gloudemans 11 powerful combination of 4MOST spec- gas reservoirs in galaxies, out to signifi-
Catherine Hale 1 troscopy and sensitive radio observa- cant cosmological redshifts that have not
Ian Heywood 6 tions from the MeerKAT deep extraga- been accessible until now.
Sheila Kannappan 12 lactic surveys to trace the evolution of
Rohit Kondapally 1 neutral gas and its lifecycle within gal- Thanks to the ongoing revolution in radio
Ross McLure 1 axies across the bulk of cosmic history. astronomy, it is now possible to probe
Leah Morabito 13 these regimes and begin to tackle a num-
Nicole Nesvadba 14 ber of key challenges in our understand-
Hengxing Pan 10 Scientific context ing of galaxy formation: what is the cos-
Anastasia Ponomareva 6 mic history of neutral hydrogen? what is
Matt Prescott 15 On galactic scales, star formation is regu- the lifecycle of gas within galaxies? and
Hayley Roberts 16 lated by the complex cycling of gas in how does feedback impact the gas res-
Huub Röttgering 11 and out of the interstellar medium (ISM; ervoirs of galaxies?
Rachel Somerville 17 Kereš et al., 2005). This cycle starts with
Madalina Tudorache 6 primordial gas accretion, which fuels star At the forefront of this revolution is the
Mattia Vaccari 3 formation and accretion onto black holes. MeerKAT telescope, its radio continuum
Imogen Whittam 6 Feedback in the form of stellar winds, surveys now simultaneously providing
John Wu 18 supernovae and active galactic nuclei one of the most reliable, obscuration-free,
Martin Zwaan 19 (AGN) jets shock heats the gas and drives probes of star formation and a means
outflows. The last stage of the cycle is the of tracing black hole accretion in AGN.
cooling of the recycled gas from the hot Furthermore, its surveys of the 21-cm HI
1
University of Edinburgh, UK halo, some of which then accretes into emission line provide a powerful new
2
Rutgers University, USA the ISM and restarts the cycle. Large probe of the ISM of galaxies up to z = 1.4.
3
University of Cape Town, South Africa investments have been made in studying
4
University of Nottingham, UK the coolest, dense molecular gas and the MeerKAT’s flagship deep extragalactic
5
University of Louisville, USA hottest, most rarefied gas halos of galax- surveys, MeerKAT International GHz
6
University of Oxford, UK ies (with, for example, ALMA, XMM and Tiered Extragalactic Exploration
7
University of Bristol, UK Chandra). However, the cycling of cold (MIGHTEE; Heywood et al., 2022) and
8
University of Hertfordshire, UK atomic gas, the cleanest probe of the Looking at the Distant Universe with the
9
Andres Bello University, Santiago, Chile ISM, is poorly understood. This gas can MeerKAT Array (LADUMA; Blyth et al.,
10
South African Radio Astronomy be traced by the 21-cm transition line of 2016), are breaking new ground in the
Observatory, South Africa neutral hydrogen (HI), but sensitivity limits study of neutral atomic gas in galaxies.
11
Leiden University, the Netherlands have restricted such studies to the local MIGHTEE will directly detect HI in several
12
University of North Carolina at Universe (Catinella & Cortese, 2015). hundred galaxies at z < 0.2, and a few of
Chapel Hill, USA the rarest, most HI-massive galaxies out
13
Durham University, UK In the local Universe, the instantaneous to z = 0.5 thanks to the large volume
14
Lagrange Laboratory, Côte d’Azur star formation rate (SFR) in galaxies is probed over its multiple fields (> 107 Mpc 3
Observatory, Côte d’Azur University, much more tightly connected to molecu- at 0.45 < z < 0.55). Reaching ~ 4.5 times
Nice, France lar gas than to the neutral atomic gas deeper in HI sensitivity at z < 0.5 and
15
University of the Western Cape, from which molecular clouds condense extending frequency coverage of HI to
South Africa (for example, Bigiel et al., 2011). Galaxy z = 1.4 in a single field, LADUMA will
16
University of Colorado at Boulder, USA evolution models that incorporate this detect larger samples of lower-mass gal-
17
Flatiron Institute, New York, USA connection have been able to account for axies and enable detection of the most
the sharp increases with redshift of the HI-massive galaxies out to z = 1. Despite
2° – 36°
– 37°
1°
– 38°
58° 57° 56° 55° 54° 53° 52° 51° 50°
151° 150° 149°
these substantial advances, however, The cosmic history of neutral hydrogen ies and key physical properties (SFR/
the 21 cm flux-limited nature of these stellar mass/metallicity; Davé et al., 2020),
surveys means that, by themselves, they Galaxy evolution is driven by the availabil- measured by 4MOST and the wealth of
will preferentially detect only the most ity of atomic and molecular hydrogen in available photometric data, over 9 Gyr of
HI-rich galaxies. the ISM of galaxies. The large, homoge- cosmic history.
neously selected and highly complete
The full potential of MeerKAT’s surveys spectroscopic samples provided by
will only be realised if they are matched ORCHIDSS will allow us to trace the The role of feedback in regulating the
with extensive optical spectroscopy. By HI content in normal galaxies out to gas reservoir of galaxies
building a large, representative sample of z = 1.4. ORCHIDSS will provide the most
galaxies with accurate spectroscopic reliable measurement of the evolution of MeerKAT will identify star formation and
redshifts, we can spectrally align and the cosmic HI density and HI mass func- AGN activity across the bulk of cosmic
stack the radio data to extract HI meas- tion out to the epoch where galaxy for- history with a single, unified dataset,
urements below the MIGHTEE and mation peaked. unbiased by obscuration. ORCHIDSS will
LADUMA noise thresholds (Chowdhury, enable us to relate that accretion and
Kanekar & Chengalur, 2022; Pan et al., SFR to gas properties. With kinematic
2022), revealing the neutral gas properties The lifecycle of gas in galaxies constraints on both the ionised gas (via
of representative samples of the full galaxy optical emission/absorption lines) and
population over much of cosmic history. The stellar growth of galaxies is regulated neutral gas components providing direct
by the continuous cycle of gas accretion, evidence of outflows, we will explore the
consumption and feedback. The parame- complex interplay between star forma-
Specific scientific goals ters governing the flow of this gas cycle, tion/AGN fuelling, activity and feedback
such as the primordial gas accretion rate for the multi-phase warm gas in galaxies.
Combining the deep radio observations and efficiency of stellar feedback, are
offered by MeerKAT and the high-multiplex highly uncertain and yet vital to obtain for
optical spectroscopy offered by 4MOST, any successful model of galaxy evolution Legacy value
the Optical, Radio Continuum and HI (Somerville & Davé, 2015). For the first
Deep Spectroscopic Survey (ORCHIDSS) time, ORCHIDSS will measure these ORCHIDSS will enable a huge range of
will address the key scientific challenges parameters via the fundamental relations additional science. For example, the
outlined below. between the neutral gas content of galax- combination of large optical spectro-
Daniel Gruen 1 16
Institute of High-Energy Physics, samples of galaxies (see, for example,
Jamie McCullough 2,3,1 Barcelona Institute of Science of Newman & Gruen, 2022 for a review). The
Alexandra Amon 4,5 Technology, Spain direct determination of the redshift
Gary Bernstein 6 17
Catalan Institution for Research and distribution with spectroscopy is advanta-
Jan Luca van den Busch 7 Advanced Studies, Barcelona, Spain geous in multiple ways. However, present
Rebecca Canning 8 18
Jet Propulsion Laboratory, California samples of spectroscopic redshifts allow
Francisco Castander 9,10 Institute of Technology, USA for a calibration of galaxy distances that
Joseph DeRose 11 is only scarcely sufficient for the ongoing
William Hartley 12 lensing experiments (Hildebrandt et al.,
Hendrik Hildebrandt 7 Accurate knowledge of the redshift dis- 2021; Myles et al., 2021; Rau et al., 2022).
Konrad Kuijken 13 tributions of faint samples of galaxies The currently required calibration uncer-
Jochen Liske 14 selected by broad-band photometry is tainties on the mean redshift are of order
Daniel Masters 15 a prerequisite for future weak lensing |Δz| ~ 0.01. But the next generation of
Ramon Miquel 16,17 experiments to deliver precision tests surveys by Euclid, Vera C. Rubin Obser-
Andrea Pocino Yuste 9,10 of our cosmological model. The most vatory, and the Nancy Grace Roman
Aaron Roodman 2,3 direct way to measure these redshift Space Telescope are predicted to need
Stella Seitz 1 distributions is spectroscopic follow-up an order of magnitude improvement, with
Roberto Saglia 1 of representative galaxies. For this to the calibrated mean redshift accurate to a
Daniel Stern 18 be efficient and accurate, targets have few parts in a thousand and stringent
Luca Tortorelli 1 to be selected such that they system requirements as well on the calibration of
Angus H. Wright 7 atically cover a space defined by appar- the width of redshift distributions. We will
ent colours in which there is little varia- thus be severely limited in testing cosmo-
tion in redshift at any point. 4C3R2 logical models unless large, systemati-
1
University Observatory, Faculty of will follow this strategy to observe over cally selected and highly complete spec-
Physics, Ludwig Maximilians University, 100 000 galaxies selected by their troscopic samples can be obtained.
Munich, Germany KiDS-VIKING ugriZYJHKs photometry
2
Kavli Institute for Particle Astrophysics over a footprint identical to that of the
and Cosmology, Stanford University, WAVES survey, to constrain the colour- Complete calibration of the colour-
USA redshift relation with high multiplicity redshift relation
3
SLAC National Accelerator Laboratory, across two-thirds of the colour space of
USA future Euclid and Rubin samples. Direct calibration of redshift distributions
4
Institute of Astronomy, University of (for example, Hildebrandt et al., 2017)
Cambridge, UK effectively consists of making a histogram
5
Kavli Institute for Cosmology, University Scientific context: the colour-redshift of spectroscopic redshifts. For this esti-
of Cambridge, UK relation for weak lensing surveys mate of the redshift distribution to be
6
Department of Physics & Astronomy, unbiased, the spectroscopic sample has
University of Pennsylvania, USA Our cosmological model predicts the to be representative of the set of galaxies
7
German Centre for Cosmological growth of large-scale structure to be in question that results from some photo-
Lensing, Astronomical Institute, Faculty highly sensitive to the densities and fun- metric selection. For the estimate to be of
of Physics and Astronomy, Ruhr damental physical laws of the different sufficiently small statistical uncertainty,
University Bochum, Germany constituents of cosmic energy density, the spectroscopic sample has to be col-
8
Institute of Cosmology and Gravitation, among them the elusive dark matter, dark lected over a large enough volume and
University of Portsmouth, UK energy, and massive neutrinos. Photo- with high multiplicity. With current instru-
9
Catalunya Institute of Space Studies, metric surveys can directly probe the ments this is prohibitively demanding of
Barcelona, Spain large-scale matter density distribution via exposure time for weak lensing source
10
Institute of Space Sciences (ICE, CSIC), the apparent distortion of the shapes of galaxy samples. This is not just because
Barcelona, Spain galaxy images. These distortions are the galaxies in these samples are faint
11
Lawrence Berkeley National Laboratory, caused by tidal gravitational forces that and thus their spectroscopic redshifts dif-
Berkeley, USA act on the distant galaxy’s light along its ficult to acquire. In addition, the large vari-
12
Department of Astronomy, University of path to us, the so-called weak gravita- ation in the redshifts of galaxies selected
Geneva, Switzerland tional lensing effect. The strength of the from noisy, few-broadband photometry
13
Leiden Observatory, Leiden University, distortion depends not only on the matter means (i) that incompleteness can
the Netherlands distribution these experiments intend to strongly bias the sample if the success of
14
Hamburg Observatory, University of measure, but also on the distances to the a spectroscopic observation depends on
Hamburg, Germany lensed galaxies. This has motivated the the redshift of a targeted galaxy (for
15
IPAC, California Institute of Technology, development of a range of techniques to example, Gruen & Brimioulle, 2017), and
USA not just estimate photometric redshifts (ii) that a very large number of randomly
but also accurately calibrate the redshift sampled spectroscopic redshifts is
distributions of photometrically selected required to achieve a sufficiently small
the same photometric catalogue and over galaxies, for example the low-redshift representative but incomplete sampling
the same area as the Wide Area Vista (z < 0.2) filament that runs across the over a wider redshift range and to some-
Extragalactic Survey (WAVES), spanning right half of the left panel of Figure 1. This what fainter objects.
more than 1170 square degrees (Driver et results in a complementary proposal of
al., 2019). Unlike the high completeness acquiring redshifts that are representative
strategy of WAVES, 4C3R2 requests only (4C3R2) by colour and magnitude where Acknowledgements
a small subset of the potential targets. they cannot be complete (as for the This work was funded by the Deutsche Forschungs-
This allows for flexible fibre assignment WAVES selection) for as much of the gemeinschaft (DFG, German Research Foundation)
and a high degree of synergy. The two physical manifold of colour and redshift under Germany’s Excellence Strategy – EXC-2094 –
surveys are currently optimising their joint as possible. Specifically, 4C3R2 aims to 390783311. DG (daniel.gruen@lmu.de) and JM
(jmccull@stanford.edu) as corresponding authors
selection through simulations, including sample targets from cells whose redshifts acknowledge support by the Bavaria California Tech-
an exploration of whether entirely dis- have a high probability of being within the nology Center (BaCaTeC).
junct target selections or partial overlap range of 0.2 < z < 1.55 in wide, and simi-
between the two surveys’ samples is the larly 0.8 < z < 1.55 in deep. The lower
References
most efficient option. The preliminary limit is motivated by the WAVES selection,
choice of the limiting magnitude of while the upper limit is constrained by Buchs, R. et al. 2019, MNRAS, 489, 820
z = 21.6 in 4C3R2-Wide is slightly deeper spectroscopic features redshifting out of Driver, S. P. et al. 2019, The Messenger, 175, 46
than WAVES, with a z = 22 selection in 4MOST’s sensitivity wavelength range. Euclid Collaboration et al. 2020, A&A, 642, A192
Gruen, D. & Brimioulle, F. 2017, MNRAS, 468, 1, 769
4C3R2-Deep that matches WAVES-Deep. While colour-redshift calibration is the Hildebrandt, H. et al. 2017, MNRAS, 465, 1454
The left panel of Figure 1 depicts the primary purpose of 4C3R2, the survey Hildebrandt, H. et al. 2021, A&A, 647, A124
colour selection of 4C3R2 targets for both data will thus also optimally complement Masters, D. et al. 2015, ApJ, 813, 53
the wide and deep surveys. Cells that WAVES. 4C3R2 empirically determines Masters, D. C. et al. 2017, ApJ, 841, 111
Masters, D. C. et al. 2019, ApJ, 877, 81
contain WAVES targets, estimated accord- the targeting incompleteness of WAVES Myles, J. et al. 2021, MNRAS, 505, 4249
ing to existing photometric redshifts, are over the redshift ranges the latter intends Newman, J. A. & Gruen, D. 2022, ARAA, 60, 363
indicated by light colours. 4C3R2 inten- to cover. In addition, the 4C3R2 sample Rau, M. M. et al. 2022, arXiv:2211.16516
tionally does not select from areas in the extends the subset of galaxy evolution
map that are dominated by low-redshift science cases that can be studied with
Zdeněk Bardon (bardon.cz)/ESO
In this picture, Venus is shining brightly over ESO’s telescopes, built by Radboud University,
La Silla Observatory in Chile. The picture was taken the Netherlands Research School for Astronomy
just before dawn, towards the East, and also fea- (NOVA), and KU Leuven in Belgium. BlackGEM will
tures the diffuse zodiacal light — sunlight scattered search for the afterglow of some of the most dra-
by dust particles in the Solar System. The three matic events in the Universe, such as the collision
domes to the left of the road are the BlackGEM of black holes and neutron stars.
Christopher Haines 1 on the cold circumgalactic medium by current concordance ΛCDM cosmology
Yara Jaffé 2 targeting 50 000 cluster–QSO pairs. is that galaxy clusters continue to grow
Nicolás Tejos 3 with cosmic time, doubling their masses
Antonela Monachesi 4 since z ~ 0.5. In fact, a large fraction of
Emanuela Pompei 5 Scientific context cluster galaxies were accreted within gal-
Alexis Finoguenov 6 axy groups (McGee et al., 2009) and fila-
Cristóbal Sifón 3 Understanding what drives the evolution ments (Kuchner et al., 2020), which can
Sebastian Lopez 7 of galaxies and determines whether they ‘pre-process’ the galaxies prior to enter-
Amrutha Belwadi Manjunatha 1 end up as star-forming spirals or quies- ing the cluster (Zabludoff et al., 1996).
Lawrence Bilton 2,3 cent early-type galaxies by the present Recent observations have confirmed that
Johan Comparat 8 day remains a fundamental task within significant pre-processing of galaxies is
Rodrigo Cuellar 7 astrophysics. Both internal energetic indeed required (Haines et al., 2015;
Giuseppe D’Ago 9 mechanisms (like active galactic nucleus Bianconi et al., 2018), but large homoge-
Ricardo Demarco 10 feedback or stellar winds) and external neous studies that are both deep and
Ciria Lima-Dias 4 environmental processes are expected to wide are still rare. Hydrodynamical simu-
Elismar Lösch 11 play major roles in transforming galaxies. lations of galaxies around massive clus-
Paola Merluzzi 12 While most isolated galaxies remain as ters find a systematic depletion of both
Analía Smith Castelli 13,14 gas-rich star-forming spirals to the pres- hot and cold gas in cluster galaxies
Laerte Sodre 11 ent day, the bulk of galaxies within mas- caused by ram-pressure stripping, that
Erik Vinicius 11 sive clusters have lost their gas and have translates into a lower fraction of star-
and the CHANCES team been transformed into quiescent elliptical forming galaxies as far out as ~ 5r200
or lenticular galaxies. A variety of physical from the host cluster, where r200 is the
mechanisms capable of transforming gal- virial radius inside which the average den-
1
University of Atacama, Copiapó, Chile axies as they fall into the clusters have sity is 200 times the critical density (Bahé
2
University of Valparaíso, Chile been proposed, including gas stripping et al., 2013), motivating large-scale envi-
3
Pontificia Universidad Católica de by the intracluster medium (ICM; for ronmental studies.
Valparaíso example ram-pressure stripping and star-
4
University of La Serena, Chile vation), and gravitational interactions CHANCES1, the CHileAN Cluster galaxy
5
ESO among galaxies or between the galaxies Evolution Survey, is a 4MOST community
6
University of Helsinki, Finland and the cluster (Boselli & Gavazzi, 2006; survey designed to uncover the relation-
7
University of Chile, Santiago, Chile Cortese, Catinella & Smith, 2021). ship between the formation and evolution
8
Max Planck Institute for Extraterrestrial of galaxies and hierarchical structure for-
Physics, Germany Adding to the complexity is the fact that the mation as it happens, through deep and
9
Pontificia Universidad Católica de Chile matter distribution in the Universe evolves wide multi-object spectroscopy. It will tar-
10
University of Concepción, Chile dynamically within a web-like large-scale get cluster galaxies out to 5r200, the dis-
11
University of São Paulo, Brazil structure in which filaments connect and tance at which environmental effects act-
12
INAF–Capodimonte Astronomical feed massive galaxy clusters located at ing on infalling galaxies are expected to
Observatory, Naples, Italy their nodes. These structures are com- be sufficient to start removing their
13
La Plata Institute of Astrophysics posed of dark matter and galaxies, as well extended hot gas atmospheres, starving
(CONICET–UNLP), Argentina as diffuse matter that dominates the baryon them of future gas, and well beyond the
14
National University of La Plata, Argentina budget. In this picture, each galaxy is sur- maximum distance of 2–3r200 to which
rounded by a multi-phase, metal-enriched, ‘back-splash’ galaxies can reach.
and diffuse circumgalactic medium (CGM) CHANCES will permit the effects of
CHANCES, the CHileAN Cluster galaxy shaped by the interplay between accretion pre-processing of galaxies in infalling
Evolution Survey, will study the evolu- from the intergalactic medium and feed- X-ray groups and filaments (detectable by
tion of galaxies in and around ~ 150 back processes occurring within galaxies. eROSITA) to be quantified in unprece-
massive galaxy clusters, from the local Any attempt to comprehend galaxy for- dented detail. The survey will also have
Universe out to z ~ 0.45. It will target mation and evolution must therefore an important legacy value, by providing
~ 300 000 rAB < 20.5 galaxies with measure observational signatures of gal- detailed spectroscopic information which
4MOST, providing comprehensive spec- axy and CGM properties against a wide will complement X-ray observations of
troscopic coverage of each cluster out range of environments — both locally and clusters and groups by eROSITA, and
to 5r 200 in synergy with eROSITA. Its on large scales. future radio observations by the Square
wide and deep scope will trace massive Kilometre Array and the Australian SKA
and dwarf galaxies from the surround- There is a growing appreciation of the Pathfinder (ASKAP), as well as current and
ing filaments and groups to the cores of need to place massive galaxy clusters in planned large optical surveys such as the
galaxy clusters, enabling the study of their full cosmological context, not only to DESI Legacy Survey DR10 (LSDR10), the
galaxy pre-processing and the role of explain their own growth and assembly, Southern Photometric Local Universe
the evolving environment on galaxies. but also to understand the evolution of Survey (S-PLUS) and the Legacy Survey
We will also study the effect of clusters their member galaxies. A corollary of the of Space and Time (LSST).
Figure 1. An example of a
(c) CGM? CHANCES target from the Low-z
sub-survey: the A3391/A3395
z = 0.0555 g alaxy cluster system at z ~ 0.05
z = 0.0511 observed by eROSITA (panel a;
from Reiprich et al., 2021) and
Legacy Survey DR10 (panel b;
from Sky Viewer). The large-scale
r200 structure around the system and
the many substructures visible
are ideal laboratories in which to
study the effect of pre-processing
of galaxies. Panel c represents
our CHANCES- CGM sub-survey
(although at z > 0.35), in which we
explore the effect of the environ-
z = 0.0518 ment in the circumgalactic medium
(CGM) of galaxies, poorly under-
r200 stood to date.
z = 0.0517
(a) (b)
Specific scientific goals veys from ASKAP and MeerKAT, since ters beyond z = 0.3. While WINGS and
most HI detections from these surveys LoCuSS have allowed us to quantify
CHANCES comprises three sub-surveys, will be from dwarf galaxies, owing to their where galaxies are being transformed
described below, each with different sci- high gas fractions. Their large HI discs within clusters and on what timescales at
entific objectives. are highly susceptible to gas and tidal a single epoch, CHANCES will permit us
stripping processes, making HI a sensi- to extend these analyses over four billion
The CHANCES Low-z sub-survey will tive trace of environmental effects. years of cosmic time within a single over-
quantify the impact of environment on arching survey.
galaxies from high masses (1011 M☉) The CHANCES Evolution sub-survey
down to the dwarf regime (10 8–9 M☉). aims to continuously track the evolution The CHANCES CGM sub-survey will
Dwarf ellipticals (dE) are the numerically- of cluster galaxies over the last four billion study the effect of group and cluster envi-
dominant population in galaxy clusters years, and measure when, where and ronments on the diffuse gaseous content
(Binggeli, Sandage & Tammann, 1988). how quickly spiral galaxies are being of galaxies traced by MgII absorption
Their formation via cluster-related pro- transformed in and around clusters. Clus- towards projected QSOs at redshifts
cesses and evolution remains largely ter galaxies have not always been as 0.35 < z < 0.7. Although MgII probes cold
unexplored outside local clusters such inactive as they are at the present epoch. (104 K) gas, given that the CGM is multi-
as Virgo and Fornax (for example, The fraction of blue (star-forming) galax- phase, it provides a good opportunity to
Eigenthaler et al., 2018; Choque-Challapa ies among cluster members increases test models of gas disruption in clusters
et al., 2021), as is their evolution in the from almost zero in the local Universe to as a function of cluster- and absorber-
infall regions beyond r200. 20% by z ~ 0.4 (the Butcher–Oemler centric distances (for example, Dutta,
effect; Butcher & Oemler, 1984), implying Sharma & Nelson, 2022). CHANCES
This sub-survey will provide a large-scale a rapid evolution in the cluster galaxy CGM is similar to the ‘quasars behind
homogenous spectroscopic dataset for population. Moreover, since more than clusters’ survey (Lopez et al., 2008), the
galaxies covering more than three orders half of the galaxy populations of local difference being that special attention
of magnitude in stellar mass, in and clusters were only accreted after z ~ 0.4, is given to biases and selection effects
around a representative sample of clus- the transformation of star-forming spirals by positioning fibres not only in cluster-
ters. It will enable us to constrain many into quiescent early-types as they selected sightlines but also on control
open questions, including: the efficiency encounter group or cluster environments samples. In addition, and in contrast to
of cluster mechanisms transforming gal- is mostly taking place at late epochs. The what has been previously done, CHANCES
axies; the timescales for quenching star previous large cluster galaxy evolution CGM will also provide a wealth of infor-
formation; the long-term survivability of surveys (WINGS, LoCuSS) were limited to mation on the projected population of
dwarfs in groups and clusters; the role of a single epoch, while attempts to track galaxies close to the QSO sightlines over
pre-processing in groups and filaments; the evolution of cluster galaxies over CGM scales. Results from this survey
and the contribution of galaxy disruption 0 < z < 0.5 have had to resort to aggre- will have implications for the relation
to intracluster light. There is also a natural gating heterogeneous datasets from the between dark matter halos and the prop-
synergy with ongoing and future HI sur- literature, and with only a handful of clus- erties of galaxies in these dense and
Franz E. Bauer 1,2 16
Space Telescope Science Institute, to weaker accretion, lower AGN-to-host
Paulina Lira 3 Baltimore, USA light ratios, broad absorption lines and
Timo Anguita 4,2 17
New York University Abu Dhabi, UAE modest dust obscuration) have yet to be
Patricia Arevalo 5 18
Lyon Astrophysics Research Centre, systematically targeted in large numbers
Roberto Assef 6 France beyond z ~ 0.1. More inclusive selection
Felipe Barrientos 1 19
Roma Tre University, Italy techniques based on variability and/or
Trystyn Berg 7 20
INAF-OAS, Bologna, Italy ultraviolet–mid-infrared (UV–MIR) spectral
Santiago Bernal 5 energy distributions (SEDs) have been
Fuyan Bian 8 shown to be more effective than optical
Médéric Boquien 9 4MOST-ChANGES will target a legacy colours alone (for example, Peters et al.,
Veronique Buat 10 sample of active galactic nuclei (AGN), 2015; Tie et al., 2017; Sánchez-Sáez et
Igor Chilingarian 11,12 based on optical continuum variability al., 2019) in selecting such ignored AGN
Paolo Coppi 13 and spectral energy distribution (SED) populations. Yet the largest spectroscopi-
Demetra De Cicco 14 selection from several existing sur- cally confirmed variability-selected sam-
Yaherlyn Diaz 6 veys, and ultimately complemented by ples number ~ 13 000 AGN from SDSS
Kirill Grishin 15,12 Rubin LSST to: 1) constrain the low- Stripe 82 (~ 14% reddened fraction;
Lorena Hernandez-Garcia 2,5 MBH, low-L/LEdd end of the accretion and Peters et al., 2015) and 1263 AGN from
Darshan Kakkad 16 black hole (BH) density functions to the Dark Energy Survey (DES) supernova
Ivan Katkov 17,12 z ~ 1, and, by extension, BH seed mod- fields (6% reddened fraction; Tie et al.,
Jens-Kristian Krogager 18 els; 2) investigate correlations among 2017), probing a tiny fraction of the parent
Elena López-Navas 5 AGN (MBH, L/LEdd, ultraviolet slope, out- population. More recent studies with
Laura N. Martínez-Ramírez 1,2 flows, variability) and host properties La Silla-QUEST (LSQ) and the Zwicky
Chiara Mazzucchelli 6 (stellar age, metallicity, kinematics); 3) Transient Facility (ZTF) to rAB ~ 20–21
Veronica Motta 5 confirm/characterise rare BH sub recover reddened fractions of 10–15%
Federica Ricci 19 samples (extreme variability, tidal dis- (for example, Sánchez-Sáez et al., 2019),
Claudio Ricci 6 ruption events, lensed, intervening while deeper high-cadence VLT Survey
Alejandra Rojas 9,6 absorption line systems) for detailed Telescope (VST) imaging in COSMOS
Benedict Rouse 1 multi-wavelength follow-up studies. and CDF-S to rAB ~ 23.5 (for example,
Paula Sánchez-Sáez 8 De Cicco et al., 2021) identify ~ 300 variability-
Victoria Toptun 12 selected AGN deg –2, with reddened AGN
Ezequiel Treister 1 Scientific context fractions of ~ 30% (the majority of which
Fabio Vito 20 notably have X-ray fluxes below the
The last ~ 60 years have seen us go eROSITA all-sky survey limit). The latter
from discovering the first black hole studies imply that the SDSS and DES
1
Pontificia Universidad Católica de Chile, (BH) in Cygnus X-1 to detecting ≳ 10 6 variability samples still suffer strong
Santiago, Chile supermassive BHs (SMBHs, ≳ 10 5 M⊙) biases and represent just the tip of the
2
Millennium Institute of Astrophysics, and hundreds of intermediate-mass iceberg of a much larger population
Pontificia Universidad Católica de Chile, BH (IMBHs, ~ 10 –10 5 M⊙) candidates awaiting confirmation with 4MOST.
Santiago, Chile across the Universe (for example, Greene,
3
University of Chile, Santiago, Chile Strader & Ho, 2020; Flesch, 2021). Never- Among upcoming surveys, the Vera C.
4
Universidad Andrés Bello, Santiago, theless, critical portions of the active Rubin Observatory (Rubin) Legacy Survey
Chile galactic nuclei (AGN) discovery space of Space and Time (LSST) is expected to
5
Universidad de Valparaíso, Chile remain poorly explored. The overwhelm- have a truly profound impact on AGN sci-
6
Diego Portales University, Santiago, ing majority of published AGN optical ence. With its ~ 18 000-deg2 footprint,
Chile spectra (≳ 600 000; for example, Pâris et ugrizy coverage (0.4–1 μm), ~ 1–3-day
7
Milan-Bicocca University, Italy al., 2018) stem from the galaxy (z ≲ 0.1) temporal sampling, and rAB ~ 24.5 single-
8
ESO or quasi-stellar object (QSO, MB < – 23; epoch 5σ-depth, the LSST will produce a
9
University of Antofagasta, Chile blue-colour-selected) samples of the high-purity sample of > 2 × 107 AGNs
10
Marseille Astrophysics Laboratory, Sloan Digital Sky Survey (SDSS), and between z = 0 and z = 7, with > 50%
France hence have poor overlap with key south- identified in the first 1–2 years. Such
11
Center for Astrophysics, Harvard and ern hemisphere observatories (for exam- numbers will quickly surpass past/current
Smithsonian, USA ple, the Very Large Telescope [VLT] AGN samples, extending to lower lumi-
12
Sternberg Astronomical Institute, M.V. and the Atacama Large Millimeter/sub nosities, higher extinction E(B-V) ~ 0.3-1,
Lomonosov Moscow State University, millimeter Array). And while SDSS has and higher z, and enlarging rare AGN
Russia likely discovered ~ 80% of bright type 1 samples. Each AGN will ultimately have
13
Yale University, New Haven, USA QSOs with iAB < 19 out to z ~ 2 (where ≥ 800 visits (≥ 130 per filter) over the
14
Federico II University of Naples, Italy the BH accretion density peaks; for 10-year survey, enabling unique analyses.
15
APC Laboratory, University of Paris, example, Aird et al., 2015), fainter and/or Deep LSST ugrizy imaging, particularly in
France redder AGN (g–r > 0.6; for example due concert with Euclid YJHK imaging, will
Table 1. Properties and selection criteria of the a sample of ~ 1.5 × 10 6 AGN with and how lower-mass BH growth occurs
ChANGES subsurveys.
rAB ≲ 21. Our SED selection derives from (as input for the next-generation Event
DELVE-DR2, VHS-DR6, VIKING-DR2, Horizon Telescope, Athena, and LISA),
offer relatively accurate photometric red- and catWISE2020 data using SED-fitting and quantify the role of downsizing and
shifts and spatially resolved SEDs for a techniques (for example, Assef et al., feedback processes in the establishment
substantial fraction of its anticipated 2010; Boquien et al., 2019), aiming for of AGN scaling relations.
~ 2 × 10 9 extragalactic detections. This ~ 30% coverage from a target pool of
will ultimately allow statistical compari- ~ 1.8 × 10 6 AGN with rAB ≲ 23. We aug- Novel time-domain spectral synergies.
sons between AGN and inactive galaxies ment these samples with ~ 10 000 multi- AGN vary on nearly all timescales. Our
across redshift, environment, luminosity, wavelength selected AGN in four 4MOST/ experimental design will yield 4MOST
and some (deblended) host properties. LSST deep-drilling fields (DDFs) and spectra for up to ~ 1.2 × 10 6 AGN, which
However, in order to extract detailed ~ 150 000 host-dominated AGN or unique when combined with intensively sampled
physical constraints (for example precise BH-related phenomena selected using LSST multi-colour light curves will: 1)
stellar properties, BH masses, accretion the LSST’s unparalleled variability con- relate their variability and SED/spectral
rates, line kinematics), and to study straints during its first few years of opera- properties; 2) enable precision measure-
accretion physics and address funda- tion (see below). The resulting spectro- ment of continuum lags for type 1 AGN;
mental evolutionary questions, spectros- scopic observations will characterise a 3) gain insight into the unobservable
copy is required. To this end, the Chilean large, representative population of relatively far-UV regime based on the emission-line
AGN/Galaxy Extragalactic Survey unobscured (i.e., potentially reddened), strengths of high ionisation species; and
(ChANGES) will acquire ≈ 1.79 Mhrs of lower-MBH, low-MBH/Mhost-ratio, and/or 4) extend our database of labelled sys-
4MOST low-resolution spectroscopy for lower-L/LEdd AGN out to z ~ 0.5–1, that tems to optimise the LSST’s photometric
a large, representative sample of AGN have yet to be well sampled by any other classification and redshifts. We seek to
within the LSST footprint (≳ 10 000 deg2), method, owing to weaker emission and observe ~ 20% of the variability-selected
selected primarily via variability and opti- host domination. AGN sample (all with rAB < 19 and 50%
cal/NIR/MIR SEDs from existing imaging with 19 < rAB < 20) at least twice, to sys-
data within the LSST’s footprint. tematically study spectral variations in
Specific scientific goals tandem with the LSST’s unprecedented
optical monitoring. Our large samples
Target selection and survey area BH accretion rate (BHAR) and MBH den- will cover a broad span of AGN physical
sities, evolution and host synergies. The (z, MBH, L/LEdd) and variability (ampli-
Table 1 specifies our current survey moderately accreting AGN sample to be tude and characteristic timescale of their
design, indicating the survey regions and probed by ChANGES (10 –4 < L/LEdd < 10 –1; light curves) parameter spaces, allowing
input catalogues. The selection functions see Figure 1) comprises ~ 50–80% of the examination of how AGN spectral prop
for our two primary samples are driven by estimated total mass accretion onto BHs erties vary across these parameters
variability features and full optical-MIR in type 1 and mildly obscured AGN and and how they comply with standard
SED modelling, respectively, allowing us strongly complements other 4MOST AGN accretion disc predictions. Differences
to overcome past optical colour selection samples. The sample will be used to for low-MBH vs. low-L/LEdd AGN may
biases. Target completeness is further characterise the low-L/LEdd and low-mass probe when and how AGN transition to
weighted by the relative number of tar- contributions to the BHAR and MBH den- advection-dominated accretion flows.
gets across bins of colour, estimated red- sity distributions and push several land-
shift/luminosity, and variability timescale mark z ≲ 0.1 constraints out to z ~ 0.5–1, Completeness and obscured fraction
and amplitude. The variability sample is as functions of several, potentially inter- corrections. Although the optical–MIR
based on ZTF-DR11, Gaia-DR3, and LSQ dependent, driving parameters (for exam- selection of ChANGES will cover a wide
data using machine-learning selection ple redshift, MBH, L/LEdd, M*, SFR, morphol- parameter space, it will still suffer from
(for example, Sánchez-Sáez et al., 2021), ogy, gas content, Mhalo, and environment). various forms of incompleteness (for
aiming to observe ~ 60% of sources from We aim to assess in better detail where example, ability to recover redshifts,
Eddington rate
MBH, L/LEdd, M*.
Low state ~10 – 2
Changing-type/EVAGN. Current sam- Standard disc
ples of extreme variability AGN (EVAGN;
ADAF
Δm/Δt > 1 mag yr–1) number in the thou-
sands (for example, Luo, Shen & Yang,
2020; López-Navas et al. 2023), with the Quiescent state ~10 – 3
LSST expected to uncover orders of
magnitude more. ChANGES will prioritise
up to ~ 100 000 such targets for single Magnetic field Jet outlflow
(and when feasible, multi-epoch) spectra
to establish an unprecedented sample for Highly quiescent
studies of detailed accretion physics and state + tidal >10 – 4
disruption event
constraints on ‘changing-type’ AGN
(Graham et al., 2020). The latter can be
compared to a well-characterised search
for changing-t ype AGN among ≳ 200 000
X-rays UV Radio
targets with high-signal-to-noise multi-
Optical
epoch splits from ChANGES’ Time-
Domain Spectral Synergy campaign to
establish absolute rates.
The Next-Generation Transit Survey (NGTS) is instruments. The telescopes focus on discovering
located at ESO’s Paranal Observatory in northern Neptune-sized and smaller planets, with diameters
Chile. This project is searching for transiting between two and eight times that of Earth. This
exoplanets — planets that pass in front of their image shows the NGTS enclosure in the day. The
parent star and hence produce a slight dimming of VISTA (right) and VLT (left) domes can also be seen
the star’s light that can be detected by sensitive on the horizon.
Jens-Kristian Krogager 1 13
Gran Telescopio Canarias, La Palma, Scientific context
Karen M. Leighly 2 Spain
Johan Peter Uldall Fynbo 3,4 14
Canary Islands Institute of Astrophysics, Atomic and molecular gas in galaxies
Kasper Elm Heintz 3,4 Tenerife, Spain
Sergei Balashev 5 15
Department of Physics, Middlebury Studying HI absorption in spectra of
Franz Erik Bauer 6,7 College, USA luminous background sources such as
Trystyn Berg 5,8 16
Inter-University Centre for Astronomy quasars remains the best way to probe
Hyunseop Choi 9 and Astrophysics, Pune, India neutral hydrogen in individual galaxies
Lise Bech Christensen 3,4 17
Centre for Extragalactic Astronomy, beyond the local Universe (z ≳ 0.1) until
Annalisa De Cia 10,11 Department of Physics, Durham the advent of the Square Kilometre
Sara Ellison 12 University, UK Array. Of particular interest for galaxy evo-
Stefan Geier 13,14 18
Institute for Computational Cosmology, lution studies are the so-called Damped
Eilat Glikman 15 Department of Physics, Durham Lyman-α Absorbers (DLAs; Wolfe,
Neeraj Gupta 16 University, UK Gawiser & Prochaska, 2005), whose large
Christina Konstantopoulou 10 19
Paris Institute of Astrophysics, France HI column densities (log[NHI cm –2] > 20.3)
Daria Kosenko 5 20
French-Chilean Laboratory for arise in self-shielded, galactic environ-
Cédric Ledoux 11 Astronomy, Santiago, Chile ments. The neutral gas at even higher
Sebastian López 8 21
Institute of Astronomy, University of column densities (log[NHI cm –2] ≥ 21.5) is
Bo Milvang-Jensen 3,4 Cambridge, UK sensitive to many physical processes,
Leah Morabito 17,18 22
Department of Astronomy & such as the formation of molecular hydro-
Palle Møller 4,11 Astrophysics, UCO/Lick Observatory, gen (H2) and subsequently star formation
Pasquier Noterdaeme 19,20 University of California, USA (Bird et al., 2014; Noterdaeme et al., 2014).
Max Pettini 21 23
Kavli Institute for the Physics and
Jason Xavier Prochaska 22,23 Mathematics of the Universe, Kashiwa, Another important aspect of galaxy evo-
Sandra Raimundo 4,25,26 Japan lution is the buildup of metals over time
Johan Richard 1 24
Dark Cosmology Centre, Niels Bohr and their redistribution in and around gal-
Raghunathan Srianand 16 Institute, University of Copenhagen, axies. Observations of DLAs offer a pre-
Ksenia Telikova 11,5 Denmark cise and model-independent way of
Donald Terndrup 27 25
Department of Physics and Astronomy, measuring metallicities at high redshift
Todd M. Tripp 28 University of California, Los Angeles, USA through the numerous metal absorption
Marianne Vestergaard 4,24 26
Department of Physics & Astronomy, lines observed in optical spectra of dis-
Tayyaba Zafar 29 University of Southampton, UK tant quasars (Prochaska et al., 2003;
27
Department of Astronomy, The Ohio De Cia et al., 2018).
State University, USA
1 28
CRAL, University Claude Bernard Department of Astronomy, University of Owing to the cross-section selection of
Lyon 1, ENS of Lyon, France Massachusetts, USA absorbers and the insensitivity of the
2 29
Homer L. Dodge Department of Australian Astronomical Optics, Lyman-a transition to temperature, DLAs
Physics and Astronomy, The University Macquarie University, Australia probe mostly the warm neutral medium
of Oklahoma, USA (T ~ 104 K). The cold neutral medium
3
The Cosmic DAWN Center, Denmark (T ~ 100 K) is best traced by H2, found in
4
Niels Bohr Institute, University of The 4MOST–Gaia Purely Astrometric only a few percent of the overall DLA
Copenhagen, Denmark Quasar Survey (4G-PAQS) will carry out population (Balashev & Noterdaeme,
5
Ioffe Institute, Saint Petersburg, Russia the first large-scale, colour-independent 2018). At high redshift, H2 is directly
6
Institute of Astrophysics and Centre for quasar survey selected solely on the detectable in optical spectra through
Astroengineering, Faculty of Physics, basis of astrometry from Gaia. Our main Lyman and Werner bands in the rest-
Pontifical Catholic University of Chile, objective is to quantify the selection frame UV (900–1100 Å). The simulated
Santiago, Chile effects of current colour-selected sam- quasar spectrum shown in Figure 1 (left
7
Millennium Institute of Astrophysics, ples. These colour-selected samples panel) illustrates the molecular and
Santiago, Chile bias our view of the neutral gas and its atomic absorption features expected in
8
Department of Astronomy, University of chemical enrichment because of dust our survey.
Chile, Santiago, Chile obscuration and reddening of optical
9
Department of Physics, University of colours. Moreover, the broad absorp-
Montreal, Canada tion-line outflows observed in quasars The impact of dust obscuration
10
Department of Astronomy, University of are under-represented by optical colour
Geneva, Switzerland selection. 4G-PAQS will provide the first If a significant amount of dust is present
11
ESO sample to overcome these challenges in the absorbing gas, the background
12
Department of Physics & Astronomy, and will constrain the physical and quasar will appear fainter and redder than
University of Victoria, Canada chemical properties of gas in galaxies it is intrinsically. Quasars with dusty
and quasars at cosmic noon. foreground systems are therefore more
Best model
75
50
25
0
4000 4250 4500 4750 5000 5250 5500 1250 1500 1750 2000 2250 2500 2750
Observed wavelength (Å) Rest wavelength (Å)
likely to fall below the flux limit of optically blue-shifted velocities. These winds Figure 1. Noiseless mock 4MOST quasar spectrum
featuring an intervening absorption system at z = 3
selected spectroscopic surveys. Such potentially contribute to the co-evolution
(left panel). The coloured marks indicate the absorp-
quasar surveys will introduce a bias of black holes and galaxies in a process tion lines that allow us to quantify the chemical and
against dust-rich absorption systems known as quasar feedback (for example, physical properties of the absorbing gas: Lyman-α,
(Pei, Fall & Bechtold, 1991). Since the Glikman et al., 2012 and references H2 lines and metal lines. The right-hand panel shows
the best-fit synthetic BAL model to the SDSS spec-
amount of dust extinction scales with therein). Another important population of
trum of a FeLoBAL quasar allowing us to extract
both metallicity and N(HI), the bias the overall quasar demographic is com- physical outflow parameters such as density, ionisa-
against dusty absorbers becomes a bias posed of intrinsically reddened quasars. tion parameter and outflow velocity (Choi et al., 2020).
against high-metallicity and high-N(HI) These red quasars have moderate red-
systems (Pontzen & Pettini, 2009). Tar- dening (E(B–V) ~ 0.25–1.5 mag), such
geted searches for quasars reddened by that their optical spectra are not fully dust example, Leighly et al., 2018). The BAL
foreground dusty absorbers have found obscured and still dominated by an AGN outflow velocity is observed to correlate
many such cases (Krogager et al., 2015). continuum and broad emission lines. Red with luminosity and because of the strong
The impact of dust bias in a purely flux- quasars may represent a distinct evolu- dependence of outflow energy on velocity
limited sample can be corrected using tionary phase between a merger-driven (outflow kinetic luminosity, Lkin ∝ v3), lumi-
statistical modelling; however, current starburst in a completely obscured AGN, nous quasars are potentially the strong-
spectroscopic samples (such as SDSS) and a normal, unreddened (i.e., blue) est sources of feedback among BALQs.
are also selected based on inhomogene- quasar (Urrutia, Lacy & Becker, 2008).
ous optical and near-infrared c olour crite- FeLoBAL quasars — a subset of low-
ria which are not easily corrected for. Large-scale optical surveys of quasars ionisation BALQs — have a myriad of
underestimate the fraction of both BAL absorption lines from Fe+ in their near-UV
Given the complex selection function of quasars (BALQs) and red quasars as a spectra (see Figure 1, right panel), and
current quasar samples, the only way to result of the imposed optical-colour crite- therefore they can be studied over a wide
make progress is by obtaining a refer- ria (for example, Glikman et al., 2012; redshift range from 0.8 to 4. This range
ence sample of quasars and DLAs that is Fynbo et al., 2013). This is further sup- includes the cosmic noon, where feed-
free from colour selection. Any fore- ported by studies of radio-selected sam- back may have been especially impactful.
ground dust obscuration effects can then ples reporting larger numbers of red- Sixty percent of red quasars show BALs
be corrected for in a statistical manner. dened quasars (for example, White et al., in their spectra, nearly all of which are
2003) and BALQs (Morabito et al., 2019). low-ionisation BALs or FeLoBALs.
The colour-selection effects limit our
Quasar feedback at cosmic noon understanding of the critical role that
these objects play in the Universe. Specific scientific goals
Quasars may play an important role in
shutting down star formation activity in The goal of 4G-PAQS is to obtain a colour-
galaxies and strong outflows are thought Broad absorption lines and their link to unbiased reference sample of quasars
to be a key ingredient during quasar red quasars and their foreground absorbers. With this
activity (Fabian et al., 2012). The broad core sample of quasars, we will address
absorption lines (BALs) observed in the Low- and high-ionisation BALQs offer an the scientific goals set out below.
rest-frame UV spectra of quasars are evi- array of absorption lines between 1020
dence for such powerful winds emerging and 3000 Å (rest-frame) that can be used
from the central engine given their large, to constrain the outflow properties (for
The unbiased cosmic mass density of optically selected samples, but it may be hole masses and accretion rates in a
metals from 2 < z < 3 as large as 40%, depending on the representative, unbiased sample will be
source luminosity (Bruni et al., 2019). crucial for quantifying any redshift evolu-
It has been shown that DLAs trace a sig- There is further observational evidence tion. If the red quasar phenomenon is
nificant fraction of metals over cosmic that BALQs are 3.5 times as common at an evolutionary phase, then determining
time, reaching ~ 100% of the inferred z = 4 as at z = 2 (Allen et al., 2011). This their fraction as a function of redshift
metal mass, ΩMet, TOT, at z = 4 (Péroux & may be evidence that black hole feed- would constrain the duration of that phase.
Howk, 2020). However, this fraction is back evolves, something that may be
observed to decrease with decreasing important for constraining feedback
redshift, based on optically selected models, if the fraction of quasars that Identifying rare quasars
samples. The colour-selection criteria of have BALs reflects the global covering
current optical quasar samples introduce fraction of the outflowing gas. Lastly, our purely astrometric selection
a redshift dependence on the dust bias with no assumptions about spectral
(Krogager et al., 2019) that directly affects shape will have significant potential to
the measured cosmic mass density of What are the gas properties of BAL discover outlying and rare objects that
metals, ΩMet. While ΩMet at z = 3 may be outflows? may be overlooked otherwise (Hall et al.,
underestimated by a factor of around 2002; Geier et al., 2019). Our survey
two, this increases to a factor of around We aim to characterise the physical aims to identify rare quasars in a repre-
three at z = 2. Such a redshift-dependent properties of the outflow in each quasar sentative sample and to quantify the
bias would counteract the observed to obtain better insight into the potential prevalence of such rare objects.
decrease in ΩMet / ΩMet, TOT. Our aim with interrelation between individual quasar
this survey is to constrain ΩMet as a func- subtypes (low-ionisation versus high-
tion of redshift from an unbiased sample ionisation BALs and red quasars). This Target selection and survey area
of nearly 2000 high-redshift DLAs. may provide a basis for understanding
the link, if present, between each quasar Candidate quasars for 4G-PAQS are
subtype and different galaxy merger selected purely on the basis that quasars
The unbiased absorption cross-section stages. We will use our novel spectral are stationary sources on the sky (Heintz
of HI and H 2 at z > 2 synthesis software SimBAL (Leighly et al., et al., 2018). We select sources from the
2018) on the sample of BALQs to trace Gaia catalog with proper motions and
Current numerical simulations are still the outflow properties and feedback over parallaxes consistent with zero at the 2σ
unable to reproduce the observed red- cosmic time. An example of the spectral level down to a limit of G < 20.5 mag. The
shift evolution of the HI absorption fitting is presented in Figure 1. Depending completeness of our survey is therefore
cross-section as inferred by the so-called on the absorption lines present, we can limited only by the fibre allocation effi-
line incidence, dn/dz (Bird et al., 2014; obtain excellent constraints on the physi- ciency (up to ~ 80%). However, since the
Hassan et al., 2020). If a dust bias is cal properties of the winds (ionisation fibre allocation does not depend on the
present in current observations, it is parameter, density) which in turn yield colour of the targets, this lower complete-
important to correct for this before draw- measurements of the outflow properties ness will not introduce any selection bias.
ing conclusions about feedback mecha- (location, mass outflow rate, kinetic lumi-
nisms in simulations. Moreover, simula- nosity; Choi et al., 2020). Since the number of stationary sources in
tions are only recently starting to resolve the Gaia catalogue increases drastically
the cold (H2) and warm (HI) neutral media when observing close to the Galactic
in cosmological volumes (for example, What is the true number of red plane, we have limited our survey area to
Nickerson, Teyssier & Rosdahl, 2019; quasars? high Galactic latitudes, b < – 60 deg.
Feldmann et al., 2022). Hence, constrain- Based on a pilot study carried out at the
ing the dn/dz of H2 and HI in an unbiased We aim to constrain black hole masses North Galactic Pole (Heintz et al., 2020),
sample will provide the first comparison and accretion rates and determine the we have estimated the stellar contamina-
to these new simulations. properties and representative fractions of tion to be around 40% over the entire
the overall quasar population (red, blue survey area.
and BAL sub-types). Intrinsically red-
What is the true number of BALQs? dened quasars make up ~ 20–30% of In total, 4G-PAQS will obtain a colour-
quasars (Glikman et al., 2012). However, unbiased sample of ~ 100 000 quasars
Another core goal of 4G-PAQS is to current representative samples are small with no prior selection on the basis of
quantify the unbiased, intrinsic fraction of (~ 150 objects), which makes it difficult redshift or spectral shape. Our sample
BALQs and the physical conditions of the to study any redshift evolution of the red will therefore be purely flux limited. To
outflowing material over the redshift quasar population. Since the red quasar reach the above scientific goals, we will
range 0.8 < z < 4. The fraction of BALQs phenomenon may be linked to an evolu- obtain spectra with a signal to noise ratio
depends on the sample selection: Hewett tionary phase with higher accretion rates of at least 5 Å –1 over 7700–8500 Å.
& Foltz (2003) infer a fraction of 20% in (Kim et al., 2015), measuring the black
This spectacular view of the VISTA telescope was survey telescope in the world and it is dedicated to curvature reduces the focal length, making the
taken from the roof of the building during the open- mapping the sky at near-infrared wavelengths. Its structure of the telescope extremely compact. VISTA
ing of the enclosure at sunset. The VLT is visible on primary mirror is 4.1 metres in diameter and is the can map large areas of the sky quickly and deeply.
the neighbouring mountain. VISTA is the largest most highly curved of its size. The extremely high
Celine Peroux 1,2 powerful synergy of absorption and the planets that we observe. Baryons
Andrea Merloni 3 emission diagnostics by observing a from the cosmic web accrete onto galax-
Joe Liske 4 large sample of background quasars to ies and cool into a phase which fuels star
Mara Salvato 3 probe the circumgalactic medium of formation, which in turn expels material
Ramona Augustin 5 foreground objects in the same sky in powerful outflows. The result is a multi-
Fabian Balzer 3 regions. The objective of the 2.8-million phase medium of pristine and enriched
Maria-Rosa Cioni 6 fibre-hour ByCycle project is ultimately material that lies in the immediate sur-
Johan Comparat 3 to characterise the physical processes roundings of galaxies, the so-called
Simon Driver 7 by which gas changes phases and circumgalactic medium. More globally,
Alejandra Fresco 3 travels into, through, and out of galaxies. the cosmic baryon cycle describes these
Antonella Garzilli 8 Such a study is essential to our under- processes of motion and phase transfor-
Aleksandra Hamanowicz 5 standing of the growth of structure in mation of the baryons (Tacconi, Genzel &
Anne Klitsch 9 the Universe. Sternberg, 2020; Peroux & Howk, 2020;
Jean-Paul Kneib 8 Walter et al., 2020).
Jens-Kristian Krogager 10
Dylan Nelson 11 Scientific context
Johan Richard 10 Specific scientific goals
Patricia Schady 12 Astronomers now know the basic constitu-
Yue Shen 13 ents of the present Universe: 73% dark Baryons as a function of time: a census
Roland Szakacs 1 energy, 23% dark matter, and 4% baryons. of the cosmic metals
Simon Weng 1 The term baryons is used to refer to the
Qian Yang 14 normal matter of the Universe. One of the Characterising the relationship between
and the ByCycle team great successes of the last decades is stars, gas and metals is a critical compo-
the excellent agreement between esti- nent of understanding the cosmic baryon
mates of the total comoving baryon den- cycle. Observations of the evolution of
1
ESO sity from Cosmic Microwave Background metals are key to constraining this global
2
Marseille Astrophysics Laboratory, (CMB) anisotropies (Planck Collaboration picture of the evolution of the Universe.
France et al., 2016), primordial nucleosynthesis The early accounting of the total metal
3
Max Planck Institute for Extraterrestrial (Cooke, Pettini & Steidel, 2018) and budget by Pettini (1999) found an order of
Physics, Garching, Germany fast radio burst dispersion measures magnitude shortfall in the comoving den-
4
University of Hamburg, Germany (Macquart et al., 2020). However unlock- sity of observed metals compared with
5
Space Telescope Science Institute, ing the physics of the normal matter con- those expected to be produced by the
USA tinues to represent a most intriguing stellar content of the Universe (Madau &
6
Leibniz Institute for Astrophysics, enigma in astrophysics and cosmology. Dickinson, 2014; Bouwens et al., 2020;
Potsdam, Germany Indeed, although the amount of matter Zavala et al., 2021). The ByCycle project
7
International Centre for Radio bound in stars grows with time, only a will observe hundreds of thousands of
Astronomy Research/University of minority of the baryonic material is found high-redshift quasar spectra with 4MOST
Western Australia there; even today, > 90% of the baryons at high resolution, providing a robust pic-
8
École Polytechnique Fédérale, are found in the gaseous phase of the ture of the metal content of the baryons
Lausanne, Switzerland Universe. Notably, progress in determin- in the Universe over much of its history.
9
Dark Cosmology Center, Denmark ing the baryons’ location and physical
10
CRAL, University Claude Bernard state will have an impact beyond under-
Lyon 1, ENS of Lyon, France standing the evolution of star formation Baryons as a function of space:
11
University of Heidelberg, Germany and galaxies. Indeed, the survival of characterising the circumgalactic
12
University of Bath, UK satellite sub-halos in cosmological hydro- medium
13
University of Illinois Urbana- dynamical simulations generating orders
Champaign, USA of magnitude more dwarf galaxies than Ultimately, a detailed map of the cool gas
14
Center for Astrophysics Harvard & observed — the ‘missing satellites prob- component around galaxies will charac-
Smithsonian, USA lem’ — relates to our capacity to accu- terise the dominant baryon reservoir
rately model galactic feedback. Further- associated with forming galaxies, as well
more, baryonic effects impact the dark as providing constraints on gas flows that
The term baryons refers to the normal matter distribution and thus the inference are an essential part of the current galaxy
matter of the Universe. Surprisingly, only of cosmological parameters from weak formation paradigm (Chen et al., 2021).
a minority of this normal matter (< 10%) lensing surveys (Semboloni et al., 2011; Recent findings indicate that the circum-
can be probed by observations of star- Chisari et al., 2018; Foreman et al., 2020). galactic medium is a major reservoir of
light from galaxies. The ByCycle project heavy elements with a mass rivaling, and
aims to study the remaining majority of From the Big Bang onwards, the baryons possibly exceeding, that of the galaxies
the baryons traced by the intergalactic collapse with dark matter to form the cos- themselves (Peeples et al., 2014). Hence,
gas. To this end, ByCycle will use the mic web, galaxies, stars and, ultimately, the gas around galaxies holds valuable
Foreground 1015
N MgII (log cm – 2)
0 galaxy 12
1014
– 20
1013
10
– 40 VISTA/4MOST
Observing limit
1012
N MgII ~ 1011 cm – 2
– 60
1011 8
– 60 – 40 – 20 0 20 40 60
Distance from centre (kpc)
1.2
Normalised spectrum
1.0
0.8
0.6
MgII quasar absorbers
0.4
0.2
0.0
2790.0 2792.5 2795.0 2797.5 2800.0 2802.5 2805.0 2807.5 2810.0
Wavelength (Å)
clues to the fueling of star formation and column density, which is tightly correlated Figure 1. Top left panel: Schematic view of the
observational setup. Absorption lines detected
feedback processes. The ByCycle project1 to the volumetric gas density that can be
against bright background quasars observed with
aims to characterise the circumgalactic detected (Rahmati et al., 2013), is set by 4MOST high-resolution fibres offer a unique oppor-
medium gas and metals of 1.5 million gal- the apparent brightness of the background tunity to study the distribution and chemical proper-
axies, active galactic nuclei, groups and sources and thus the detection efficiency ties of the circumgalactic medium gas of foreground
galaxies observed with 4MOST low-resolution fibres.
clusters. is independent of the redshift of the fore-
The top panel shows results from zoom-in FOGGIE 2
ground host galaxy. Figure 1 (right panel) simulations with exquisite resolution (~ 0.1 kpc scales)
illustrates the three orders of magnitude in the circumgalactic medium of galaxies (Peeples et
Building a sizeable legacy sample gain in magnesium (MgII) column density al., 2019; Augustin et al., 2021). The ByCycle project
is designed to probe the whole range of MgII column
that will be reached with the ByCycle
densities displayed. The bottom left panel presents a
Today, the circumgalactic medium is best survey compared to what is available today. 4MOST normalised mock quasar spectrum with typ-
probed by the powerful technique of We will use the powerful synergy of absorp- ical rest-frame MgII doublet lines. Right panel: Illus-
examining the gas in absorption against tion and traditional emission diagnostics tris TNG503 simulations of the circumgalactic metal
distribution (Szakacs et al., 2023). This is a 2D stack
background sources (such as quasars or by observing one million background qua-
of 200 z = 0.5 galaxies like the ones probed by ByCy-
gamma-ray bursts) whose lines of sight sars and 1.5 million foreground objects in cle. The red contour illustrates the current MgII col-
pass through an identified foreground the same sky regions with 4MOST. umn density detection limit from SDSS (Anand, Nel-
galaxy’s halo. The absorption lines offer son & Kauffmann, 2021), while the white contours
correspond to the MgII column density limit of ByCy-
the most compelling way to study the dis- The high-redshift element of ByCycle
cle. The white circle shows the r 200 radius, the Virial
tribution, chemical properties and mass focuses on several dozen thousand radius assuming a critical overdensity constant of
budget of the mixed halo gas by probing strong Lyman-α absorbers tracing both 200. The ByCycle project will provide an improve-
matter at different densities, metallicities neutral and ionised gas observed over ment of three orders of magnitude in the MgII column
density probed in the extended circumgalactic
and temperatures (Figure 1, left panels). a large area. The ByCycle high-redshift
medium of galaxies, thanks to its large multiplexing
In these quasar absorbers, the minimum quasars will be used to probe the capability and spectral resolution of R = 20 000.
Lyman-α absorbers tracing the neutral Finoguenov et al., 2019). These observa- length coverage. To assess both the
gas phase and its metal content with the tions will be stacked to measure the nature and redshift of a large sample of
goal of providing a modern metal census. radial profile of both Lyman-α gas and targets we revert to a prior, likelihood,
The metal content of gas traced by metals as a function of galactocentric and posterior probability in a Bayesian
Lyman-α absorbers will be modelled with radii (see left panels of Figure 1 and Turner analysis (see Yang & Shen, 2023). The
the column density of metal ions through et al., 2017; Chen et al., 2021). Specifically, selection algorithm assigns probabilities
Voigt-profile fitting (for example, Quiret et we will quantify the covering fraction of for quasars, galaxies and stars, and
al., 2016). We will determine the dust- the neutral gas and numerous ion spe- simultaneously calculates photometric
corrected and ionised-gas metallicity by cies (including MgII and CIV) as a function redshifts (Salvato, Ilbert & Hoyle, 2019) for
performing a full photoionisation model- of impact parameter. We will compute the all extragalactic sources. We make use of
ling (Hamanowicz et al., 2020) informed number density of metal absorbers down the latest Gaia release (DR3) to classify
by the presence of elements with various to small rest-frame equivalent width, thus known stars as well as the most recent
degrees of ionisation. By counting the probing the most diffuse and extended DESI Legacy Imaging Survey4 (DR10) in
number of detected systems of various gas. Together these measurements will concert to estimate photometric redshifts.
strengths, we will provide a robust esti- put direct limits on the clumpiness of the The comparison of the results with availa-
mate of the number density per unit red- medium and provide new constraints on ble quasar spectroscopic redshifts con-
shift of strong Lyman-α absorbers which the scales of metal mixing in the circum- firms that the resulting sample has both
will allow us to calculate their gas mass galactic medium (Nelson et al., 2020; high completeness and high purity, secur-
content, Ωgas. This quantity is also critical Augustin et al., 2021). These quantities ing the requirement of z > 0.55. This
for estimates of the meta-galactic ultra relate directly to the missing satellites computation results in a total of 3.2 mil-
violet background used in radiative trans- problem (Byrohl et al., 2021; Mitchell & lion quasars within the 4MOST project
fer models of the high-redshift Universe Schaye, 2022). Furthermore, these find- footprint (for example, with declination
(Kollmeier et al., 2014; Fumagalli et al., ings will put new constraints on state-of- < +5 deg). To ease scheduling, the
2017). With the measurements of the the-art cosmological hydrodynamic simu- ByCycle survey provides targets to
metallicity and gas mass, we will directly lations with increased spatial resolution in choose from a more extended catalogue
test whether the strong Lyman-α quasar the circumgalactic medium regions that of sources. In addition, the ByCycle sur-
absorbers at intermediate redshifts do aim to characterise the physical proper- vey has no requirement on complete-
indeed contain the reservoir of missing ties of the gas (van de Voort et al., 2019; ness, cadence of observations or selec-
metals. This novel census of the metal Peeples et al., 2019). We will further tion function. Instead, the strategy aims
mass density will provide a new perspec- assess the strength of the absorbers by at maximising the number of low- and
tive on the missing metals problem and computing the optical depth of the metal high-spectral-fibre pairs.
existing tensions with models (Yates, doublets for different phases of the gas
Peroux & Nelson, 2021). from cold, at temperature T = 104 K (MgII), A key component of ByCycle is the sharing
to warm, at T = 10 5 K (CIV). We will locate of the focal plane with other low-redshift
With the low-redshift part of the ByCycle the gas spatially and in velocity space extragalactic surveys. Indeed, ByCycle
project, we will crosscorrelate quasar with respect to the associated foreground will combine high-resolution fibre obser-
absorbers with galaxies detected in emis- galaxies (Turner et al., 2014; Chen et al., vations of bright background quasars
sion in the same fields. The primary tar- 2021). In so doing, we will probe the net with low-resolution fibre observations of
gets are magnesium (MgII) quasar effect of inflow and outflow interactions, foreground galaxies, active galactic
absorbers, that trace cooler gas at 104 K revealing recycled gas and the history nuclei, groups and clusters that are in the
and whose doublet nature makes their of circumgalactic medium metal enrich- same field. The ByCycle survey com-
signature in quasar spectra unambiguous. ment. We will constrain the source of prises both a deep and a wide compo-
The strong complementarity of absorp- ionisation, the dynamical state of the cir- nent. The former is driven by the deep
tion and emission diagnostics will be cumgalactic medium gas and hence the observations that will be carried out by
used to characterise the low-density gas total baryonic content of galaxy halos. low-resolution galaxy surveys in five dis-
which is key to our understanding of gal- We stress that baryonic effects have a tinct cosmological fields. The latter
axy formation and evolution. We will significant impact on the distribution of extends to an area designed in synergy
gather a sample of 1 million z > 0.55 qua- matter, which needs to be incorporated with other efforts to cover > 15 000 deg2.
sar spectra, exploring the low-density into ongoing and future weak lensing
gas traced by 250 000 MgII absorbers. experiments. In summary, the 2.8-million fibre-hour
The physical properties of the metal ByCycle survey will use the powerful
absorbers will be crosscorrelated with up synergy of absorption and emission diag-
to a million foreground galaxy spectra Target selection and survey area nostics by observing a sizeable sample
observed in the same fields (Driver et al., of background quasars and foreground
2019; Richard et al., 2019). An additional The ByCycle survey strategy requires an galaxies in the same fields. At high red-
600 000 active galactic nuclei and a priori knowledge that the quasar’s red- shift, we will build an unprecedented cata-
20 000 groups and clusters are ideally shift is greater than z = 0.55 to ensure the logue of Lyman-α absorbers, measuring
suited to probing the rarer, denser fore- detection of intervening absorbers along their dust-free and ionisation-corrected
ground objects (Merloni et al., 2019; its line of sight within the 4MOST wave- metallicity to reappraise the missing
This picture shows the grand skies of the Chajnantor creates a dramatic display of reds and blues, as well the Sun in line with a telescope. This large antenna is
plateau in the Chilean Atacama desert. The rare as a sun pillar — an optical phenomenon caused by a part of the Atacama Large Millimeter/submillimeter
sight of clouds in this typically dry and arid region ice crystals in the atmosphere — that emanates from Array (ALMA), which is co-owned by ESO.
Edward N. Taylor 1 10
ARC Centre of Excellence for All Sky result is a dataset with exceptional and
Michelle Cluver 2 Astrophysics in 3 Dimensions, Australia long-lasting legacy value.
Eric Bell 3 11
Cosmic Dawn Center, Denmark
Jarle Brinchmann 4 12
Hamburg Observatory, University of
Matthew Colless 5 Hamburg, Germany Scientific context
Hélène Courtois 6 13
Tartu Observatory, University of Tartu,
Henk Hoekstra 7 Estonia Census-class surveys of the low-z galaxy
Sheila Kannappan 8 14
School of Mathematics and Physics, population are central to the study of gal-
Claudia Lagos 9,10,11 University of Queensland, Brisbane, axy formation and evolution in at least
Jochen Liske 12 Australia three ways. First, empirical mappings of
Elmo Tempel 13 15
School of Physics and Astronomy, demographic distributions and scaling rela-
Cullan Howlett 14 University of Birmingham, UK tions at z ~ 0 are vital in the calibration and/
Sean McGee 15 16
South African Astronomical or validation of numerical simulations and
Khaled Said 14 Observatory, Cape Town, South Africa theoretical models. Second, local samples
Rosalind Skelton 16 17
Sydney Institute for Astronomy, School provide the essential t ~ 0 reference point
Madusha Gunawardhana 17 of Physics, University of Sydney, to anchor evolutionary studies over higher
Sabine Bellstedt 9 Australia redshifts. Third, statistically representative
Leslie Hunt 18 18
INAF–Arcetri Astronomical Observa- samples are critical as control groups for
Thomas Jarrett 19 tory, Florence, Italy studies of interesting or unusual classes or
Chris Lidman 5 19
Department of Astronomy, University of populations, for example group and cluster
John Lucey 20 Cape Town, South Africa populations, radio- or HI-selected samples,
Shadab Alam 21 20
Department of Physics, Durham transient hosts, etc.
Maciej Bilicki 22 University, UK
Anna de Graaff 23 21
Tata Institute of Fundamental Research, The last decades have shown how low-z
Wojciech Hellwing 22 Mumbai, India science is driven by — or can be limited
Sarah Leslie 7 22
Center for Theoretical Physics, Polish by — the availability of spectroscopic red-
Ilani Loubser 24 Academy of Sciences, Warsaw, Poland shifts and high-quality multiwavelength
Lucia Marchetti 19,25 23
Max Planck Institute for Astronomy, imaging over wide areas. Imagine a
Michael Maseda 26 Heidelberg, Germany Universe in which we have eRosita, LSST,
Moses Mogotsi 16 24
Centre for Space Research, North VISTA-VHS, Euclid, Roman, WISE, SKA,
Peder Norberg 20 West University, Potchefstroom, South and LIGO/LISA, but where low-z science in
Alessandro Sonnenfeld 27 Africa the south is limited to 6dFGS (Colless et al.,
Jenny G. Sorce 28,29,30 25
INAF–Institute for Radio Astronomy, 2001; KAB < 14.5), WAVES (~ 1000 deg2;
and the 4HS Team Bologna, Italy Driver et al., 2019) and/or the margins of
26
Department of Astronomy, University of DESI-BGS (Hahn et al., 2022).
Wisconsin–Madison, USA
1 27
Centre for Astrophysics and Department of Astronomy, School of In this context, the 4MOST Hemisphere
Supercomputing, Swinburne University Physics and Astronomy, Shanghai Jiao Survey (4HS) addresses an urgent need
of Technology, Melbourne, Australia Tong University, China for uniform, wide-area, local-Universe
2 28
Department of Physics and Astronomy, University of Lille, CNRS, France spectroscopy in the southern hemi-
29
University of the Western Cape, Institute of Space Astrophysics, sphere, to support low-z science with
South Africa University of Paris-Saclay, CNRS, flagship European and global facilities.
3
Department of Astronomy, University of Orsay, France Accordingly, 4HS is designed to provide
30
Michigan, USA Leibniz Institute for Astrophysics, the best possible description of the low-z
4
Institute of Astrophysics and Space Potsdam, Germany galaxy population in the southern
Science, University of Porto, Portugal Universe, with particular emphasis on
5
Research School of Astronomy & science activities that absolutely require
Astrophysics, Australian National The 4MOST Hemisphere Survey (4HS) will spectroscopy and/or spectroscopic (as
University, Canberra, Australia obtain uniform spectroscopy and red- opposed to photometric) redshifts.
6
Lyon Institute of Physics of the shifts for approximately six million galax-
2 Infinities, University of Lyon, France ies over ~ 2π steradians, and with high In particular, 4HS takes full advantage of
7
Leiden Observatory, Leiden University, and unbiased completeness for z < 0.15. the unique capabilities of the 4MOST
the Netherlands 4HS aims to 1) complete the map of mass facility and consortium structure to
8
Department of Physics and Astronomy, and motion in the Local Volume, 2) map achieve high and unbiased spectroscopic
The University of North Carolina at the influence of environment on galaxy completeness for low-z galaxies over the
Chapel Hill, USA evolution through overwhelming statis- widest possible area. Sharing the focal
9
International Centre for Radio tics, and 3) define the local (z < 0.15) gal- plane enables 4HS to efficiently obtain
Astronomy Research, University of axy reference sample for the era of LSST, high completeness, and stellar partners
Western Australia, Perth, Australia Euclid, and ASKAP/MeerKAT/SKA. The enable us to press into the historic Zone
main, PV, and 2MASS XSC sub-surveys are shown ness for a large statistical sample of
1) V
02 - P
SS
in black, red, and dark red, respectively. Note how z < 0.15 galaxies, 4HS will deliver the
SD
(2
DESI-PV
4MOST enables us to operate efficiently even deep
2MRS
(2021-26) comprehensive suite of environment met-
into the historic Zone of Avoidance for PV science.
B: 4HS in the context of major PV cosmology sur-
(2012) rics (for example, local density, halo
mass, situation with respect to group/
7
critical for controlling systematics between different WALLABY void/filament/node, etc.) that is needed to
z~
(2022-27) 6dFGS
surveys/teams. On its own, 4HS will be the largest (2014) advance the field. Further, 4HS will iden-
southern PV experiment by a factor of about 10 in 4HS
volume, and about 50 in the number of galaxies, and tify > 60 000 galaxy groups, including
the largest globally by ~ 30% in area and a factor of
4HS
(2024-29)
~ 30 000 in the low-to-intermediate mass
two in number. Together with DESI-PV, 4HS will range (1011–1013 M☉; multiplicities in the
complete the map of the cosmic density and velocity 100 Mpc/h
z~
0.1
5
range 1–10) where group and galaxy
fields in the Local Volume.
properties seem to be most sensitive to
the relative strengths of interacting pro-
of Avoidance. (Equally, 4HS supports z ~ 0.12 represents the limit of where the cesses (for example, Davies et al., 2019).
stellar halo science around the Galactic uncertainties in FP-derived distances, Thus 4HS is designed to provide a solid
caps and the Sagittarius stream.) As well which scale with redshift, begin to blow empirical basis to describe exactly how
as extending samples and areas for other up.) The unique contribution of 4HS will environment influences galaxy properties,
participating extragalactic surveys, be to map the cosmic velocity field to in order to challenge and constrain the
shared observing costs for common tar- distances of ~ 500 Mpc, covering virtually next generations of hydrodynamical sim-
gets means that 4HS carries efficiency the full southern extent of the Local ulations, and so to resolve the environ-
dividends within the 4MOST consortium Volume, and including the most massive mental processes/effects that most influ-
amounting to ~ 25–30%. cosmological structures (i.e., Laniakea, ence galaxy formation and evolution.
Shapley, Vela, etc.) that are found in
the south. When combined with the
Specific scientific goals cosmological density field (as traced Establish the benchmark local galaxy
by galaxy positions), this provides a reference sample for the next
The key science goals for 4HS are set out snapshot of the cosmic web in formation generation of wide-area surveys and
below. via gravitational collapse. 4HS will map large/high-resolution simulations
the late-time (< 3 Gyr) evolution of the
growth rate of structure through 5–10% With an increase in statistical power of
Map the mass, motion, and gravitational precision measurements of the parameter about seven over the Sloan Digital Sky
growth of structure in the southern combination f(z),σ8 which is composed of Survey and about 65 over the Galaxy And
Universe the cosmological growth rate and mass Mass Assembly (GAMA) survey, 4HS will
clustering parameters, in each of six establish the definitive benchmark sample
4HS will measure velocity dispersions to independent redshift bins over the interval of galaxies at z < 0.15, including ~ 400 000
enable distance and peculiar velocity (PV) 0 < z < 0.3. The relation between the galaxies in the 8.5 < log M* < 9.5 dwarf
determinations from the Fundamental cosmic density and velocity fields also regime, all the way up to the very rarest
Plane (FP) for up to ~ 650 000 early-type represents a direct measurement of and most massive galaxies. Relative to
galaxies at z < 0.12. (The soft limit of gravitation on the largest scales DESI-BGS, the southern advantage gives
4HS access to areas covered by LSST, the two WAVES-Wide fields this gives a Figure 1a shows a preliminary 4HS input
ALMA, and ASKAP/MeerKAT/SKA, plus total area of ~ 18 000 deg2. catalogue in the context of other current
extensive overlap with eRosita, and and future wide-area surveys. We also
with Euclid in the South Galactic Cap, to Our observing strategy is to obtain emphasise 1) the strong synergies
add to existing data from all of GALEX, approximately uniform spectroscopic between 4HS and each of VRO-LSST,
VISTA-VHS, and WISE. Following the depth for all targets, equivalent to Euclid, and ASKAP-WALLABY, 2) the
successful model of GAMA (Driver et al., ~ 18 minutes integration in nominally grey complementarity of 4HS in the south with
2022), and together with 4MOST partners conditions. This strategy guarantees DESI-BGS in the north, and 3) the lasting
including WAVES and CHANCES, 4HS > 98% redshift success per observation legacy value of uniform and highly com-
is the keystone galaxy redshift survey for our z < 0.15 targets. Further, we plete spectroscopic redshifts across the
necessary to establish a truly transfor- expect S/N ≳ 12 for a high fraction hemisphere.
mational laboratory for empirical studies (> 80%) of log M* > 10 early-type galaxies
of galaxy formation and evolution with at z ≲ 0.12 to enable the velocity disper- As a near-infrared-selected spectroscopic
comprehensive, panchromatic vision sion measurements that underpin our redshift survey of the southern hemi-
across the entire baryon cycle. PV cosmology aims. Having roughly uni- sphere, 4HS is more than three magni-
form spectral data quality brings other tudes deeper than 6dFGS for a factor of
significant science benefits: for example, > 50 increase in target density and num-
Target selection and survey area easy-to-characterise sensitivity limits for ber of redshifts. In combination with
emission-line science (star formation high-quality optical-to-infrared photometry
The 4HS input catalogue is derived from a rates, metallicities, emission-line diagnos- from LSST, Roman, Euclid, VISTA, and
reanalysis of JK imaging from the VISTA tics, etc.), and additional science oppor- WISE — plus eRosita X-ray and ultimately
Hemisphere Survey (VHS) plus several tunities using high-S/N subsets of the ASKAP/MeerKAT/SKA radio data — 4HS
other programmes to fill gaps as far as data (beyond our key PV cosmology sci- can equally be viewed as providing
possible. For target characterisation (for ence goals, including Galactic/Magellanic GAMA-like depth and panchromatic cov-
example, masses, star formation rates, dust attenuation, etc.). erage over > 60 times the area, to estab-
etc.), we currently include optical imaging lish an incredible legacy dataset.
from Pan-STARRS1 where available For PV cosmology, more than complete-
(Dec > –30°). We plan to extend to infrared ness, the object is to span the largest pos- As a PV cosmology survey, 4HS expands
imaging from WISE in the next year, and sible area/volume. To this end, we define a the current state of the art (for example,
then to LSST optical and Euclid infrared PV subsample as JAB < 16.5 and 0 < (J–K) Qin et al., 2018; Kourkchi et al., 2022;
imaging as they become available. Our < 0.3 (~ 65 deg –2), which extends beyond Howlett et al., 2022) by factors of > 60
photometric pipeline is built around the the footprint of the main survey to E(B–V) and > 10 in number and volume, respec-
Scarlet deblender (Melchior et al., 2018) to < 1.5, for a total of ~ 19 890 deg –2. Our tively (see Figure 1b). More importantly
obtain robust and signal-to-noise (S/N)- goal with this sub-survey is to capture than ‘factor of’ comparisons, in combina-
optimised extended source photometry. those early-type galaxies at z < 0.12 for tion with DESI in the north, and with com-
We are committed to releasing this new which we can obtain useful velocity dis- plementary data from ASKAP-WALLABY
photometry as a public resource before the persion measurements. As a means of at lower redshifts, 4HS will complete the
start of 4MOST survey operations. pre-selecting suitable FP targets from the map of gravitational mass and motion in
main sample, this selection is ~ 60–65% Local Volume.
The main 4HS sample is selected as complete and ~ 30–35% reliable. In con-
JAB < 18 and (J–K) < 0.45 (foreground cor- trast to the main sample, the J-magnitude
rected; AB), with a notional mean target selection for this sub-survey is applied to Acknowledgements
density of ~ 325 deg –2. J-band selection magnitudes as observed (i.e., without fore- We acknowledge the support of the 4MOST
is a nearly optimal basis for pre-selecting ground correction), to focus on targets Consortium, and Astronomy Data and Computing
stellar-mass-limited samples: 4HS is vol- where we can get sufficient S/N for a Services, Australia.
ume limited for log M* > 9.3/10.0/10.3 velocity dispersion measurement to
and z < 0.05/0.10/0.15. As a means to ≲ 0.1 dex. We anticipate this sub-survey
References
pre-select z < 0.15 galaxies with near will provide about an additional 30 000
total completeness, our simple (J–K) col- useful PV measurements over and above Colless, M. et al. 2001, MNRAS, 328, 1039
our selection is maximally transparent, the main sample, extending our area/ Davies, L. J. M. et al. 2019, MNRAS, 483, 5444
Driver, S. P. et al. 2022, MNRAS, 513, 439
and is only marginally less efficient than a volume for PV science by ~ 10%. Driver, S. P. et al. 2019, The Messenger, 175, 46
photometric z selection. For star/galaxy Hahn, C. et al. 2022, arXiv:2208.08512
separation, we use a combination of both Further, we include the 2MASS Extended Howlett C. et al. 2022, MNRAS, 515, 953
colour and resolved-ness in the VISTA Source Catalogue (Jarrett et al., 2000) as Jarrett T. H. et al. 2000, AJ, 119, 2498
Kourkchi, E. et al. 2022, MNRAS, 511, 6160
imaging, with the opportunity to include a supplemental sample to push to very Qin, F. et al. 2018, MNRAS, 477, 5150
information from Gaia. To cover the wid- low Galactic latitudes towards the Melchior, P. et al. 2018, Astronomy & Computing,
est possible area, the main survey foot- Galactic anticentre, through the 24, 129
print is defined as the area where E(B–V) Magellanic Clouds, and wherever there
< 0.3 in the Planck dust map; including are gaps in the VISTA imaging.
Thomas E. Collett 1 7
University of Valparaìso, Chile will follow up strong lens candidates
Alessandro Sonnenfeld 2 8
University of Oxford, UK discovered in the Euclid survey and the
Chris Frohmaier 3 9
Andres Bello University, Santiago, Chile Legacy Survey of Space and Time.
Karl Glazebrook 4 10
University of Groningen, 4SLSLS will provide pairs of redshifts
Dominique Sluse 5 the Netherlands for 10 000 strong-lensing galaxies
Veronica Motta 7 11
INAF–Capodimonte Astronomical (lenses) and background galaxies
Aprajita Verma 8 Observatory, Naples, Italy (sources). Velocity dispersions will also
Timo Anguita 9 12
École Polytechnique Fédérale, be measured for 5000 lenses. This sam-
Leon Koopmans 10 Lausanne, Switzerland ple will enable discoveries about the
Crescenzo Tortora 11 13
Midi Pyrénées Observatory, Toulouse, evolution of galaxies, the study of intrin-
Frederic Courbin 12 France sically faint objects and of the cosmo-
Remi Cabanac 13 14
University of Arizona, USA logical model.
Brenda Frye 14 15
University of Birmingham, UK
Graham P. Smith 15 16
Cantabria Institute of Physics,
Jose Maria Diego 16 Santander, Spain Scientific context
Bruno Alteiri 17 17
ESA – European Space Astronomy
Sebastian Lopez 18 Centre, Madrid, Spain General Relativity (GR) is based on the
Chris Fassnacht 19 18
University of Chile, Santiago, Chile principle that massive objects warp space-
Asantha Cooray 20 19
University of California at Davis, USA time. Because of this, light passing close
Ariel Goobar 21 20
University of California at Irvine, USA to massive objects is deflected. If the sur-
Dan Ryczanowski 15 21
Stockholm University, Sweden face mass density of the object is great
Stephen Serjeant 22 22
The Open University, UK enough, then multiple images of a single
Johan Richard 23 23
Lyon Astrophysics Research Centre, background source can form; this is the
Tommaso Treu 24 France regime of strong gravitational lensing.
Leonidas Moustakas 25 24
University of California at Los Angeles,
Rui Li 26 USA For each strong lens system, the images
Colin Jacobs 4 25
Jet Propulsion Laboratory, Pasadena, observed depend on the light profile of
Cameron Lemon 12 USA the background source, the mass and
Lucia Marchetti 27 26
University of Chinese Academy of mass distribution of the foreground lens,
Phillipa Hartley 28 Sciences & National Astronomical the cosmological distances between
Eric Jullo 29 Observatories, PR China observer, lens and source and the nature
Chien-Hsiu Lee 30 27
University of Cape Town, South Africa of gravity. Strong lenses are therefore
Simon Birrer 31 28
Square Kilometre Array Observatory, sensitive probes of both astrophysics and
Alexander Fritz 32 UK cosmology. As such, strong lenses have
James Nightingale 33 29
Marseille Astrophysics Laboratory, been used to constrain the masses and
Nicola Napolitano 11 France density profiles of galaxies (Auger et al.,
Andres Alejandro Plazas 34 30
NSF’s NOIRLab, USA 2010; Shajib et al., 2021), the dark sub-
Sandor Kruk 35 31
Stony Brook University, USA halo and field-halo populations (Vegetti et
Chiara Spiniello 8 32
Max Planck Institute for Extraterrestrial al., 2014; Ritondale et al., 2019), cosmo-
Claudio Grillo 36 Physics, Garching, Germany logical parameters (Collett & Auger, 2014;
Sherry Suyu 37 33
Durham University, UK Wong et al., 2020), the high-redshift lumi-
Anowar Shajib 38 34
Princeton University, USA nosity function (Barone-Nugent et al.,
Georgios Vernardos 12 35
ESA – European Space Research and 2014), the nature of high-redshift sources
Simon Dye 39 Technology Centre, Noordwijk, (Newton et al., 2011) and the validity of
Tansu Daylan 40 the Netherlands GR (Schwab, Bolton & Rappaport, 2010;
Jeffrey Newman 41 36
University of Milan, Italy Collett et al., 2018). For many of these
Stefan Schuldt 36 37
Max Planck Institute for Astrophysics, analyses, the shortage of suitable strong
Garching, Germany lenses is a major limiting factor.
38
University of Chicago, USA
1 39
University of Portsmouth, UK University of Nottingham, UK To date, several hundred galaxy-galaxy
2 40
Shanghai Jiao Tong University, Washington University in St. Louis, USA strong lenses have been discovered in
41
Shanghai, China University of Pittsburgh, USA heterogeneous searches of photometric
3
University of Southampton, UK and spectroscopic survey data (Myers et
4
Swinburne University of Technology, al., 2003; Bolton et al., 2006; Gavazzi et
Australia Almost all science that can be done al., 2012; Negrello et al., 2014; More et al.,
5
University of Liège, Belgium with strong gravitational lenses requires 2016; Hartley et al., 2017; Petrillo et al.,
6
Paris Institute of Astrophysics & knowledge of the lens and source red- 2019; Jacobs et al., 2019; Li et al., 2020).
Marseille Astrophysics Laboratory, shifts. The 4MOST Strong Lensing The reason known lenses are rare is
France Spectroscopic Legacy Survey (4SLSLS) because even the most massive galaxies
are only capable of deflecting light by an highlight three main science goals that (Schaller et al., 2015; Xu et al., 2017;
arcsecond or two. Only a small fraction of are a priority for the project team. Peirani et al., 2017). Current strong lensing
the sky has been observed to sufficient observations suggest a larger diversity
depth and with good enough image reso- in dark matter profiles (Oldham & Auger,
lution to identify a typical Einstein ring. Galaxy formation — the inner structure 2018), although the number of galaxies
Several ongoing, and forthcoming, wide of lenses for which such measurements are availa-
and deep sky surveys offer improved ble is still very limited. Confirming or ruling
depth, area and resolution compared to ark matter and baryons exist in compa-
D out this tension can be very useful for
existing data (Miyazaki et al., 2006; rable amounts within the Einstein radius improving the accuracy of the subgrid
Ivezic et al., 2008; Laureijs et al., 2011). (Auger et al., 2010) and as a result galaxy- physics recipes adopted in simulations.
These surveys have the potential to scale lenses are well suited to the study
increase the current galaxy-scale lens of both components. Strong lensing is a Strong lensing is one of the very few
sample by orders of magnitude (Collett, particularly useful tool to simultaneously methods available for simultaneously
2015; Marshall, Blandford & Sako, 2005). address questions related to the inner constraining galaxy stellar masses and
Between Euclid and the Vera C. Rubin structure of massive galaxies. dark matter density profiles inde-
Observatory Legacy Survey of Space pendently of a stellar population model.
and Time (LSST), up to 300 000 lenses 4SLSLS will statistically combine strong
are discoverable. However, these sur- What fraction of the cosmological baryon lensing measurements from a large set of
veys will not provide adequate redshift content is converted into stars? systems (for example, Sonnenfeld &
information for science: this is what Cautun, 2021). Our forecasts show that
4SLSLS will deliver. Virtually all current measurements of gal- 4SLSLS will constrain the average stellar
axy stellar masses are obtained by con- IMF normalisation of a sample of
With a large increase in the known strong verting observed luminosities into masses 5000 lenses with a precision of 5%, while
lens population, current work could be on the basis of synthetic stellar popula- measuring the average dark matter den-
extended to new regimes, including lower tion models which have not been cali- sity slope with a precision of 0.1 (see
lens masses, higher-redshift lenses and brated against model-independent stellar Figure 1). With such precision, we will be
intrinsically fainter sources. A critical mass measurements. These models able to settle the dark matter core vs.
component of any results derived from introduce a systematic uncertainty in the cusp issue in massive lens galaxies and
this increased discovery space is the total baryonic mass locked in stars. The put much tighter constraints on the stellar
confirmation of any identified strong lens main uncertainty is the stellar initial mass IMF of these objects than is currently
candidate with a spectroscopic redshift function (IMF), the choice of which can possible. Moreover, by combining lensing
for the lens and the source behind it. plausibly vary the stellar mass-to-light constraints with 4SLSLS velocity disper-
ratio of a galaxy by a factor of two. Con- sion measurements it will be possible to
Spectroscopy is needed for two main straining stellar mass-to-light ratios of address additional questions, such as how
reasons: galaxies provides a unique way of gaining the IMF varies as a function of galaxy prop-
1. F
eatures on the sky often look similar insight into the stellar IMF and the physics erties and within each galaxy as a function
to strong lenses but are not. Ring gal- of star formation. of position (Sonnenfeld et al., 2018).
axies, star-forming tidal features and
interacting galaxies are often particu-
larly hard to distinguish. Spectroscopy What is the inner density profile of the Probing the cosmological parameters
confirms that the putative lens is dark matter halo? with 4SLSLS
indeed closer than the putative arcs.
2. S
trong lensing observables are natu- The determination of the inner dark mat- he Einstein radius of a strong lens is a
T
rally in angular units (Einstein radius, ter density profile can provide unique function of lens mass and cosmological
unlensed angular size of source, lens constraints both on dark matter physics, distances. Measuring cosmological dis-
mass, etc). Converting them into physi- for example by testing the self-interaction tances as a function of redshift can teach
cal quantities requires precise spectro- scenario (Elbert et al., 2018), and on us about the underlying cosmological
scopic redshifts. baryonic physics: the dark matter distribu- model of the Universe, but doing this with
tion responds to processes such as adia- strong lenses requires additional informa-
batic contraction from gas infall (Gnedin tion to break the degeneracy between
Specific scientific goals et al., 2011), feedback from the central mass and distances. 4SLSLS will provide
supermassive black hole (Martizzi, Teyssier two pathways to break this degeneracy:
The primary purpose of 4SLSLS is to & Moore, 2013) and dynamical friction directly by providing a lensing-independent
enable the strong lensing community to (Romano-Díaz et al., 2008). Cosmological measurement of the mass through veloc-
exploit the large increase in sample sizes. hydrodynamical simulations generally ity dispersions, and indirectly by confirm-
Having a homogeneous sample of predict, in the centres of massive galax- ing exotic strong lenses with information
10 000 lenses with a well understood ies, dark matter profiles steeper than beyond a single Einstein radius. Grillo et
selection function represents a huge step the Navarro–Frenk–White model used to al. (2018) showed that velocity dispersions
forward for the community. Here we describe dark matter-only simulations can in principle be used to constrain the
from Euclid and a further ~ 8 per square References More, A. et al. 2016, MNRAS, 455, 1191
Myers, S. T. et al. 2003, MNRAS, 341, 1
degree from the LSST; see Collett, 2015).
Auger, M. W. et al. 2010, ApJ, 724, 511 Napolitano, N. R. et al. 2020, ApJL, 904, L31
By sharing the focal plane of a multi Barone-Nugent, R. L. et al. 2014, ApJ, 793, 17 Negrello, M. et al. 2014, MNRAS, 440 1999
object spectrograph with larger surveys, Bolton, A. S. et al. 2006, ApJ, 638, 703 Newton, E. R. et al. 2011, ApJ, 734, 104
strong lens redshifts can be obtained Collett, T. E. et al. 2012, MNRAS, 424, 2864 Oldham, L. J. & Auger, M. W. 2018, MNRAS,
Collett, T. E. 2015, ApJ, 811, 20 476, 133
efficiently. 4SLSLS targets will be discov-
Collett, T. E. & Auger, M. W. 2014, MNRAS, 443, 969 Peirani, S. et al. 2017, MNRAS, 472, 2153
ered across the LSST and Euclid Collett, T. E. et al. 2018, Sci, 360, 1342 Petrillo, C. E. et al. 2019, MNRAS, 484, 3879
Southern footprints. We will target lens Dye, S. et al. 2022, MNRAS, 510, 3734 Pettini, M. et al. 2002, Ap&SS, 281, 461
candidates with a photometric redshift Elbert, O. D. et al. 2018, ApJ, 853, 109 Rigby, J. R. et al. 2017, ApJ, 843, 79
Frye, B., Broadhurst, T. & Benítez, N. 2002, ApJ, Ritondale, E. et al. 2019, MNRAS, 485, 2179
consistent with zs < 1.5, where the OII
568, 558 Romano-Díaz, E. et al. 2008, ApJL, 685, L105
doublet falls out of the 4MOST wave- Frye, B. et al. 2012, ApJ, 754, 17 Schaller, M. et al. 2015, MNRAS, 451, 1247
length band. This represents about a Gavazzi, R. et al. 2012, ApJ, 761, 170 Schwab, J., Bolton, A. S. & Rappaport, S. A. 2010,
third of the LSST and Euclid sample. All Gnedin, O. Y. et al. 2011, arXiv:1108.5736 ApJ, 708, 750
Grillo, C. et al. 2018, ApJ, 860, 94 Shajib, A. J. et al. 2021, MNRAS, 503, 2380
but the most massive 4SLSLS targets will
Hartley, P. et al. 2017, MNRAS, 471, 3378 Sonnenfeld, A. & Cautun, M. 2021, A&A, 651, A18
be lens-source blends, with the fibre cen- Ivezic, Z. et al. 2008, SerAJ, 176, 1 Sonnenfeld, A. et al. 2018, MNRAS, 481, 164
tred to maximise source flux. Jacobs, C. et al. 2019, ApJS, 243, 17 Swinbank, A. M. et al. 2015, ApJL, 806, L17
Jones, T. et al. 2015, AJ, 149, 107 Vegetti, S. et al. 2014, MNRAS, 442, 2017
Jones, T., Stark, D. P. & Ellis, R. S. 2018, ApJ, Wang, X. et al. 2017, ApJ, 837, 89
863, 191 Wong, K. C. et al. 2020, MNRAS, 498, 1420
Acknowledgements
Laureijs, R. et al. 2011, arXiv:1110.3193 Wuyts, E. et al. 2014, ApJ, 781, 61
Li, R. et al. 2020, ApJ, 899, 30 Xu, D. et al. 2017, MNRAS, 469, 1824
TEC is funded by a Royal Society URF. This work
Livermore, R. C. et al. 2015, MNRAS, 450, 1812 Yuan, F., Bu, D. & Wu, M. 2012, ApJ, 761, 130
is supported by the European Research Council
Marshall, P., Blandford, R. & Sako, M. 2005, NewAR, Yuan, T.-T. et al. 2011, ApJL, 732, L14
grant 945536 LensEra. Data are available from the
49, 387
first author upon request. For the purpose of open
Martizzi, D., Teyssier, R. & Moore, B. 2013, MNRAS,
access, the authors have applied a Creative
432, 1947
Commons Attribution (CC BY) licence to any Author
Miyazaki, S. et al. 2006, SPIE, 6269, 62690B
Accepted Manuscript version arising.
P. Horálek/ESO
Patrice Martinez 1 understanding of the properties of extreme adaptive optics (XAO) supported
Mathilde Beaulieu 1 exoplanets’ atmospheres, for which few by a second-stage AO to reduce the tem-
Carole Gouvret 1 data are available. The majority of poral delay, an HCI system, and high-
Alain Spang 1 exoplanets so far known were discovered dispersion spectroscopy (HDS). Various
Aurelie Marcotto 1 by indirect methods and, despite the pro- activities are under way in the community
gress with the second generation of on these topics, for example SPHERE+
high-contrast imaging (HCI) instrumenta- (Boccaletti et al., 2020), HiRISE (Vigan et
1
Lagrange Laboratory, Côte d’Azur tion, direct imaging detections remain al., 2018), RISTRETTO (Blind et al., 2022),
Observatory, Côte d’Azur University, marginal (~ 1%). Directly imaging and KPIC (Mawet et al., 2018), MagAO-X
CNRS, Nice, France characterising a fairly large number of (Males et al., 2018), SCExAO (Lozi et al.,
exoplanets requires an optimal combina- 2018) as well as many laboratory-based
tion of telescope and instrument to pro- developments1.
The Segmented Pupil Experiment for vide extremely high contrast, very good
Exoplanet Detection (SPEED) optical sensitivity, and exquisite image quality In this context, the Segmented Pupil
test bed is now ready for operation. and stability. Among the target stars, low- Experiment for Exoplanet Detection
SPEED is dedicated to high-contrast mass M-stars are particularly attractive (SPEED)2 facility offers an ad hoc environ-
imaging at short angular separations because of the considerable number they ment to improve HCI considering the ELT
with segmented telescopes. Its optical represent within 20 light-years of the Sun, pupil fragmentation and to assess the
design allows a wide range of applica- and because they are arguably the most impact on control and control stability of
tions and the immediate goal is to common environment with the potential quasi-static speckles. SPEED is posi-
demonstrate a high-contrast dark hole to harbour life. Observing habitable plan- tioned within a three-body problem:
close to the stellar vicinity. This will be ets around nearby M-stars is of the great-
achieved by a combination of optimal est scientific importance but represents Figure 1. 3D CAO view of the SPEED test-bed.
wavefront shaping architecture, small an extreme instrumental challenge. The C olour code: telescope simulator and common path
inner working angle coronagraphy, and Planetary Camera and Spectrograph (yellow), visible path (blue) and near-infrared path
efficient complex field sensor for fine (PCS; Kasper et al., 2021) for ESO’s (red). Abbreviations: TTM — tip/tilt mirror, OAP —
off-axis parabola, ASM — active segmented mirror,
cophasing and dark hole generation. Extremely Large Telescope (ELT) must DM — deformable mirror, FM — flat mirror, DIC —
provide an imaging contrast of 10 –8 at dichroic, L — lens, SCC-PS — self-coherent camera-
15 milliarcseconds angular separation phasing sensor, FPM — focal plane mask, PIAA-M1
Science case and challenges from the star (1.5 to 2 l/D in the near- & PIAA-M2 — phase induced amplitude apodisation
mirror 1 & 2, LS — Lyot stop, APOGEE — visible
infrared) and 10 –9 at 100 milliarcseconds camera, NIT — near-infrared camera, Basler — pupil
Direct observation of exoplanets is chal- and beyond. A promising path to reach camera (visible and NiR), FF — fl ip-flop mirror, FW —
lenging but essential to the study and these capabilities is a combination of filter wheel.
FW4
FW3 TTM
L4
Fiber FM1
DIC
ASM
APOGEE Camera FF1 OAP0
L3 OAP2
SCC-PS
FM3
L2
DM1
DM2
L1
FPM1
FM4
FM2
OAP3
PIAA-M1
OAP4
Coronagraphic bench
PIAA-M2
LS
NIT Camera FPM2
FW5 OAP5
FF2 Basler Camera (NiR) ZYGO verifire
L5
FW6
FM5
Robert J. De Rosa 1 climate. When injected into the strato- the 1810s and the volcanic eruption of
Angel Otarola 1 sphere, SO2 is transformed into sulphuric Tambora was not recognised at the time.
Thomas Szeifert 1 acid (H2SO4) by the photochemical effect This connection was only realised after
Jonathan Smoker 1,2 in the presence of water vapour. While the eruption of the Krakatoa in Indonesia
Fernando Selman 1 Earth’s surface effectively cools because in 1883 (Royal Society, 1888), at a time
Andrea Mehner 1 the Sun’s short wavelength radiation is when news could be quickly reported via
Fuyan Bian 1 scattered by the sulphate aerosols in the telegraphy across the world.
Elyar Sedaghati 1 stratosphere, heat radiation from Earth’s
Julia Victoria Seidel 1 surface is efficiently absorbed by the same
Alain Smette 1 particles, resulting in an altered weather The eruption of the Hunga Tonga–
Willem-Jan de Wit 1 pattern and climate. The sulphate aerosol Hunga Ha‘apai volcano
particles in the stratosphere circulate glob-
ally and are only removed by precipitation The submarine HTHH volcano in the
1
ESO on timescales of several years. South Pacific erupted violently on 15 Jan-
2
UK Astronomy Technology Centre, uary 2022. An ash plume shot 57 km into
Royal Observatory, Edinburgh, UK In terms of the energy released, the the mesosphere, shockwaves rippled
world’s biggest volcanic incident in the through the atmosphere, and the eruption
past 1300 years was the eruption of triggered a tsunami with heights of more
The Hunga Tonga–Hunga Ha‘apai vol- Mount Tambora, in what is now Indone- than 19 m above sea level 3, causing mas-
cano erupted on 15 January 2022 with sia, in April 1815. In addition to the imme- sive infrastructure destruction on the
an energy equivalent to around 61 meg- diate and direct destruction, a huge dust nearby islands and the death of four peo-
atons of TNT. The explosion was bigger cloud entered the stratosphere, which ple in Tonga and two in Peru. The ener-
than any other volcanic eruption so far disrupted weather systems in 1816 and getic output from the volcano has been
in the 21st century. Huge quantities of for the following three years in the north- estimated to be approximately 61 mega-
particles, including dust and water ern hemisphere. The year 1816 was the tons of TNT equivalent (Diaz & Rigby,
vapour, were released into the atmos- second-coldest year on record since the 2022). Ocean floor maps showed that the
phere. We present the results of a pre- Middle Ages and is known as the ‘year volcano spewed out at least 9.5 km 3 of
liminary study of the effects of the without a summer’. The change in climate material in total4. By comparison, the
explosion on observations taken at was followed by famine, disease, poverty 1815 eruption of Tambora in Indonesia
Paranal Observatory using a range of and civil unrest, with many social and ejected more than 100 km 3 of erupted
instruments. These effects were not political consequences (D’Arcy Wood, material, the 1883 eruption of Krakatoa in
immediately transitory in nature, and a 2014; Behringer, 2019). The years after Indonesia 25 km 3, the 1991 eruption at
year later stunning sunsets are still the Tambora eruption also sparked the Mount Pinatubo in the Philippines
being seen at Paranal. imaginations of artists (for example Mary 5.5 km 3, and the CE 79 eruption of Mount
Shelley’s Frankenstein, the dark poetry Vesuvius 4 km 3. In the case of HTHH,
of Lord Byron, paintings by J. M. W. Turner 1.9 km 3 of material ended up in the
Introduction and Caspar David Friedrich, and music atmosphere, which caused the stunning
by Beethoven and Schubert)1,2. Interest- sunsets observed from Paranal following
Many astronomers were likely unaware of ingly, the link between climate change in the eruption.
the impact of volcanic eruptions on astro-
nomical observations when the Hunga
Pressure Anomaly (in 10-minute time intervals), hPa/hPa
10 000 2022-05 elevation of the Sun variability in the sky extinction. The sea-
below the horizon. The
2022 07 & 08 sonal variability is due to the change in
green line is a model for
8000 2022 09 & 10 the atmospheric absolute humidity and
March 2022, which esti-
2022-11 mates a column height barely visible cirrus condition induced
be
0.2
0.1
Volcano eruption date
0.1
0
59 300 59 400 59 500 59 600 59 700 59 800 59 900 60 000 59 300 59 400 59 500 59 600 59 700 59 800 59 900 60 000
MJD MJD
0.25 0.2
FORS2 RSPECIAL filter extinction
0.15
0.1
0.1
0.05
0.05
0 0
59 300 59 400 59 500 59 600 59 700 59 800 59 900 60 000 59 300 59 400 59 500 59 600 59 700 59 800 59 900 60 000
MJD MJD
path to the LHATPRO detector on Earth’s and barely visible cirrus, induced by the Figure 5. FORS2 extinction time series 5 between
1 January 2021 and 30 December 2022. The vertical
surface. We computed the atmospheric altiplanic winter atmospheric condition.
line in each subplot shows the date of the eruption.
brightness temperature using the atmos- The detection of the increase in water (top-left) bHIGH filter, (top-right) v HIGH filter, (bottom-
pheric model am (Paine, 2022) and find vapour, injected by the volcano in the left) RSPECIAL filter, (bottom-right) IBESSEL filter.
that a change in brightness temperature stratospheric level, was not possible
at the peak of the 183 GHz water vapour because of the low sensitivity of our
line induced by a 25 ppmv water vapour water vapour radiometers to emission Proud, S. R., Prata, A. T. & Schmauß, S. 2022,
layer, with respect to the background from such a high altitude. A more detailed Science, 378, 554
condition, is only around 0.07%, corre- analysis of the data is ongoing to deter- Royal Society 1888, The Eruption of Krakatoa: And
sponding to a change of 0.04% in precip- mine the composition of the aerosols and Subsequent Phenomena (London, Trübner &
Company)
itable water vapour, which is well below to monitor the longer-term effects of the Rufener, F. 1986a, The Messenger, 44, 32
the daily temporal variability of water explosion on astronomical observations. Rufener, F. 1986b, A&A, 165, 275
vapour at Paranal. Given the intensity of the eruption and Paine, S. 2022, The am atmopsheric model (v. 12.2),
the large amount of water vapour injected https://doi.org/10.5281/zenodo.6774376
in the stratosphere, it is believed the
Conclusions effects will last for several years. Links
1
The HTHH volcanic eruption had several Ian Ritchie’s website: https://www.ianritchie.org/
References the-year-without-a-summer
observable effects at Paranal Observa- 2
T he Guardian website:
tory, located more than 10 000 km away Behringer, W. 2019, Tambora and the Year without a https://www.theguardian.com/music/2016/
from the eruption site. These included Summer: How a Volcano Plunged the World into jun/16/1816-year-without-summer-dark-
measurements of the shockwave and a Crisis (Medford, MA: Polity Press) masterpieces-beethoven-schubert-shelley
3
striking change in the colour of the sky Burki, G. et al. 1995a, The Messenger, 80, 34 T he BBC website: https://www.bbc.com/news/
Burki, G. et al. 1995b, A&AS, 112, 383 science-environment-61567521
visible in routine calibration sky flats taken D’Arcy Wood, G. 2014, Tambora: The Eruption That 4
National Institute of Water and Atmospheric
during twilight at optical and infrared Changed the World (Princeton: Princeton Research website: https://niwa.co.nz/news/tonga-
wavelengths. The detection of variations University Press) eruption-confirmed-as-largest-ever-recorded
in the extinction, observed via standard Diaz, J. S. & Rigby, S. E. 2022, Shock Waves, 32, 553 5
FORS2 QC1 processed zeropoint and extinction
Freudling, W. et al. 2007, The Messenger, 128, 13 data: http://archive.eso.org/qc1/qc1_
star observations, is harder to assess Grothues, H.-G. & Gochermann, J. 1992, cgi?action=qc1_browse_table&table=fors2_
since the volcanic eruption took place The Messenger, 68, 43 photometry
during the austral summer. This is a Harrison, G. 2022, Weather, 77, 87 6
See the section FORS2 Absolute Photometry
period when we observe a natural varia- Kerber, F. et al. 2012, Proc. SPIE, 8446, 84463N Project at https://www.eso.org/sci/facilities/
Legras, B. et al. 2022, Atmos. Chem. Phys., 22, 14957 paranal/instruments/fors/doc.html
bility and increase of the atmospheric Millán, L. et al. 2022, Geophys. Res. Lett., 49, 99381
extinction owing to an increase in humidity Moreno, H. & Stock, J. 1964, PASP, 76, 55
Tereza Jerabkova 1 but rather by fellow PIs who have also to assign proposals to the reviewers.
Ferdinando Patat 1 applied for observing time, hence actually The ESO DPR in P110 and P111 aimed for
Francesca Primas 1 staying true to the very concept of peer expert peer review in which proposals
Dario Dorigo 1 review, in which the applicants and the were assigned to reviewers with expertise
Fabio Sogni 1 referees are at the same level. Gemini as close as possible to the science case
Lucas Astolfi 1 Observatory pioneered the adoption of the proposals. Thus, the expertise level
Thomas Bierwirth 1 of DPR for telescope time allocation in for each proposal-reviewer pair needed
Michael Prümm 1 their Fast Turnaround channel in 2015 to be evaluated and the assignment algo-
(Andersen, 2020). DPR was first consid- rithm was then responsible for an optimal
ered at ESO following a report by the Time proposal distribution that maximised the
1
ESO Allocation Working Group (Patat, 2018), match overall.
which turned into a successful DPR exper-
iment that was conducted on a voluntary
The European Southern Observatory basis alongside the panel review (Patat et Reviewer-proposal expertise score
(ESO) implemented a new paradigm al., 2019). DPR was deployed for ALMA
called Distributed Peer Review (DPR) as time allocation in 2021 for Cycle 8 (Meyer The Phase 1 tool (P1)1 introduced by
part of its proposal evaluation process et al., 2022), followed by ESO in 2022 for Primas et al. (2019) has revolutionised the
in Period 110. Under DPR, Principal Period 110. In this article we present a sta- way proposals are submitted to ESO. In
Investigators who submit proposals tistical analysis of the first two semesters. order to submit a proposal, users must
agree to review a certain number of have a User Portal account, where they
proposals submitted by their peers and In P110 and P111 proposals requesting must provide information about their
accept that their own proposal(s) are more than 16 hours of observing time career stage, affiliation, and scientific
reviewed by their peers who have also and special programmes (for example, expertise through keywords. These key-
submitted proposals in the same cycle. Targets of Opportunity, Calibration Pro- words are grouped into classes such as
This article presents a brief overview of posals, and joint XMM-Newton propos- Cosmology, Galaxies and Galactic Nuclei,
the DPR process at ESO, and its out- als) were evaluated by expert panels and Interstellar Medium, Star Formation and
comes based on data from periods 110 approved by the Observing Programmes Planetary Systems, and Stellar Evolution.
and 111. Committee (OPC). The 16-hour limit was Each proposal submitted through the P1
chosen to produce a balanced distribu- tool must include a subset of these key-
tion between DPR and panels, effectively words that describe the proposal’s science
DPR introduction and ESO DPR P110 reducing the load on the panels by a fac- scope (up to five and a minimum of three).
and P111 overview tor of two, while leaving about 80% of the
time under their control. The threshold In each case, the keywords are specified
The Distributed Peer Review (DPR) para- value was defined in the Call for Propos- by the applicant in decreasing order of
digm should be seen not only as an inno- als for the corresponding period, and it relevance. This information is then used
vative concept, but above all as a natural might be adjusted in future. The DPR to compose a knowledge vector for each
consequence of the increased number channel evaluated 435 proposals distrib- proposal and each reviewer. For each
of proposals requiring review. Different uted to 379 reviewers in P110 and 417 reviewer-proposal pair, it is then possible
opinions on whether the expert panel proposals with 362 reviewers in P111. to compute the scalar product of their
review model needed revision have been respective knowledge vectors, resulting
put forward more than once since its The DPR scheme was set up so that PIs in what we will refer to as the match
inception a couple of centuries back. (or delegated PIs) submitting proposals score. In other words, the match score is
However, the recently increased numbers that qualified for DPR had to agree — at a figure of merit describing how parallel
of applications for observing time made time of submission — to evaluate 10 pro- the two vectors are. The match scores
this issue more pressing and challenging posals per submitted (DPR) proposal. They range from 0 (no match), through 1 (match
for all large astronomical facilities and also had the option of selecting one of the within a science category), to 2 (perfect
their time allocation processes. To keep proposal co-Is as the delegated DPR full match). This information provides
logistical aspects manageable and, at the reviewer. Failing to do this by the set dead- essential input for the assignment process
same time, to ensure a high-quality pro- line would result in the automatic rejection and is used to ensure that proposals are
posal evaluation, peer review by standard of their proposal/s. This guaranteed that assigned in a controlled way to the most
expert panels has become a less and each proposal received 10 independent qualified reviewers.
less viable option, calling for suitable and grades (from 1 to 5) and comments.
sustainable alternatives.
Proposal-reviewer matching
The DPR concept was first introduced Expertise evaluation and proposal
and formalised by Merrifield & Saari assignment The proposal assignment is a thorough
(2009). The general idea underlying it and methodical process that takes into
is that submitted proposals are not An important aspect of any peer review account various factors to ensure the
reviewed by pre-selected expert panels process is the set of criteria that are used best possible match between proposals
# of 0 = 2065 (49%) same figure). The hori- common keywords for the reviewer-pro-
1500
zontal axis shows the posal pair. The final score value for the
match score, and the
proposal_distributor is more than three
vertical axis shows the
number of proposals times larger than in the case of the ran-
ESO DPR assigment
assigned in each score dom assignment. Another comparison to
1000 Σi score i = 6085.65
Asym. max = 6408.72
bin. The blue histogram gauge the performance of the matching
shows the assignment
# of 0 = 25 (1%) algorithm can be made with the best dis-
produced by the pro-
posal_distributor. In tribution one can have with the available
500 grey, we show a random set of reviewers and proposals. To this
assignment that is rep- end, for a given distribution of scores,
resentative of the overall
one can construct a hypothetical asymp-
score distribution.
totic assignment which maximises the
0 final score. This ideal proposal distribu-
0.0 0.5 1.0 1.5 2.0
Reviewer-proposal keyword match score
tion takes the ten best assignments for
each proposal, excluding conflicts, but
ignoring the boundary constraint that
2.0
Figure 2. The reviewer- each reviewer can have only 10 proposals
proposal match score
to assess. This therefore represents an
calculated based on
Reviewer-proposal expertise score
0.2
0.1
0.0
Not Somewhat Mostly Fully Not Somewhat Mostly Fully
DPR comment usefulness Panel comment usefulness
posals. To monitor the performance of responsible for providing useful feedback Summary and future prospects
the algorithm, the DPR proposal evalua- to the PIs. This feedback should clearly
tion interface requires each reviewer to point out the strengths and weaknesses DPR was deployed smoothly at ESO in
express their self-evaluated expertise of the proposal and provide indications periods P110 and P111, and the outcome
level (non-expert, intermediate, expert) for for improving it when applicable. For the is statistically comparable to that of the
each proposal assigned to them. Figure 2 first time in P110, we invited ESO PIs/dPIs expert panels. The user feedback on the
shows a comparison between the to evaluate the usefulness of the received usefulness of the comments received by
reviewer-proposal match score and the comments, on a voluntary basis. The PIs’ the PIs clearly favours DPR. In this
self-evaluated expertise level. For each response rate was the same for both the respect, it is worth noticing that the
self-evaluated expertise value, the match expert panels and DPR: around 30% pro- decreasing level of user satisfaction with
score distribution is plotted, with its mean vided their evaluation of the feedback the feedback provided by the panels was
value denoted by a cross. There is an they received. This fraction is not very one of the most quantitative drivers for
overall trend, indicated by a dashed line satisfactory, and measures to increase it considering DPR as a possible alternative
fitted to the mean values, showing a posi- in future semesters are under discussion. to the classical paradigm. It is reassuring
tive correlation between the two indica- However, the data already allow some to see that DPR has improved this situa-
tors. Typically, low match scores are con- basic analysis. tion. At the same time, the data seem to
sistently evaluated as ‘non-expert’ by the indicate that, following the introduction of
users. However, several users catego- Figure 3 shows the distribution of the col- DPR, the level of satisfaction with the
rised themselves as ‘non-expert’ while lected responses for DPR (left) and pan- panels’ feedback has increased, a benefi-
their match score was as high as 2, thus els (right). These first results indicate a cial result of the significant decrease in
indicating a perfect keyword match (the higher level of satisfaction for DPR. This the number of proposals reviewed by
same keywords specified in exactly the becomes even more evident when look- each panel member.
same order). These cases can have sev- ing only at the rejected proposals, which
eral possible explanations, such as incor- are arguably the most relevant cases for As recommended by ESO’s Scientific and
rectly assigned keywords (either in the this specific aspect. While in the case of Technical Committee, we will keep the
User Profile or in the proposal), keywords the panels most of the comments are same setup for P112 and P113, to collect
that were too broad, or subjective bias. judged as ‘not’ or ‘somewhat’ useful, an more information under the same homo-
In the data collected in P110, we noticed opposite trend is seen in the case of geneous conditions. Nevertheless, we are
that early-career scientists were less DPR. When considering this result, one planning to further monitor and improve
likely to claim to be an expert, even in must emphasise that the panels produce the process and to keep presenting our
their own field. On the other hand, senior one single joint comment for each pro- findings in order to remain as transparent
scientists, such as professors and staff posal, emerging from the discussion at as possible. More complex assignment
astronomers, were more likely to claim to the meeting, while in the case of DPR, algorithms are being tested (for example,
be an expert even in a field that is far each reviewer writes an independent Faez, Dickerson & Fuge, 2017; Stelmakh,
from their declared expertise. evaluation, which is passed directly and Shah & Singh, 2018) and we are working
unedited to the PI. Thus, in the case of on improving reviewer and proposal profil-
DPR the feedback on the proposals cap- ing via machine learning approaches
Feedback on the reviewers’ proposal tures more information, which is the likely (Patat et al., 2019; Kerzendorf et al., 2020).
comments explanation for the observed trend. We are also looking into ways of increas-
ing the rate of feedback return from the
In addition to grading proposals, both in PIs, in order to have a larger statistical
DPR and panels, the reviewers are basis for future analyses.
As part of the development process, elderly and conservative scientists from References
we organised the conference Peer- occupying crucial administrative positions.
Andersen, M. 2020, Nature Astronomy, 4, 646
Review Under Review, held at ESO’s Indeed, it was doubted if the problem Faez, A., Dickerson, J. P. & Fuge, M. 2017,
Headquarters in Garching, Germany, ever would be solved.” The promising Proceedings of the Twenty-Sixth International
on 6–10 February 2023, the outcome of outcome of the deployment of DPR at Joint Conference on Artificial Intelligence, 35
Kerzendorf, W. E. et al. 2020, Nature Astronomy, 4,
which will be presented in a future Mes- ALMA and ESO indicates that this may
711
senger article. The idea was to bring indeed be one viable and valid alternative Merrifield, M. R. & Saari, D. G. 2009, A&G, 50, 4.16
together not only the astronomers, but to the standard expert panel review pro- Meyer, J. D. et al. 2022, Bull. Am. Ast. Soc., in press
also representatives of the wider scientific cess. For once, we might be ahead of Patat, F. 2018, The Messenger, 173, 7
Patat, F. et al. 2019, The Messenger, 177, 3
community, in order to start a discussion science fiction.
Primas, F. et al. 2019, The Messenger, 176, 41
about peer-review processes in general, Stelmakh, I., Shah, N. B. & Singh, A. 2018,
and to identify ways to cope with a con- arXiv:1806.06237
tinuously growing scientific community Acknowledgements
and the appearance of intriguing new The authors are grateful to ESO’s Director General Links
technologies. and Director for Science and to Markus Kissler-Patig
for supporting this enterprise, which represents the 1
T he ESO P1 tool: https://www.eso.org/sci/
Back in 1973, in his science fiction novel most significant change in the proposal evaluation observing/phase1/p1intro.html
process at ESO since it was established. The sup-
Rendezvous with Rama, Arthur C. Clarke port and the positive attitude shown by the Scientific
wrote prophetically about an expert panel Technical Committee, the Users Committee and the
evaluating a proposal for a space mission: Observing Programmes Committee are also acknowl-
“Even by the twenty-second century, no edged. We would like to acknowledge the valuable
assistance of OpenAI’s language model, ChatGPT.
way had yet been discovered of keeping
ENEL
Located on the outskirts of the Chilean Atacama middle of a desert is in fact ESO’s first observatory,
Desert, 600 km north of Santiago and at an altitude La Silla Observatory.
of 2400 metres, this seemingly tiny village in the
Claudia Paladini 1
Konrad R. W. Tristram 1
Itziar de Gregorio-Monsalvo 1
1
ESO
Figure 2. Workshop photo with all participants. interferometrists in the respective fields. colleagues joined to listen to the various
In this way, the knowledge was also presentations.
18 students identified themselves as spread to more senior colleagues, and
female, 17 as male, and 2 as non-binary. collaborations were fostered. The setup The programme was organised to leave
Most of the students had no previous was very successful as some of the pro- enough time for networking and for
knowledge of interferometry. Despite the fessors decided to stay on during the in-person interactions between students
workshop’s being in person, all the lec- second week and join the practical ses- and lecturers during coffee breaks,
tures were broadcast in real time via sions themselves. This enabled them to lunches, and other activities: a welcome
Teams and about 10 additional students help the students with the scientific inter- reception on the first day, a barbecue on
joined the workshop lectures online. pretation of the data. the first Friday, an excursion on Saturday,
and a wine and cheese reception on
The second week was dedicated to prac- Wednesday of the second week. Several
Programme tical hands-on sessions. The students short talks during the first week were
were divided into groups according to dedicated to soft-skills development such
The first week of the workshop was dedi- their scientific interests. Each team had a as unconscious bias, scientific writing,
cated to lectures introducing the students VLTI expert available to work with them as job hunting, career development, job
to the basics of optical and infrared inter- tutor during the entire week. On Monday, applications and interviews, among others.
ferometry, the VLTI instruments, and the the students were assigned a scientific Since the workshop was held at the ESO
many tools commonly used for VLTI target observed with a specific VLTI offices in Santiago and the telescopes
observation preparation and data reduc- instrument; they were shown how to were too far away to visit, a live connec-
tion. Additionally, various scientific lec- browse the archive and how to download tion with Paranal was hosted on Thursday
tures were given covering different their data set, how to obtain and install of the second week. The students were
research areas where optical and infrared the respective data reduction pipeline, shown the control room and could
interferometry has made significant con- and how to reduce the data. On Tuesday, observe in real time during twilight. Cali-
tributions, such as exoplanets, active they used the tools presented during the brator observations were prepared by
galactic nuclei, massive young stellar first week for model-fitting. Wednesday some students, and some volunteers
objects, protoplanetary discs, stellar evo- was dedicated to image reconstruction; reduced the data in real time. Moreover,
lution, evolved massive stars and funda- the students attended a tutorial and then during the second week the last hour of
mental stellar parameters. The lectures received a special data set suitable for the day was dedicated to presenting vari-
were organised in two parts: first a gen- image reconstruction. At the end of the ous scientific profiles working at an
eral introduction to the scientific topic day each group made a five-minute pres- observatory: the operator, the engineer
understandable by everyone, and then entation of their results. On Thursday they working on the instruments, the astrono-
specific lectures presenting VLTI results continued working on their initial data mer. Special attention was paid to select-
on the given scientific topic. Several pro- set, finished any pending tasks, and pre- ing the speakers so as to maintain the
fessors from Latin America were invited pared their final presentation which gender balance and also include women
to give the general introductions, while was given on Friday in front of the whole from Latin America to act as role models
the VLTI lectures were given by expert school. Students, lecturers and ESO for the students.
This picture shows a beautiful meeting between the other. This shot was taken at ESO’s La Silla Obser-
Swedish–ESO Submillimetre Telescope (SEST) and vatory, located on the outskirts of the Chilean
the Milky Way, apparently almost touching each Atacama Desert, at an altitude of 2400 metres.
Fellows at ESO
Michael Abdul-Masih
After quite a bit of moping, I decided I As with my life plan, my interests in astron- Working at ESO has been a dream come
wasn’t done and I that I would see if it was omy have evolved quite a bit over the true. When showing my friends the moons
possible to continue. I spoke with the years. During my bachelor’s, I was primar- of Jupiter through my 8-inch telescope, I
heads of both the Astronomy and ily interested in astrobiology and how life never imagined that 10 years later, I would
Biochemistry departments and after a bit on other planets might look, bridging my get to point a telescope 40 times bigger
of finagling, I managed to come up with interests in biochemistry and astronomy. at those same moons. I feel truly lucky to
a schedule that would allow me to finish By the time I was applying for my mas- have the opportunity to work at one of the
both the biochemistry major and an ter’s, I was more interested in planetary largest and most advanced observatories
astronomy major on time. The day that I science and the detection of exoplanets. in the world. Being able to meet world-
submitted the form to officially declare the During my master’s, however, I worked leading experts from around the world,
astronomy major was one of the happiest with eclipsing binaries and realised that I getting to interact with new cutting edge
of my life. The rest is history. During the really enjoyed stellar astronomy. Continu- instruments, and seeing how a world-
final year of my bachelor’s, I decided that I ing in that direction, my PhD focused on a class observatory operates behind the
would try to continue in astronomy and particularly interesting class of eclipsing scenes has been an eye-opening experi-
applied for a few master’s programmes. I binaries known as overcontact binaries. ence. On a more personal level, working
was accepted at Rensselaer Polytechnic The stars in these systems are so close to at ESO has only enhanced my love for
Institute and started in September and each other that they are actually touching astronomy and the night sky. I am still an
graduated with my master’s a year and a and sharing material, making them roughly avid stargazer and the night sky on a
half later. From there, I accepted a PhD peanut shaped. Recently, my interests Moonless night at Paranal is unrivalled.
position at KU Leuven, where I spent the have broadened a bit and I have begun From time to time, I will grab a mug of hot
next four years. I defended my thesis in working on other classes of non-spherical chocolate and go out to the platform in the
2020 and now I can say that I am officially stars such as rapidly rotating stars and hopes of catching a meteor or two. The
a doctor, albeit a different doctor than I semi-detached binaries. only thing missing is a hammock.
originally planned!
ESO/M. Kornmesser, L. Calçada
This artist’s impression shows a two-star system the disc, the white dwarf uses its strong magnetic
where micronovae may occur. The blue disc swirling fields to funnel the hydrogen towards its poles. As
around the bright white dwarf in the centre of the the material falls on the hot surface of the star, it trig-
image is made up of material, mostly hydrogen, sto- gers a micronova explosion, contained by the mag-
len from its companion star. Towards the centre of netic fields at one of the white dwarf’s poles.