TARSIS. Science Document

Code: TEC/TAR/007 Issue: 1.A Date: 17/05/2023 Pages: 60 TARSIS. Science Document TARSIS. Science requirements Document TEC/TAR/007 1.Α - 17/05/2023 2 Authors: Patricia Sánchez-Blázquez Revised by: Armando Gil de Paz Ana Pérez Calpena Marisa García Vargas Jorge Iglesias Alfredo Montaña Mónica Relaño José Oñorbe Approved by: Patricia Sánchez-Blázquez TARSIS. Science requirements Document TEC/TAR/007 1.Α - 17/05/2023 3 Distribution List: Name Affiliation Date Armando Gil de Paz UCM 17/05/2023 Jorge Iglesias IAA 17/05/2023 Mónica Relaño UGR 17/05/2023 Alfredo Montaña INAOE 17/05/2023 José Oñorbe US 17/05/2023 TARSIS. Science requirements Document TEC/TAR/007 1.Α - 17/05/2023 4 Acronyms: ACT Atacama Cosmology Telescope (ACT) BCG Brightest Cluster Galaxy CAHA Centro Astronómico Hispano en Andalucía CATARSIS Calar Alto Tetra-ARmed Super-Ifu spectrograph Survey CGM Circumgalactic Medium CMB Cosmic Microwave Background CoD Conceptual Design DES Dark Energy Survey DLA Damped Lyman Alpha EM Electromagnetic ETG Early-Type Galaxy FoV Field of View GW Gravitational Wave HSC Hyper-Suprime Cam IA Intrinsic Alignment IFS Integral-Field Spectrograph/Spectroscopy IFU Integral-Field Unit IMF Initial Mass Function LAE Lyman α emitter LMT Large Millimeter Telescope LSS Large Scale Structure MAR Mass Accretion Rate MAH Mass Accretion History MUSE Multi-Unit Spectroscopic Explorer NFW Navarro, Frenk & White (profile) NS Neutron Star PD Preliminary Design PDF Probability Distribution Functions PI Project Investigator PM Project Manager PS Project Scientist PSF Point Spread Function RPS Ram-Pressure Stripping SDSS Sloan Digital Sky Survey TARSIS. Science requirements Document TEC/TAR/007 1.Α - 17/05/2023 5 SED Spectral Energy Distribution SFG Star-Forming Galaxy SFH Star Formation History SFR Star Formation Rate SMBH Supermassive Black Hole SPT South Pole Telescope SSP Simple/Single Stellar Population SZ Sunyaev-Zel’dovich TARSIS Tetra-ARm Super-Ifu Spectrograph TTT Tidal Torque Theory WL Weak Lensing TARSIS. Science requirements Document TEC/TAR/007 1.Α - 17/05/2023 6 Change Control Issue Date Section Page Change description 1.A 17/05/23 All All First version Reference Documents Nº Document Name Code R.1 TARSIS. Operational Concept Document TEC/TAR/003 R.2 TARSIS. Dithering and mapping strategies TEC/TAR/009 R.3 TARSIS. Conceptual Design. Site. UV Characterization TEC/TAR/008 TARSIS. Science requirements Document TEC/TAR/007 1.Α - 17/05/2023 7 TABLE OF CONTENTS 1. SUMMARY ........................................................................................................................ 9 2. GALAXY CLUSTERS AS COSMOLOGICAL PROBES ............................................ 9 2.1 The mass profile of galaxy clusters ................................................................................ 10 2.1.1 Strong lensing ................................................................................................................ 10 2.1.2 Weak lensing .................................................................................................................. 11 2.1.3 Dynamic of galaxies ...................................................................................................... 12 2.2 The nature of dark matter .............................................................................................. 17 2.2.1 The inner mass profile ................................................................................................... 17 2.2.2 Anisotropy in the velocity distribution of DM haloes ................................................... 17 2.3 Substructure in galaxy clusters ...................................................................................... 18 2.4 Galaxy alignments ........................................................................................................... 19 2.5 Mass Accretion Rates (MAR) ........................................................................................ 22 2.6 Intracluster Light ............................................................................................................ 24 2.7 Galaxies' velocity dispersion function ........................................................................... 27 3. EVOLUTION OF GALAXIES IN CLUSTERS ........................................................... 28 3.1 Transformations of galaxies in clusters ........................................................................ 29 3.1.1 Measuring star formation histories in CATARSIS ........................................................ 30 3.1.2 The importance of 2D-spectroscopy .............................................................................. 33 3.2 Chemical evolution in clusters and the star formation history of galaxies ................ 34 3.2.1 Stellar and interestellar medium chemical abundances ................................................. 36 3.2.1.1 Evolution of the IMF .............................................................................................. 38 4. BEHIND THE CLUSTERS ............................................................................................ 39 4.1 The Lyman-alpha transition .......................................................................................... 40 4.1.1 The physics of the Lyman-alpha transition .................................................................... 40 4.1.2 Lyman-alpha emitters .................................................................................................... 41 4.1.3 The faint-end of the LAE luminosity function .............................................................. 43 4.1.4 Measuring the properties of LAE in CATARSIS .......................................................... 44 4.1.5 The emision of the Circumgalactic medium .................................................................. 45 TARSIS. Science requirements Document TEC/TAR/007 1.Α - 17/05/2023 8 4.1.6 Counterparts of Damped Lyman-alpha .......................................................................... 46 4.2 Dissecting HeII emitters: spectral templates for sources of the Cosmic Dawn ......... 46 4.3 Lensed galaxies ................................................................................................................ 47 5. SYNERGIES WITH OTHER FACILITIES ................................................................. 47 6. SCIENCE CASES OUTSIDE CATARSIS .................................................................... 48 6.1 Stellar evolution Science Cases ...................................................................................... 48 6.1.1 Spectral classification of stars in OB-Type clusters (J. Maíz Apellániz - CAB) ........... 48 6.1.2 A new view of the evolution of metal-poor massive stars with TARSIS (M. García - CAB) 49 6.1.3 Structure of Extended Planetary Nebulae (M. A. Guerrero - IAA-CSIC) ..................... 50 6.1.4 Extinction law in OB clusters (J. Maíz Apellániz - CAB) ............................................. 51 6.2 Solar System Science ....................................................................................................... 51 6.2.1 Structure and properties of Small Bodies in the Solar System (R. Duffard et al - IAA) 51 6.3 Targets of Oportunity ..................................................................................................... 51 6.3.1 CAHA response to GW alerts (A. Carramiñana - INAOE) ........................................... 51 7. SUMMARY OF THE SCIENTIFIC REQURIMENTS ............................................... 53 7.1 Justification of the most critical scientific requirements ............................................. 53 7.1.1 Blue coverage and FoV .................................................................................................. 54 7.1.2 Integral Field Spectrograph with a large field of view .................................................. 54 8. JUSTIFICATION OF THE REQUESTED OBSERVING TIME .............................. 55 8.1 Description of the sample and the observing plan ....................................................... 55 8.2 Justification of the requested observing time ............................................................... 57 8.2.1 Justification of dark night request .................................................................................. 59 TARSIS. Science requirements Document TEC/TAR/007 1.Α - 17/05/2023 9 1. SUMMARY The Calar Alto Tetra-Armed Super-IFu spectrograph Survey, CATARSIS, will survey 16 galaxy clusters in the redshift range 0.15 < z < 0.23 to understand the formation of structures and the evolution of galaxies in a dynamic and growing environment. CATARSIS will sample each cluster from the center to 2xR200, obtaining 2D spectra in the TARSIS full spectral range (320810 nm) of all the targets brighter than mr,AB = 22 mag and galaxies with even fainter magnitudes but with emission line fluxes above 1-2 x 10−17 erg s−1 cm−2. There are two major far-reaching interests that motivate our understanding the population of galaxy clusters. Firstly, they are ideal test objects to check the likelihood of a given cosmological model to describe our Universe; secondly, they are very important probes of the astrophysical and chemical evolution of the baryonic component of the Universe (e.g. Rosati et al. 2002; Schuecker et al. 2003a,b; Voit 2005; Vikhlinin et al. 2003, 2009; Rozo et al. 2007; Henry et al. 2009; Mantz et al. 2008). A good understanding and observational characterisation of galaxy clusters is required to attain these goals. Furthermore, being a blind survey, CATARSIS has an important component of serendipitous science related to galaxies at high redshift. Due to the long exposure times, the large field of view (FoV) of TARSIS and its wavelength coverage, CATARSIS will be very competitive in the detection of high-redshift targets and, in particular, in the detection of Lyman-α emission at redshifts 1.63 < z < 3, even when compared to other surveys using instruments in 8-m class telescopes. This document describes the science cases of CATARSIS that motivated the instrument requirements of the instrument TARSIS. We will relate them to the state of the art, emphasizing those where we expect a more important contribution of CATARSIS. The document also outlines other science cases that, not being part of the survey, show the capabilities of TARSIS to improve our knowledge of a variety of astrophysical phenomena. The unique characteristics of this instrument ensure the scientific return of the CAHA observatory for the years to come. 2. GALAXY CLUSTERS AS COSMOLOGICAL PROBES Galaxy clusters have been used for cosmology in a variety of ways, by e.g., i) measuring their gas fraction or mass-to-light ratio (e.g. Carlberg et al. 1996; Allen et al. 2008), ii) using their X-ray luminosity function (e.g., Allen et al. 2002; Pierpaoli et al. 2003; Mantz et al. 2008), iii) using information of their spatial distribution, through the Sunyaev-Zel’dovich (SZ) power spectrum/non-gaussianity and clustering analyses (e.g., Hong et al. 2012; Planck collaboration 2015; Bolliet et al. 2018; Marulli et al. 2018; note the increased relevance of such non-gaussianity studies to the light of the latest JWST results on the density of z ≳"8 galaxies; see Biagetti et al. 2023 and references therein), or iv) estimating their abundances as a function of mass and redshift (e.g, Vikhlinin et al. 2009; Planck collaboration 2015; Bocquet et al. 2019; Costanzi et al. 2019). The last two methods have gained prominence in recent years thanks to the increased size and/or depth of the cluster catalogues obtained from past or ongoing optical (SDSS, DES, HSC) or mmwavelength surveys (Planck, ACT, SPT). TARSIS. Science requirements Document TEC/TAR/007 1.Α - 17/05/2023 16 Figure 6: Number of cluster members inside spheres of radius = R200 (green), 1.36 ´ R200 (blue) and 3 ´ R200 (black) in Illustris TNG300. The number of members is plotted as a function of the virial radius of the cluster at z = 0.15, The cyan arrows represent the number of spectroscopic members in the first data release of HeCS cluster redshift survey, while the yellow ones, the number of members with spectra in the CATARSIS mother sample. Figure 7: Number of galaxies in clusters at z = 0.15 (top panels) and z = 0.20 (bottom panels) as a function of the Hα luminosity (left), [OII] luminosity (central) and r-band magnitude (right) in Illustris TNG300 cosmological simulations. Different colors represent the averages for clusters with different virial radius, as indicated in the insets. As mentioned before, using different methods to determine cluster mass profiles is fundamental since different methods suffer from different systematics. For instance, ICM determinations of cluster masses assumes the gas is in hydrostatic equilibrium and tend to be underestimated if bulk gas motions and the complex thermal structure of the Intra-Cluster Medium (ICM) are ignored (Rasia et al. 2004, 2006; Lau et al. 2009; Molnar et al. 2010; Cavaliere et al. 2011). Cluster triaxiality, substructures and interlopers tend to bias the mass profile estimates obtained by gravitational lensing (e.g. Meneghetti et al. 2011; Becker & Kravtsov 2011; Feroz & Hobson 2012) while triaxiality and substructures in the velocity space are the main biases for cluster galaxy kinematics (Cen 1997; Biviano et al. 2006) analysis. Comparing different mass profile determinations can therefore help assessing the contribution of nonthermal pressure to the ICM TARSIS. Science requirements Document TEC/TAR/007 1.Α - 17/05/2023 17 and the elongation along the line-of-sight (e.g., Morandi & Limousin 2012; Sereno et al. 2012). If systematics are well under control, the comparison of independent determinations of cluster mass profiles from gravitational lensing and the kinematics of cluster members can shed light on the very nature of DM (Faber & Visser 2006; Serra & Domínguez Romero 2011). 2.2 The nature of dark matter Uncovering the nature of dark matter (DM) is one of the major goals of science in the 21st century. The standard cosmological model, ΛCDM, assumes that DM particles are collisionless, which is to say that the only DM interactions relevant for structure formation are gravitational. Selfinteracting dark matter (SIDM) is an interesting alternative to CDM where DM particles can scatter with one another at astrophysically important rates. 2.2.1 The inner mass profile A non-vanishing DM self-interaction probability would impact the internal structure of halos, as collisional processes transport heat across high-density regions, thereby homogenizing the mass distribution and flattening the observed profiles in the cores of halos (Yoshida et al. 2000; Rocha et al. 2013; Robertson et al. 2019, 2021). Self-interacting dark matter (SIDM) was proposed as a possible to the core-cusp problem in dwarf spheroidal galaxies (see Tulin & Yu 2018 for a review). SIDM cross sections in the range σ/m ∼ 1 − 5 cm2 g-1 would produce central densities in broad agreement with the values observed in dwarf galaxies (e.g., Spergel & Steinhardt 2000; Davé et al. 2001; Valli & Yu 2018; Ren et al. 2019). SIDM also provides a potential solution to the observed diversity of galaxy rotation curves (Kamada et al. 2017; Creasey et al. 2017; Ren et al. 2019; Kahlhoefer et al. 2019; Sameie et al. 2020; see, however, Santos-Santos et al. 2020) and the anti-correlation between Milky Way satellites’ pericentric distances and their central densities (Kaplinghat et al. 2019; Correa 2020). Evidence for a large DM–DM scattering cross-section would rule out many popular DM candidates and would therefore alter the most promising regions of DM parameter space at which to target direct and indirect detection experiments (Zentner 2009; Kaplinghat et al. 2014; Boddy et al. 2014; Kouvaris et al. 2015; Del Nobile et al. 2015). These tests however require a very accurate determination of the mass density profile of a representative sample of clusters over a wide radial range, from kpc to Mpc scales. CATARSIS observations will allow to improve the determination of inner mass profiles using strong lensing techniques (Section 2.1.1), and it will complement these measurements with dynamical measurements (Section 2.1.3). The advantage of CATARSIS observations for the determination of masses with these two methodologies have been outlined above, and they should result in errors below a few percent. 2.2.2 Anisotropy in the velocity distribution of DM haloes The three-dimensional shape of collisionless halos is predicted to be generally triaxial with a preference for prolate shapes (Warren et al. 1992; Jing & Suto 2002), reflecting the collisionless nature of DM (Ostriker & Steinhardt 2003). Older halos tend to be more relaxed and thus to be rounder. Since more massive halos form later, on average, cluster-size halos are expected to be more elongated than less massive systems (Shaw et al. 2006; Ho et al. 2006; Despali et al. 2014, 2017; Bonamigo et al. 2015). Accretion of matter from the surrounding large-scale environment also plays a key role in determining the shape and orientation of halos. The halo orientation tends TARSIS. Science requirements Document TEC/TAR/007 1.Α - 17/05/2023 18 to be in the preferential infall direction of the subhalos and hence aligned along the surrounding filaments (Shaw et al. 2006). The shape and orientation of galaxy clusters thus provide an independent test of models of structure formation. The DM velocity anisotropy profile also holds important information about the nature of DM, such as its collisionless nature (e.g., Host et al. 2009) and decaying time (Peter, Moody, & Kamionkowski 2010). N-body simulations using CDM suggest a nearly universal velocity anisotropy profile (Cole & Lacey 1996; Carlberg et al. 1997; Colin, Klypin, & Kravtsov 2000; Diemand, Moore, & Stadel 2004; Rasia et al. 2004; Wojtak et al. 2005), where β increases radially from zero in the central region to roughly 0.5 in the outer region (Carlberg et al. 1997; Cole & Lacey 1996; Hansen & Moore 2006). For collisional systems, in contrast, the velocity anisotropy is explicitly zero in the equilibrated regions. The velocity anisotropy of a DM is defined by the β parameter as β" ≡ "1" −"𝜎! "/𝜎# " , where 𝜎! " and 𝜎# " are the 1-dimensional tangential and radial velocity dispersions in a spherical system (Binney & Tremaine 1987). Efforts to measure the DM velocity anisotropy have been done using the ICM temperature, a method which is applicable at intermediate radii and relaxed systems (Host et al. 2009), or by examining galaxy velocities (Lemze et al. 2011), with constrains on the self-interaction cross-section per unit mass of σ/m ≲ 1 cm2 g−1. 2.3 Substructure in galaxy clusters The presence of substructures can substantially affect the estimate of the cluster velocity dispersion and mass (Girardi et al. 1996; Pinkney et al. 1996) but, on the other hand, it can provide insights into the formation process of the cluster, and the expansion rate of the Universe (Richstone et al. 1992; Kauffmann & White 1993; Mohr et al. 1995; Thomas et al. 1998). From an astrophysical point of view, the mass growth of clusters brings new (generally gas-rich) galaxy populations into clusters (e.g., Moran et al. 2007) and may lead to shock heating of the ICM and/or disruption of cooling in cluster cores (e.g., Poole et al. 2008). The presence of substructures appears to be a fundamental ingredient of the galaxy-environment connection and for shaping the morphology-density relation (e.g., Fasano et al. 2015; Girardi et al. 2015). Reliable measurements and interpretation of cluster substructure are, therefore, of broad interest. The identification of substructure has been done using X-ray surface brightness morphology to either separate relaxed from merging clusters (e.g., Parekh et al. 2015b) or to quantify the systematic errors affecting cluster mass measurements (Nagai et al. 2007; Piffaretti & Valdarnini 2008). The substructure identification with X-rays is limited to the cluster central region within the R500 radius, which is the typical largest distance where the ICM can be detected reliably (Piffaretti & Valdarnini 2008), and to substructures that contain a quantity of hot gas large enough to produce a detectable X-ray emission. In addition, the identification of substructures is complicated by the fact that viscosity and magnetic fields can displace the hot gas from the dominant mass distribution, as indicated by the observations of numerous merging clusters (e.g., Markevitch et al. 2004; Mahdavi et al. 2007; Menanteau et al. 2012). The existence of substructures in the DM halos of clusters can also be revealed by the anomalous images of strong gravitational lensing systems (Kneib et al. 1996; Mao & Schneider 1998; Mao et al. 2004) or by peculiar features of the halo density profiles of weak lensing systems (Hoekstra et al. 2000; Clowe et al. 2006; Okabe et al. 2010; Pastor Mira et al. 2011; Oguri et al. 2013; TARSIS. Science requirements Document TEC/TAR/007 1.Α - 17/05/2023 19 McCleary et al. 2015; Shirasaki 2015), although the contamination by chance alignments of unrelated massive systems along the line of sight can be severe (Hoekstra 2003; Hoekstra et al. 2011; Geller et al. 2013). The redshift distribution of member galaxies exposes substructure in the form of asymmetrical velocity distributions and dynamically distinct subgroups (e.g., Dressler & Shectman 1988; Hou et al. 2012; Cohn 2012; Einasto et al. 2012). Not only does this type of analysis offer key insights into the dynamical state of cluster, but it also provides an estimate of cluster mass. However, the Caustic techniques offers a more promising avenue, not only to detect substructure, but to measure their masses. Yu et al. (2015) tested the Caustic technique as a substructure detector on two samples of 150 mock redshift surveys of clusters with M200 masses in the range between M200∼(1014-1015) h−1 M⊙ using a large cosmological N-body simulation of a ΛCDM model. Figure 8: Completeness vs. 3D substructure mass for three different N3R (randomly selected redshifts in the FoV, including clusters and no cluster galaxies) in clusters with a median mass M200 = 1014 M ⨀ using the Caustic methodology. The error bars show the deviation for different random selections of the redshifts (from Yu et al. 2015). The success of the method to identify substructures depend on the mass of the cluster, the mass of the structures, and the number of available redshifts for galaxy members. Figure 8 illustrates this by showing the completeness in the identification of structures in a cluster of 1014 M⨀ as a function of the structure mass for different number of redshifts (N3R) randomly selected in the FoV of a cluster at z = 0.2. Deep observations using TARSIS will allow the detection and characterization of small structures in the CATARSIS cluster sample. The main requirement is having redshifts for a large and unbiased number of galaxies with a homogeneous distribution in the whole area. The expected errors in the line-of-sight velocities are very small compared to the velocity dispersion of the cluster and the cluster structures, and they will not add important sources of errors. 2.4 Galaxy alignments The cosmic web is shaped by the gravitational tidal field, which determines the directions of anisotropic mass collapse. The same tidal field is also responsible for spinning up haloes and galaxies. For example, during the linear phase of structure formation, the tidal torque theory (TTT; Hoyle 1949; Peebles 1969; Doroshkevich 1970; White 1984) describes how the angular TARSIS. Science requirements Document TEC/TAR/007 1.Α - 17/05/2023 20 momentum of a proto-halo is generated by the gravitational shear of the surrounding matter distribution. This produce that galaxy shapes are not randomly oriented, rather they are statistically aligned in a way that can depend on formation environment, history, and galaxy type. Studying the alignment of galaxies can, therefore, deliver important information about the physics of galaxy formation and evolution as well as the growth of structure in the universe. In particular, the current paradigm states that galaxies acquire their angular momentum by the torque exerted from large-scale structure onto the protogalactic object prior to gravitational collapse (Hoyle 1949). The angular momentum, which is conserved during collapse, will align with the filaments, but only for galaxies below a certain mass. Figure 9: Expected alignments in the TTT between the spin of disc & early-type galaxies and the filaments. Indeed, high-mass haloes have spins preferentially perpendicular to filaments while low-mass haloes show the opposite trend, with their spins being preferentially parallel to filaments (e.g., Navarro et al. 2004; Aragón-Calvo et al. 2014). The halo mass at which this transition happens is known as the spin-flip transition mass, or, in short, spin-flip mass. This transition mass increases with decreasing redshift (e.g., see Codis et al. 2012; Wang & Kang 2018) and is ∼1 × 1012 h−1 M⊙ at present day. The results from the observations are not so clear, despite the large number of studies in the literature (e.g., Trujillo et al. 2006; Wang et al. 2018, to cite a few). Some suggest that spiral galaxies are aligned perpendicular to filaments, in agreement with predictions from cosmological simulations based on the ΛCDM model (Lee & Pen 2002; Lee & Erdogdu 2007; Jones et al. 2010; Zhang et al. 2015), while others find just the opposite (e.g., Tempel et al. 2013; Tempel & Libeskind 2013; Pahwa et al. 2016). The study of alignments can also give us some insights into the nature of DM. Despite the success of the cold and collisionless DM paradigm (CDM), we know very little about its particle nature, other than its interactions with the Standard Model (protons, neutrons, neutrinos, etc.) must be exceptionally weak. Interestingly, the self-interaction of dark matter is not constrained by the same limits and provides a unique avenue to understand forces within the dark sector without having to assume any coupling to the Standard Model (for a review see Tulin & Yu 2018). TARSIS. Science requirements Document TEC/TAR/007 1.Α - 17/05/2023 21 Recently, it has been shown that the intrinsic alignments of galaxies are sensitive to the nature of DM up to scales of several Mpc/h. Using hydrodynamical simulations with different models of self-interaction, Katz et al. (2018) have showed that the differences should be detectable at a 5σ level and it could be recognized due to its dependency with halo mass and cluster-centric distance. Furthermore, as we mentioned before, the study of “intrinsic alignments” (IA) of galaxies with overdensities is crucial as it represents a major contaminant of weak lensing cosmic shear studies, which rely on the assumption that galaxies are randomly oriented (Hirata & Seljak 2004; Bridle & King 2007; Kirk et al. 2010, 2012; Krause et al. 2016). Frameworks to mitigate the impact of IA on cosmological model inference have been developed, whereby nuisance parameters can be marginalised over (King 2005; Joachimi et al. 2011). This requires insights into the origin of the intrinsic alignments signal (see Figure 10). Figure 10: Sketch of the gravitational lensing signal and its IA contamination. Light travels from the top of the sketch downwards, from the source plane, via the lens plane, to the plane at the bottom, containing the images as seen by an observer. The matter structure (green ellipsoid) deflects the light from the background source galaxies (blue discs) and distorts their images tangentially with respect to the apparent center of the lens (as seen in the bottom plane). Consequently, the galaxy images become aligned (GG signal). Galaxies which are physically close to the lens structure (red ellipsoids) may be subjected to forces that cause them to point towards the structure, which results in the alignment of their images. Images of galaxies close to the lens are then preferentially anti-aligned with the gravitationally sheared images of background galaxies (from Joachimi et al. 2016). Measuring alignments have been usually analysed using galaxy images by assuming that galaxies spin along their minor axis. However, due to projection effects, it is not possible to know which half of the two in which the galaxy is divided by its major axis, is closer to the observer, so it is not straightforward to measure the 3D orientation of the galaxy's angular momentum. To deal with this problem some authors take all the possible directions and average the results while others just take one possibility, plaguing the results with uncertainties. Thanks to the 2D spectroscopy provided by TARSIS, we will be able to obtain a coarse velocity field, that will be enough, if galaxies have trailing spiral arms, to determine, unambiguously, the real alignment of galaxies. The study of alignments between galaxies and filaments requires to know the rotational direction of galaxies with respect to the line of sight. TARSIS minimum requirements, with a pixel size of 2 arcsec and spectral resolution of R~1000 is enough to obtain the 2D velocity field. To illustrate this, we resampled MUSE observations of NGC6902 to these values and convolved it with a PSF TARSIS. Science requirements Document TEC/TAR/007 1.Α - 17/05/2023 22 of 1-arcsec FWHM. We then run Ppxf (Cappellari & Emsellem 2004) to measure line-of-sight velocities. Figure 11 shows the results, demonstrating that it is possible to measure the galaxies' global rotation pattern using CATARSIS observations. Furthermore, it is also possible to build maps of emission line fluxes and line-strength indices. Even if the resolution is not very high, these measurements will be invaluable to understanding the physical processes that influence the evolution of galaxies in clusters (see Section 3.1.1). Figure 11: Line-of-sight velocity map for nearby galaxy NGC 6902 (see direct image in the top-left corner) shifted to the CATARSIS limiting redshift of z = 0.23 (right), and as it was observed with the TARSIS instrument, for three different spaxel scales. The thin-disk modelling results (a µ Vmax, v=Vsys –Vz=0.23, pa=position angle) for the TARSIS spaxels (~2 ´ 2 arcsec2) is shown in the upper right corner. 2.5 Mass Accretion Rates (MAR) In the current model of the formation of cosmic structure, where DM halos form from the aggregation of smaller halos, the MAR of dark matter halos is a stochastic process whose average behavior can be predicted with N-body simulations (e.g., van den Bosch 2002). In principle, the MAR of dark matter halos is a valuable tool for testing different models of structure formation. Specifically, the mass evolution M(z) of a DM halo, which describes its mass assembling history, or its time derivative, the mass accretion rate Ṁ(z) can be used to test the predictions of the ΛCDM cosmological model, i.e., MAR should be correlated with halo properties, including concentration (Tasitsiomi et al. 2004; Zhao et al. 2009; Giocoli et al. 2012; Ludlow et al. 2013), shape (Kasun & Evrard 2005; Allgood et al. 2006; Bett et al. 2007; Ragone-Figueroa et al. 2010); spin (Vitvitska et al. 2002; Bett et al. 2007), degree of internal relaxation (Power et al. 2012), and fraction of substructures (Gao et al. 2004; van den Bosch et al. 2005; Ludlow et al. 2013). The model also predicts a correlation with the halo formation redshift (Lacey & Cole 1993; van den Bosch 2002; Ragone-Figueroa et al. 2010; Giocoli et al. 2012). The mass accretion history (MAH) can be a probe of the cosmological parameters. Hurier (2019) uses the thermal Sunyaev-Zel’dovich (SZ) effect as a proxy for the mass of the clusters from the 2nd Planck SZ catalogue (Planck Collaboration et al. 2016) and the fit by Correa et al. (2015) to the MAH of dark matter halos in simulations to derive values for the power-spectrum TARSIS. Science requirements Document TEC/TAR/007 1.Α - 17/05/2023 23 normalization (σ8), the cosmic mass density (Ωm), and the Hubble parameter (H0), which result in σ8(Ωm/0.3) − 0.3 (H0/70 km s-1 Mpc-1) − 0.2 = 0.75 ± 0.06. This value is in rough agreement with other analyses of galaxy cluster samples and of the power spectrum of the cosmic microwave background (de Haan et al. 2016; Planck Collaboration VI 2020; Zubeldia & Challinor 2019). A unique probe of the mass assembly of galaxy clusters relies on the observational detection of their edges. Examining the logarithmic slope of the cluster halo profile highlights a density jump at the location where recently accreted material is reaching its first apocenter, associated with the last density caustic (Bertschinger 1985) density profile of both matter and galaxies, termed the splashback radius. This radius, that correspond to the location of the first orbital apocenter of satellite galaxies after their infall, offers a robust way to gain insight into dynamics properties of the halo and to the nature of DM itself. In ΛCDM, the location of the splashback radius or the turnaround radius crucially depend upon the MAR of the collapsing halo (Vogelsberger et al. 2011; Diemer & Kravtsov 2014; Adhikari et al. 2014). For halos of the same mass, large accretion rate results in a smaller splashback radius (see Shim & Diemer 2023 and references therein). The physical reason is simple: the deeper the halo potential well gets during the orbit of a DM particle, the smaller is the value of its apocenter. A change of slope that is consistent with the expectations from the simulations is indeed present in the profile of the surface number density of galaxies from the Dark Energy Survey (DES) cross-correlated with the SZ clusters from the South Pole Telescope (SPT) and the Atacama Cosmology Telescope (ACT) (Shin et al. 2019), as well as in the deprojected cross-correlation of the SZ clusters from the Planck Survey with galaxies detected photometrically in the PanSTARRS survey (Zürcher & More 2019). Similarly, the splashback radius is detected in the inferred DM density profiles of the redMaPPer clusters (More et al. 2016) and in clusters from either SDSS (Baxter et al. 2017) or DES (Chang et al. 2018). Although interlopers along the line of sight affect the inference of this feature both from optically-selected clusters (Busch & White 2017; Shin et al. 2019; Sunayama & More 2019) and from weak lensing analyses of X-ray selected clusters (Umetsu & Diemer 2017; Contigiani et al. 2019), galaxy redshift surveys covering the whole area of interest, like CATARSIS, can overcome the effects of this contamination and constrain the relation between the splashback radius and the accretion rate (Xhakaj et al. 2020). Furthermore, while the splashback radius is observed as a feature in the spatial distribution of matter and galaxies (and their velocities), it also has an inherent timescale associated with it: the time for a particle or galaxy to reach the apocenter of its first orbit. In other words, DM particles or galaxies that form the splashback region have been inside the halo for approximately one orbital time, from accretion to first apocenter. Unlike DM particles, galaxy properties like star formation rates (SFRs) and morphology can evolve significantly on similar timescales if the cluster environment plays a role in galaxy evolution. Since the splashback feature is closely related to the orbital time it is possible to use it as a clock to study different populations of galaxies and their time-evolving properties. The MAR can also be derived in more model independent way. Beyond the splashback radius, the velocity of the DM within a radial shell reaches a minimum, and so most closely represent the infall of new material onto the cluster (De Boni et al. 2016). Therefore, assuming an infall velocity (e.g., one based on a spherical-collapse model) is, in principle, possible, to derive the MAR. Diaferio & Geller (1997) applied this method to numerical simulations obtaining very good agreement. To derive MARs, however, it is necessary to obtain a good coverage of redshifts in TARSIS. Science requirements Document TEC/TAR/007 1.Α - 17/05/2023 24 the external parts, which is very rare in studies using multi-object spectrographs (see also Rines et al. 2013). Figure 12: Relation between the splashback radius and the accretion rate predicted by three different cosmological simulations based on the ΛCDM model. ! 2.6 Intracluster Light The most revealing signature of galaxy cluster assembly is contained within a diffuse component occupying the space between the galaxies in clusters. This component is composed of a substantial fraction of stars (5−20% of the total light of the cluster, Krick & Bernstein 2007) that constitute the so-called intra-cluster light (ICL, see Mihos 2016 for a review). This diffuse light is thought to form primarily by galaxies that interact with the cluster medium or merge during the hierarchical accretion history of the cluster (e.g., Gregg & West 1998; Mihos et al. 2005; Conroy et al. 2007; Presotto et al. 2014; Contini et al. 2014). The amount of ICL, stellar population properties (age, metallicity) and the spatial variation of the ICL are the results of the assembly history of the cluster (Rudick et al. 2010), and that is why its analysis is crucial. Different physical mechanisms may be at play in the formation of the ICL, and their relative importance can vary during the dynamical history of the cluster and at different mass ranges. Therefore, the fraction of light in this component and the evolution with redshift will provide information of the efficiency of the interactions that form the ICL. The determination of the mass fraction of this component can also help to alleviate the tension between the observed growth rate of brightest cluster galaxies (around 2 in the last 7-8 Gyr) and that expected in theoretical models based on a ΛCDM cosmology (3–4 since z =1, de Lucia & Blaizot 2007, blue shaded area in Figure 4). The tension between simulations and observations can be alleviated if we assume that a significant percentage of the accreted stellar mass ends up in the cluster’s ICL rather than in the Brightest Cluster Galaxy (BCG) (between 30−80%, Conroy et al. 2007; Laporte et al. 2013; TARSIS. Science requirements Document TEC/TAR/007 1.Α - 17/05/2023 25 Contini et al. 2018). This brings the predictions of the models into better agreement with the observations (purple area in Figure 13). Figure 13: Mass growth rate of the BCG with redshift. Green dots are observations from: Whiley et al. (2008); Collins et al. (2009); Stott et al. (2010); Lidman et al. (2012); Lin et al. (2013); Oliva-Altamirano et al. (2014); Burke et al. (2015); Bellstedt et al. (2016); Zhang et al. (2016). The blue and purple shaded areas are simulations from de Lucia & Blaizot (2007) and Contini et al. (2018), respectively (figure from Montes 2019). The relative importance of the mechanisms responsible for the ICL will likely vary during the evolution history of the cluster and, possibly, within different regions of the cluster. A useful tool to determine the properties of the ICL is the study of its stellar populations. In fact, the ages and metallicities of the ICL population reflect the properties of the progenitor galaxies from which the stars got stripped. For example, Contini et al. (2014) predicted that the bulk of the ICL light is stripped from the most massive (M∗ ∼ 1010 −1011 M⊙) galaxies as they fall into the cluster core (see also Rudick et al. 2011; Cooper et al. 2013). If this is the case, the ICL should exhibit a mean metallicity like in the outer regions of these massive satellites. Additionally, the age of the ICL stellar populations should give us an upper limit on when the formation of the ICL took place. This is because we do not expect any star formation in the ICL component after their stars have been stripped from their progenitor objects. In this sense, knowing the age and metallicity of the ICL of the clusters allow us to infer how (and when) the assembly history of these clusters was, ranging from the shredding of dwarf galaxies (Purcell et al. 2007; Contini et al. 2014), to violent mergers with the central galaxies of the cluster (Murante et al. 2007; Conroy et al. 2007), or in situ formation (Puchwein et al. 2010). Studies of ICL colors show clear radial gradients (Iodice et al. 2017; Mihos et al. 2017; DeMaio et al. 2015, 2018) indicating radial variations in metallicity and, in some cases, age. These gradients point to tidal stripping of massive satellites as the dominant process of ICL formation. In fact, the metallicities found in the ICL region indicate that the progenitors of the unbound stars are the outskirts of galaxies of masses around 3 × 1010 M⨀. Spectroscopic studies of stellar populations in ICL (Coccato et al. 2010; Melnick et al. 2012; Edwards et al. 2016) offer contradictory results. Melnick et al. (2012) found that the ICL is dominated by old, metal rich stars (75%) while Coccato et al. (2011), using Lick indices, found that 100% of the ICL stars in the Hydra-I cluster were old and metal poor. Most spectroscopic studies have used long slit TARSIS. Science requirements Document TEC/TAR/007 1.Α - 17/05/2023 32 Figure 17: Top: UV-optical SED of two stellar populations, one Simple Stellar Population (SSP) of 10 Gyr (in black) and one SSP with a 0.01% (in mass) young burst (63 Myr-old, in red). Bottom: Ratio between the two models. Note that only in the UV range the two spectra show significant differences. CATARSIS will allow measuring the whole sequence of galaxy transformations, relating them to the epoch of accretion into the cluster and to thermodynamical conditions of the ICM gas, and the substructure of the cluster. We will measure detailed SFH that will allow to study very low levels of star formation and, therefore, trace the quenching processes over longer periods of time than Hα SFR measurements, relating it with the properties of the cluster. To derive detailed SFH we need to obtain spectra in the rest-frame region 280-320 nm, as this region is extremely sensitive to the presence of young stars. Studies of stellar populations using the optical part of the spectra require signal-to-noise (S/N) per angstrom above 30 to minimize the effects of the age-metallicity degeneracy. However, the information contained in the NUV part of the spectra allows relaxing this condition. This is illustrated in Figure 18 (Costantin et al. 2019), that shows the probability distribution function of the ages derived with optical and optical+NUV line-strength indices in two spectra with different SFH and S/N. It is clear that when NUV indices are used, unbiased ages can be measured in spectra with S/N = 10, which we will be able to obtain in the spectrum of galaxies with small fractions of young stars, due to the dramatic increase in the flux that we expect at these wavelengths (see Figure 17). TARSIS. Science requirements Document TEC/TAR/007 1.Α - 17/05/2023 33 Figure 18: Examples of Probability Distribution Functions (PDFs) of r-magnitude weighted mean age for a simulated galaxy with a SFH that includes a small fraction of young population. PDFs retrieved using only optical indices are in red and those obtained with both ultraviolet and optical indices in blue. The blue shaded region represents the confidence interval corresponding to the 16-84 per cent percentile range of the PDF. The solid green line corresponds to the true value, <age>r-mag = 4.44 Gyr, and the dashed vertical red (blue) line indicates the median value of the corresponding distribution. 3.1.2 The importance of 2D-spectroscopy There is mounting evidence that ram-pressure stripping (RPS) is one of the most effective mechanisms quenching galaxies in clusters. Simulations show that RPS can effectively remove cold gas from galaxies, and in some cases temporarily enhance the star-formation activity before quenching it completely (see, e.g., Steinhauser, Schindler & Springel 2016, and references therein). Evidence supporting RPS comes from observations of neutral gas (HI) (Haynes, Giovanelli & Chincarini 1984; Cayatte et al. 1990; Chung et al. 2009; Vollmer et al. 2001; Jaffé et al. 2015) showing cold gas in the process of being stripped from the galaxy (see Figure 19). In some cases, stars are formed in the stripped gas (Kenney & Koopmann 1999; Kenney et al. 2014), and can thus be identified from UV or optical images. The most striking examples of stripped galaxies with new stars tracing the stripped tails are the so-called “Jellyfish” galaxies. Estimates based on small blind Hα surveys of Coma and A1367 indicate that the fraction of cluster late-type galaxies with these features is close to 40% (Boselli & Gavazzi 2014). TARSIS. Science requirements Document TEC/TAR/007 1.Α - 17/05/2023 34 RPS is inversely proportional to radius of the cluster, as both the ICM density and the free-fall velocity increase as a galaxy approaches the cluster centre. The second term, however, is the binding energy of the galaxy, that decrease at increasing radius. This is evident in Figure 19, showing that the gas is being removed from outside-in. Figure 19: False-color SDSS images of galaxies falling into the Virgo cluster. The HI emission (coming from the VIVA survey; Chung et al. 2009) is indicated with the white contours, showing the neutral gas of the galaxy is being stripped from outside-in (from Wong et al. 2014). If we limit our analysis to the central apertures, we will be missing the important effects of RPS during a large period of the crossing time, when the density is not high enough to remove the gas for the galaxy center. The spatial resolution of CATARSIS will allow us to resolve regions between 5 kpc and 7.5 kpc at z = 0.15 and 0.25 respectively, enough to study the effects of this process on galaxy properties from much earlier. Furthermore, a 2D analysis will give important insight into the stellar and gas kinematics in these highly disturbed galaxies (e.g. Fumagalli et al. 2014; Bellhouse et al. 2017; Kalita & Ebeling 2019), detect the presence of star formation induced by the hot gas pressure, and the fraction of new stars that form outside the galaxy. 3.2 Chemical evolution in clusters and the star formation history of galaxies Theoretical models of chemical evolution, coupled with state-of-the-art galaxy evolution models will also play an increasingly important role. The complexity and non-linearity of the processes driving the chemistry at high redshift require better and more accurate numerical simulations. TARSIS. Science requirements Document TEC/TAR/007 1.Α - 17/05/2023 35 From the effect of stellar feedback, the initial mass function (IMF), and detailed nucleosynthesis to the birth and growth of supermassive black holes (SMBHs), observations need to guide these theoretical predictions. Clusters of galaxies are unique laboratories for the study of the nucleosynthesis and chemical enrichment of the Universe as their deep gravitational potential wells keep all the metals produced by the stellar populations of the member galaxies within the cluster, making them the closer the closed-box model approximation. The dominant fraction of these metals reside within the hot ICM. The chemical abundances measured in the intra-cluster plasma thus provide us with a “fossil” record of the integral yield of all the different stars (releasing metals in supernova explosions and winds) that have left their specific abundance patterns in the gas prior and during cluster evolution. The detection of an overabundance of α-elements with ASCA suggests that the metals in the ICM might come mainly from Type II SNe (Loewenstein & Mushotzky 1996; Mushotzky et al. 1996). Thus, one of the possible mechanisms for the metal enrichment of the ICM is a wind from proto-galaxies driven by Type II SNe explosions in their early evolutionary phase (aka early galactic wind). In a hierarchical universe, it has been shown that the epoch of galaxy formation is earlier than that of cluster formation, that is it, in the protocluster. Some simple theoretical attempts of chemical and population synthesis modelling failed to simultaneously match the constraints on ICM metallicity and on the stellar mass-to-light ratios (Renzini & Andreon 2014; Loewenstein 2013; Ghizzardi et al. 2021). These authors claimed that the iron yield in the largest clusters (M500,crit > 1014 M⊙) must be a few times the solar value. Two main solutions have been suggested, one is a time-dependent IMF and the other is related to variations in the Type-Ia SNe yields. However, either solution should explain not only the mass distribution of elliptical galaxies, but also their detailed chemical abundances. If the ICM enrichment is due to strong winds in protocluster galaxies with very high SFR, we will expect that bright elliptical galaxies in the center of clusters have a high Mg/Fe. Indeed, elliptical galaxies in the center of clusters show a uniform age and a [Mg/Fe] that increases with galaxy mass (e.g., Kuntschner et al. 1998, Sánchez-Blázquez et al. 2006). This relation seems to be the same independent of the cluster mass or X-ray luminosity (Carretero et al. 2004). Contrary to this, CN in elliptical galaxies have also shown that galaxies with the same mass (and [Mg/Fe]) show a much more prominent CN feature in X-ray luminous clusters, indicating that the SFHs were not the same for galaxies in all clusters (Figure 20). As C and N are mainly release by intermediatemass stars, this has been interpreted as the consequence of slightly more extended SFHs in less massive clusters. This fine-tuned SFH would have been long enough for the Type II SNe to explode, so their products could be incorporated in the next generation of stars but, in the most massive clusters, it was stopped before intermediate mass stars had time to reach the AGB. If this happened at the proto-cluster stage, then the mechanism that caused it must be related with its present-day cluster mass. For example, a time-dependent IMF whereby a strongly star-forming system starts with a topheavy IMF, followed by a later phase where a large amount of gas is locked into low-mass stars (Vazdekis et al. 1996, 1997; Davé 2008) could explain the observed abundances. However, this scenario produces high values of [Mg/Fe], chemical enrichment properties, X-ray binary fractions, as well as the IMF signatures in the spectra of early-type galaxies (ETGs, Weidner et TARSIS. Science requirements Document TEC/TAR/007 1.Α - 17/05/2023 36 al. 2013; Ferreras et al. 2015), that should be examined in detail. If we invoke variations in the SNIa yields to explain the Fe abundance in the ICM, they should also be compatible with the chemical abundance ratios of ETGs in clusters. Figure 20: Cluster X-ray luminosity vs. overabundance values of [ZCN/ZFe] (top), [ZMg/ZFe] (middle) and [ZMg/ZCN] (bottom). Circles, squares, and triangles indicate clusters with richness classes of 1, 2 and 3, respectively. Each point corresponds to one individual cluster and represents the interpolated abundance ratio for σ = 200 km s−1. The Spearman rank-order correlation coefficients and its significance values are written in top and bottom panels (from Carretero et al. 2004). X-ray observations of galaxy clusters show that the ICM contains large amounts of heavy elements, such as O, Ne, Mg, Si, S, Ar, Ca, and Fe (e.g., Werner et al. 2008). The constant ratios of abundances of several of these elements found throughout the Virgo cluster (Simionescu et al. 2015) as well as in the radial profiles of 44 clusters observed out to intermediate radii with XMMNewton (Mernier et al. 2017) indicate that, during the early period of metal enrichment, the products of core-collapse and Type Ia SNe were well mixed. The estimated ratio between the number of Type Ia SNe and the total number of SNe enriching the ICM is about 15–20%, percentage that will need to be consistent with the stellar chemical abundances in galaxies, that we will measure with CATARSIS. A consistent chemical evolution model that incorporates but stellar and hot-gas components should put strong constrains on the early epoch of protocluster formation and it will allow to probe or discard the need for a time changing IMF. 3.2.1 Stellar and interestellar medium chemical abundances In the local universe, the chemistry of galaxies is generally investigated by targeting two distinct types of galaxies. The simplicity and high-surface brightness of massive, quiescent galaxies makes them ideal candidates for studying stellar absorption features, especially in their central regions. Consequently, they are traditional benchmarks for stellar population modelling and the study of the stellar abundances in external galaxies. The chemistry of the ISM, on the other hand, is studied in star-forming galaxies, where the gas is excited in HII regions by the presence of massive stars. This dichotomy has led also to contradictory results and to significant gaps in our TARSIS. Science requirements Document TEC/TAR/007 1.Α - 17/05/2023 37 understanding of the chemical evolution of galaxies. The wavelength range covered by CATARSIS will allow the measurement of two chemical elements, Nitrogen and Magnesium in both components: the stars and the nebular gas. Stellar Nitrogen abundances can be obtained from the NH3375 index defined by Serven (2010), while Mg can be obtained from Mg2800, Mg2852, Mg3334 and also from Mgb (see Figure 21). Abundances of the Nitrogen ions in the warm gas can be obtained from [NII]6584Å while the presence of MgII2800 emission will give us access to the abundance of this ion. As the NH3375 line will be dominated by the youngest component, the comparison with the abundance of NII can be used to calibrate the ionization corrections for integrated spectra, when individual HII regions cannot be resolved, for different metallicities and SFRs. These corrections are a very important source of uncertainty in measurements of nebular abundances, but they are difficult to constrain observationally, especially in external galaxies. The comparison of nebular and stellar Mg or the nebular O will be interesting to study the destruction of dust by the hot ICM. Figure 21: Sketch of the most prominent absorption features in the NUV: Mg II 2800 Å, the Fe I3000 Å which trace four neutral magnesium lines (2966.90; 2984.73; 2999.81 and 3021.37 Å), the two iron FeII features at 2402 and 2609 Å and MgI at 2852 Å. We also be able to test common interpretations of nebular line ratios. According to theoretical models, the gas-phase N/O abundance ratio in galaxies is strongly influenced by the SFH of galaxies (e.g., Henry et al. 2000; Mollá et al. 2006; Vincenzo & Kobayashi 2018). Therefore, the N/O–O/H diagram (for individual star forming complexes and also for global average values for integrated galaxy spectra) has been suggested as an alternative SFH chemical abundance diagnostic of galaxies (e.g., Vincenzo et al. 2016; Chiappini, Matteucci & Ballero 2005; Mollá et al. 2006). In the nebular component, The N/O abundance ratio will be derived from the ratio of the [NII]l6584Å and [OII]ll3726,3729Å emission lines, via the N2O2 parameter (PérezMontero & Contini 2009), while the gas-phase O/H chemical abundance can be derived from a number of strong emission-lines present in the spectral range (namely [OII]ll3726,3729Å, Hβ, [OIII]l5007Å, Hα, [NII]l6584Å). The comparison of N/O with the SFH derived with a full spectral fitting (see previous section) will tell us if the first is, indeed, an alternative to measure the second. TARSIS. Science requirements Document TEC/TAR/007 1.Α - 17/05/2023 38 The CATARSIS survey will offer an excellent data sample to test chemical evolution models' predictions using both ionized gas and stellar abundances. The TARSIS wavelength range contains a plethora of absorption and emission lines that, combined with the derived SFHs, will give us new information to calibrate and interpret commonly used features to measure chemical abundances. 3.2.1.1 Evolution of the IMF As we mentioned above, a possible solution to reconcile the chemical abundances in hot halos in clusters with the mass-to-light ratios of galaxies is to use a top-heavy IMF. This is also a condition for semi-analytical models of galaxy formation and chemodynamical simulations to reproduce the relation between the stellar [Mg/Fe] and velocity dispersion observed in elliptical galaxies (Yates et al. 2013; Arrigoni et al. 2010). However, in the last few years, improved technology and population synthesis modelling have allowed to use spectral line-strength indices that are sensitive to the giant vs. dwarf stellar ratio as a discriminant of the IMF in passive populations. Early results (Cenarro et al. 2003; van Dokkum & Conry 2010), later confirmed with independent data and analysis (e.g., Ferreras et al. 2013; La Barbera et al. 2013) indicate the opposite, a bottom-heavy IMF in elliptical galaxies. These results were supported by dynamical constraints, produced at around the same time but based on kinematic studies of IFU data of nearby galaxies (Cappellari et al. 2012, 2013), that revealed high values of the stellar mass-to-light ratios with respect to the expectations from a standard, Milky Way-type IMF). Gravitational lensing over galaxy scales provides a third, independent probe of the bottom-heavy IMF in massive early-type galaxies (Treu et al. 2010; Smith & Lucey 2013). A way to solve this problem is by invoking a time-dependent IMF whereby a strongly starforming system starts with a top-heavy IMF, followed by a later phase where a large amount of gas is locked into low-mass stars (Vazdekis et al. 1996, 1997; Davé 2008). This scenario produces chemical enrichment properties, X-ray binary fractions, as well as the IMF signatures found in massive ETGs (Weidner et al. 2013; Ferreras et al. 2015), although it is admittedly contrived. The relation between SFR and IMF was explored by Hoversten & Glazebrook (2008) in a sample of star forming galaxies using a combination of Hα EW and g-r color (see Figure 22). They found that low SFR galaxies lie below models with a Salpeter IMF (α = –2.35), towards the bottom track for α = –3.0. In contrast, those with high SFR lie towards the top-heavy model with α = –2.0. TARSIS. Science requirements Document TEC/TAR/007 1.Α - 17/05/2023 39 Figure 22: Galaxies from the GAMA survey with three different SFR. The evolutionary tracks predicted by PEGASE (solid) and maraston (dashed) for three IMF (-3, -2.35 and -2, top to bottom) are indicated. On the other hand, a correlation between the IMF and the metal content has been found in elliptical galaxies, both globally and as a function of radius. To disentangle between a strong star formation and metallicity is difficult as one goes usually associated to the other at a given time. Thanks to the rest-frame NUV observations of CATARSIS, we can contribute to shed some light to this problem in an independent way. The FeI300 spectral index, with values that are nearly twice for dwarf stars than for super-giants, is a strong feature, especially in star-forming galaxies. Besides, it has the advantage that is not affected by dust reddening and it is not contaminated by older populations. Furthermore, the cluster environment may give us the possibility of finding galaxies with similar SFR and different metallicities, or vice-versa, for example by comparing galaxies in merging clusters with others that are just infalling into the cluster. 4. BEHIND THE CLUSTERS Observations of deep extragalactic fields performed with MUSE (HDFS, Bacon et al. 2015; UDF, Bacon et al. 2017; CDFS, Urrutia et al. 2019; MUDF, Lusso et al. 2019) have demonstrated how a wide-field optical IFU is a game-changer for the study of distant galaxies. It provides a comprehensive spectroscopic view of the sky, i.e., high-quality spectra for all sources in the FoV with no prior selection. This approach has produced an order of magnitude increase in the number of spectroscopic redshifts measured in these deep fields (Herenz et al. 2017; Inami et al. 2017; Brinchmann et al. 2017), thereby revealing systematically groups and associations of galaxies that would never have been targeted for spectroscopic follow-up otherwise (e.g., Ventou et al. 2017). Our deep exposures will serendipitously observe a large variety of different objects at high redshifts, and thanks to its blue coverage, the current MUSE redshift desert (z = [1.5 − 3], corresponding to the redshift of [O II]ll3726,3729Å emitters at the red end and of Lyman Alpha Emitters –LAEs– at the blue end) will be largely filled down to z = 1.63 by CATARSIS. Among other things, this will allow us to better characterize the nature Damped Lyman Alpha systems, [OII], HeII or Lyα emitters. We will be able to observe Lyα and MgII emission at the same time in objects between 1.63 < z < 1.9, which will be a giant step to understand the resonance nature of both lines and to quantify the Lyα escape fraction. Furthermore, our detection limit will allow the study the emission of the intergalactic medium and, therefore, to understand better the physics TARSIS. Science requirements Document TEC/TAR/007 1.Α - 17/05/2023 40 of gas inflows and outflows in galaxies. CATARSIS will provide new ways of exploring the physics associated to high-redshift objects allowing the same study to be done at much lower redshifts. Although these projects are normally carried out in large aperture telescopes, the large field of view, the much smaller cosmic dimming, and the efficiency of TARSIS make this project highly competitive. Using the Ultra-Fast Image Generator (UFug) for wide astronomy surveys (Berge et al. 2013) and a limiting magnitude of r = 22 mag, we calculate that the number of galaxies observed serendipitously will be 8000 to 9000 per square degree. This adds up to a total of 12000-15300 spectra at the end of the survey. The detection of galaxies with strong emission lines will more than triple this number. 4.1 The Lyman-alpha transition With a vacuum rest-frame wavelength of 1215.67 Å, the Lyman-α (Lyα) recombination line (n = 2 → n = 1) is intrinsically the strongest emission line in the rest-frame UV and optical (e.g., Partridge & Peebles 1967; Pritchet 1994). Lyα emission from stellar populations and AGN is a powerful diagnostic of high-redshift objects, enabling the effective identification and spectroscopic confirmation of these sources (Bromm & Yoshida 2011; Finkelstein 2016). The blue wavelength coverage of CATARSIS will allow the detection of LAEs at redshifts in the range 1.63 < z < 3, which coincides with the rise and fall of the SFR density in the Universe (i.e. the "Cosmic Noon"). The spectral resolution of CATARSIS will allow to study the Lyα profile containing important information about the radiative transfer on the clouds and about the escape of Lyα photons, which is of vital importance to understand the epoch of reionization. CATARSIS observations will also detect the Lyα emission of the circumgalactic medium (CGM; Steidel et al. 2011; Hayes et al. 2013; Momose et al. 2016; Leclercq et al. 2017; Erb et al. 2018) which will allow studying the gas accretion processes into the galaxies. 4.1.1 The physics of the Lyman-alpha transition Due to the complex physics of the Lyα radiative transfer process, it has proven difficult to interpret the results from theoretical models and observational data (Dijkstra 2014) despite the fact that understanding the physics of Ly 𝛼 escape has major implications. Ly 𝛼 equivalent width (WLyα) and the Lyα escape fraction (fescLyα) are higher (WLyα ≳ 20Å and fescLyα ≳ 10%) in galaxies that represent the less massive and younger end of the distribution of galaxies in the local universe. This relates to various properties, such as the fact that Lyαemitting galaxies have lower metal abundances (median value of 12+log(O/H) ∼ 8.1) and dust reddening. However, the presence of galactic outflows/winds is also vital to Doppler shift the Lyα line out of resonance with the atomic gas, as high WLyα is found only among galaxies with winds faster than ∼50 km s−1. In fact, self-consistently incorporating non-ionizing continuum and other hydrogen, nebular, and metal absorption lines into Ly 𝛼 studies helps to disentangle certain radiative transfer effects that are either amplified or suppressed by resonant scattering. For example, different types of line and continuum radiation may have unique spatial, spectral, and angular escape or absorption features that help to determine the relative importance of turbulent ISM porosity compared to smoother spherical and disc-like density gradients. This is also significant because ionizing radiation and stellar feedback act to clear low-column density TARSIS. Science requirements Document TEC/TAR/007 1.Α - 17/05/2023 41 channels that facilitate the escape of Ly 𝛼 photons as suggested by several theoretical works (e.g., Yajima et al. 2014; Dijkstra et al. 2016; Kimm et al. 2019; Kakiichi & Gronke 2021; Mauerhofer et al. 2021) and observational studies (e.g., Nakajima & Ouchi 2014; Henry et al. 2015; Chisholm et al. 2018; Gazagnes et al. 2020; Jaskot et al. 2019). The spectral resolution of CATARSIS will be enough to resolve the double peak of the Lyα emission and therefore, study the escape fraction in relation to both the dynamics and the metallicity of the cloud (see Section 4.1.4). This is easily seen in Figure 23¡Error! No se encuentra el origen de la referencia., that shows the Lyα profile of a LAE at z = 1.88 observed with the LRIS (Oke et al. 1995) spectrograph with a grism that gives an average resolution of FWHM ~ 4.0 Å (275 km s−1), very similar to what we will achieve in CATARSIS (from Berg et al. 2018). Figure 23: Velocity profile of the double-peaked Lyα emission in SL2S0217, a LAE at z = 1.8. The overall profile is reasonably well fit using two Gaussians (blue lines). Interestingly, the blue component is stronger than the red one and has a slightly larger velocity width. The Lyα peak separation is relatively small, but can be resolved with observations made with a spectral resolution FWHM = 4 Å (from Berg et al. 2008). 4.1.2 Lyman-alpha emitters Already more than 50 years ago, the Lyα emission line of hydrogen was predicted to be a superb tracer for galaxy formation and evolution studies in the high redshift universe (Partridge & Peebles 1967). Meanwhile the study of LAEs has provided a route to identify low-mass galaxies at high redshifts that possibly constitute the progenitors of present-day L* galaxies such as the Milky Way (Gawiser et al. 2007). Most LAE samples have so far been constructed from narrowband imaging (e.g. Hu & McMahon 1996; Rhoads et al. 2000; Ouchi et al. 2003; Shibuya et al. 2012; Sobral et al. 2017; Ouchi et al. 2018), but significant efforts need to be spent on confirming LAE candidates by spectroscopy. LAE samples have also been built from large multiobject spectroscopic surveys (Stark et al. 2010; Cassata et al. 2015), but to be efficient, such samples rely (by construction) on a very stringent photometric preselection of high-redshift candidates. The all-in-one approach of using TARSIS as a survey instrument obviates the need of any pre-selection and follow-up spectroscopy (see below). Previous results include a measurement of the clustering properties of LAEs (Diener et al. 2017), an estimate of the Lyα TARSIS. Science requirements Document TEC/TAR/007 1.Α - 17/05/2023 48 on the CMB photon energy distribution induced by inverse-Compton scattering off the ICM hot free electrons. The relatively high angular resolution of the TolTEC SZE observations (~6 arcsec at 1.1mm) will potentially detect pressure discontinuities in the ICM (i.e., shocks) produced by the merging of substructures, and probe the gas dynamics and astrophysical processes out to the low-density regions in the outskirts of galaxy clusters (Mroczkowski et al. 2019). These mm-wavelength observations will therefore complement those from CATARSIS to provide a multi-wavelength picture of the mass assembly of clusters and the growth of structure. Furthermore, mmwavelength observations are sensitive to the dust and the obscured star-formation in galaxies. Optical-UV observations are known to underestimate the total SFR of dusty galaxies, even when mid-IR data (e.g., 24 μm MIPS/Spitzer) are available. The combination of CATARSIS and TolTEC/LMT follow-up observations will therefore allow us to understand the process of star formation and its impact on the evolution of galaxies within large scale structures more accurately. Observations with the suite of heterodyne instruments of the LMT towards samples of galaxies identified by CATARSIS will trace their cold molecular gas to better understand the mass buildup in galaxies over time (e.g., Cybulski et al. 2016). Since clusters of galaxies act as strong gravitational lenses, targeted observations towards these massive structures increase the probability of serendipitously discovering faint distant galaxies, which are elusive to current mmwavelength facilities (e.g., Pope et al. 2017). This population of galaxies, with modest SFR (~10 M⨀ yr-1), is expected to dominate the SFH of the Universe at z > 2. Having access to OpticalUV and mm-wavelengths data is therefore crucial to have a complete view of the SFH and better understand the impact of dust at high redshifts. 6. SCIENCE CASES OUTSIDE CATARSIS Although the call from the CAHA Observatory was to carry out a survey and TARSIS was selected specifically for carrying out the CATARSIS survey, we have collected some science cases outside CATARSIS to illustrate the versatility of TARSIS and its usefulness for a variety of different science cases, not only in the field of “galaxies and cosmology” but also for studies of stellar evolution and classification, the ISM, or the Solar System. 6.1 Stellar evolution Science Cases 6.1.1 Spectral classification of stars in OB-Type clusters (J. Maíz Apellániz - CAB) Gaia is allowing us to identify the membership in stellar clusters but the wavelength region it uses for its spectrograph is the Ca triplet window, where few useful lines exist for OB stars. The blue spectrographs in TARSIS could be used to spectroscopically characterize those stars, as the blueviolet region is rich in OB stellar lines. As this case does not require spectrophotometric calibrations, it can be done under poor conditions and even be used as a filler program. Each cluster could be covered in a time scale of 1 hour for compact systems to several hours for more extended ones. As there are several tens of OB clusters to be analyzed, the case could be applied to a fraction of the sample or to the full one as a filler program. In this case the advantage of TARSIS is the access to the whole blue-violet spectral region using a large IFS on a 3.5 m telescope. TARSIS. Science requirements Document TEC/TAR/007 1.Α - 17/05/2023 49 6.1.2 A new view of the evolution of metal-poor massive stars with TARSIS (M. García - CAB) The role of massive stars as dynamizing agents of the Universe begins already at the reionization epoch, being among the first sources capable of providing a strong field of UV photons. Since there on, characterizing the evolution of galaxies and the Cosmos relies on our knowledge of the physics of massive stars, which must be provided as a function of the varying environment and extended to the conditions of the early Universe. One of the most important parameters changing in the evolution of the Universe is metallicity, ever increasing since the formation of the first stars. Our current standard for low metallicity is the Small Magellanic Cloud (SMC), with about 1/5 of the solar value. That is far from the very low metallicities of the early Universe and even from the metallicity of the peak of star formation, that does not exceed Z⨀/10 (Madau & Dickinson 2014). Efforts are being made to surpass the SMC frontier by characterizing massive stars in Local Group and nearby galaxies with metal content akin to earlier stages of the Universe (García et al. 2019), but progress is slow due to their distance (≥750 kpc). Photometric candidates must first be confirmed with low-resolution spectroscopy, which requires long observing programs on multi-object spectrographs and has the risk of interesting stars slipping through the photometric cuts. The interest in metal-poor massive stars has been revitalized by the results from the revolutionary LIGO and VIRGO experiments. The masses of the black holes (BH) whose collapse has been registered by gravitational wave (GW) events evince the possibility that massive stars keep most of their mass until the end of their evolution, leading to pre-BH masses 20 or 30 M⊙ larger than previously predicted by stellar evolution models (Abbott et al. 2016). Since massive stars lose mass to radiation-driven winds (RDWs), propelled by absorption and re-scatter of photons by metallic ions, this new scenario immediately invokes metal-poor chemical composition with the ensuing weaker winds. Re-visiting the evolution of massive stars within binary systems has also led to interesting new channels. In particular, there is a new view according to which long-known post-main sequence stages such as Be stars, Red Supergiants (RSG), Luminous Blue Variables (LBV), and WolfRayet stars (WR), could be the result of the interaction of two stars within a binary system (Beasor et al. 2020; Götberg et al. 2018). This emerging scenario is in line with the detection of isolated WR/LBV stars in metal-poor galaxies (Herrero et al. 2010; Tramper et al. 2015), hard to explain in the context of RDWs. However, there is much work ahead to characterize the mass donors and gainers within the system, and the relative incidence of these stellar classes due to binary versus classical-single star evolution (Shenar et al. 2020). The interest of these findings exceeds the academic interest on massive star evolution and impact the estimated production of ionizing photons and kinetic energy by a stellar population, the estimated rates of SNe/GRBs within a galaxy, the inferred IMF from integrated populations, and the interpretation of nebular lines in distant galaxies. Yet, the incidence of binary stars, and their typical period and mass fractions, is uncharted territory beyond the Magellanic Clouds. The Local Group and nearby galaxies could enable us to explore this topic but only a couple of binary stars have been confirmed in these systems. Again, TARSIS. Science requirements Document TEC/TAR/007 1.Α - 17/05/2023 50 the main reason is that the distance to the galaxies impose strict sensitivity requirements that make multi-epoch observations prohibitive. A large field of view IFU such as TARSIS can help us advance significantly in our studies of metal-poor massive stars. On the one hand it will enable untargeted studies of galaxies, enlarging the scarcely populated catalogs of metal-poor massive stars while fully sampling the upper IMF. On the other hand, it will enable multi-epoch observations and the detection of spectroscopic binaries in Local Group galaxies. We therefore envision observations of massive stars in star-forming regions in Local Group galaxies accessible from Calar Alto (and with a density adequate to the TARSIS spaxel size), covering a range in metallicity from the Large Magellanic Cloud (as anchor point) to well beyond the Small Magellanic Cloud. Our previous work on these galaxies will allow us to identify the best suited regions to obtain significant statistical information on metal-poor massive stars and on their massive binary population. 6.1.3 Structure of Extended Planetary Nebulae (M. A. Guerrero - IAA-CSIC) Planetary nebulae (PNe) represent the final stages of the stellar evolution of lowand intermediate-mass stars with initial mass from 1 to 8 M⊙ (Kwok 2000). These short-lived nebulae encode information on the chemical enrichment of the progenitor stars and their yields to the interstellar medium to contribute to the chemical evolution of galaxies. Their characteristic emission line spectra make them easily detectable in external galaxies (or even in the ICM of clusters of galaxies) where they can be used to trace the kinematics and chemical abundances and to be used as secondary distance indicators (Ciardullo et al. 1989). The spectral coverage and resolution of TARSIS are ideally suited for spectral analysis of PNe to investigate their physical conditions and key abundances ratios such as He/H, O/H, N/O and Ne/O among others. Compared to “traditional” long-slit and small-sized integral field spectroscopy, the large field of view of TARSIS will be a game changer in the investigation of the largest structures of PNe, low surface brightness haloes surrounding the bright main PNe (Chu et al. 1987). These haloes correspond to material ejected during the final stages of the asymptotic giant branch, most likely associated with thermal pulses (Stanghellini & Pasqualli 1995), so they can be used to diagnose the chemical evolution on the surface of lowand intermediate-mass stars as the result of the second and third dredge-up (Monreal-Ibero et al. 2005). The large FoV of TARSIS can also be used to investigate PNe in resolved external galaxies. These investigations are traditionally performed on a two-step way, the first one in imaging mode to detect the population of PNe (e.g., Merrett et al. 2006) and the second one in a follow-up spectroscopic mode to characterize them (Kwitter et al. 2012; Fang et al. 2015). TARSIS observations can go through the whole process in a single step, identifying the population of PNe and obtaining spectral information of them simultaneously. The results derived from the population of PNe can be added to those obtained from other galactic components (stars, dust, ionized nebulae) in the same observation. TARSIS. Science requirements Document TEC/TAR/007 1.Α - 17/05/2023 51 6.1.4 Extinction law in OB clusters (J. Maíz Apellániz - CAB) Most of the previous work on the topic has been done with photometry, not with spectrophotometry, so the detailed behavior with wavelength is poorly constrained. TARSIS allows for the simultaneous observation of some cluster stars using the blue spectrographs and others using the red spectrograph. Using either a rotating ("Staring mode") or mapping ("Mapping mode") strategy, it should be possible to cover an OB cluster (including calibrations) with ~10 useful stars in ~1 hour. Therefore, using 1-2 nights we could produce a significant sample of 100200 stars with 3200-8100Å spectrophotometry and derive the extinction law (and its variations) in that region. This case would require photometric conditions to be carried out. The biggest advantage of TARSIS with respect to other spectrographs for this case is the easy access to the U band. 6.2 Solar System Science 6.2.1 Structure and properties of Small Bodies in the Solar System (R. Duffard et al - IAA) CATARSIS is a survey optimized for extra-galactic astronomy and cosmology, aimed at obtaining deep spectroscopy of galaxy clusters and filaments at z~0.15-0.23 through selected pointings. Its proposed observational strategy favors higher-quality data in the bluer range (320520nm) rather than in the redder part (510-810nm). Serendipitous observations of small bodies (SB) of the Solar System can be considered as transient phenomena due to their apparent motion in the sky that will produce random observations in some pointings. The most interesting data that TARSIS will provide us in the context of the Solar-System science is in the blue range, as this has not been thoroughly explored yet because most of the attention has been directed to the red and near-infrared spectra where absorption features of silicates and ices appear. Random observations of Small Bodies within CATARSIS may lead to the observation of extended objects, perhaps surrounded by a coma due to cometary-like activity. Some cometary emission lines of interest appear in the 320-520nm region, in particular lines of C2, NH2, and CN (e.g., Brown et al. 1996), although most will be blended due to the moderate resolving power of TARSIS. On the other hand, the IFU capabilities can be used to study the morphological properties of these extended objects in regions close to the nucleus. The redder part of the spectrum will also be interesting to characterize the continuum of the spectra. 6.3 Targets of Oportunity 6.3.1 CAHA response to GW alerts (A. Carramiñana - INAOE) August 17th, 2017 represented an unforgivable event in the history of astrophysics, thanks to the detection, for the first time, of the electromagnetic (EM) counterpart of a Gravitational Wave detection event, GW170817 (Abbott et al. 2017; Pian et al. 2017; Smartt et al. 2017), that was caused by the merger of two neutron stars (NS-NS event). As a highlight of its relevance is worth noting that during the period between August and September 2017 a total of 5,000 images and spectra were taken with the 14 instruments of the seven ESO telescopes (Levan & Jonker 2020). That moment represented the crack of the starter's pistol for the race for being first and being best in detecting these EM counterparts, even although it had already started back in 2015 with the detection of the first BH-BH GW event with LIGO+VIRGO (GW150914; Abbott et al. 2016). TARSIS. Science requirements Document TEC/TAR/007 1.Α - 17/05/2023 52 The onset of GW170817 set a level for the LIGO+VIRGO NS-NS detection rate of 1.5 yr–1 within 100 Mpc3 (with 90% credible range of 0.3-4.7 yr–1; Abbott et al. 2017). The Advanced LIGO/Virgo detector was sensitive to NS-NS mergers up to 70-100 Mpc. Indeed, the GW170817 (NGC 4993) is an early-type galaxy with sub-solar metallicity at an approximate distance of 4045 Mpc, depending on the distance estimator used (see Hjorth et al. 2017). The actual GW170817 NS-NS counterpart was found ~2 kpc (0.64´Reff) away from the galaxy center (see Figure 28). Since the most efficient EM-counterpart identification strategy as of today is to target the mostprobable hosts (based on the position and redshift probability distributions of the GW event and the host's stellar mass content and SFH; see Belczynski et al. 2018), having the possibility of directly catching the EM-counterpart in spectroscopy with a wide-FoV IFU such as TARSIS would allow to be first and to be best in this race. In that regard, TARSIS would put the CAHA 3.5m as a unique facility for following up this kind of NS-NS events without the need of carrying out pre-imaging or even of comparing these imaging observations with previous imaging data. Note, that with the increase in sensitivity of future GW detectors previous imaging data might not reach the required depth to perform such identification, especially at high declinations where the Rubin Observatory will not be observing. Figure 28 below shows the image of NGC 4993 with the TARSIS FoV over-imposed to emphasize the unique capabilities of TARSIS for performing direct spectroscopic observations of GW EM counterparts even if they are found in hosts located significantly closer to us than NGC 4993. Figure 28: TARSIS FoV over-imposed on the discovery image of the EM counterpart of the GW170817 NS-NS Gravitational Wave (GW) event in nearby elliptical galaxy NGC 4993. TARSIS. Science requirements Document TEC/TAR/007 1.Α - 17/05/2023 53 7. SUMMARY OF THE SCIENTIFIC REQURIMENTS Table 1 lists the design requirements of TARSIS in terms of FoV, spectral range per spectrograph, spaxel size, resolving power, and efficiency. The scientific requirements for each of the major science cases of CATARSIS described elsewhere in this document are listed in Table 2. Blue Red Field of View (FoV) 3 × (1.36 × 1.36) arcmin2 1.36 × 1.36 arcmin2 Spectral range 320 nm – 520 nm 510 nm – 810 nm Spatial resolution element 2.05 × 2.05 arcsec2 2.05 × 2.05 arcsec2 Resolving Power (R) 740 at 320 nm, 1030 at 420 nm, 1290 at 520 nm 750 at 510 nm, 980 at 660 nm, 1210 at 810 nm Total average efficiency without the telescope (but including slicer + spectrograph + detector) ³ 20% ³ 30% Table 1: Main characteristics of the two sets of spectrographs of the TARSIS instrument. Table 2: Summary of the broad scientific requirements imposed by the CATARSIS survey and level of criticality of each for the science cases described above. 7.1 Justification of the most critical scientific requirements The two requirements that have been a major challenge when designing the instrument and that have a significant impact on the final budget are (1) the short blue limit of the wavelength range (λ = 3200Å) and its optimization of the instrument in the blue, and (2) its large FoV (2.8 arcmin TARSIS. Science requirements Document TEC/TAR/007 1.Α - 17/05/2023 54 ´ 2.8 arcmin). At the same time, those are the features that make this instrument unique and the ones that make CATARSIS scientifically strong. Besides, we also require (see Table 1 above) to reach a moderate spectral resolution or R~1000 for most of the TARSIS wavelength range and the highest efficiency compatible with these previous top-level requirements. 7.1.1 Blue coverage and FoV One of the characteristics that makes TARSIS unique is the wavelength coverage. Its wavelength range allows to study the rest/frame NUV in the spectra of our targeted galaxy clusters. We have shown that thanks to this wavelength range we can detect very small amounts (< 1%) of star formation and study the timescales of star formation quenching with a precision that will allow, for the first time, to distinguish between the different physical processes responsible for it. The optical region does not have sensibility to detect small amounts of star formation and the SFR indicators, like the Hα luminosity, become undetectable for SFRs below a certain threshold. The extension to the blue part of the spectra will allow us to study differences in the chemistry of young and old stars, compare chemical abundances in the stellar and ionized gas components and explore the variations of the IMF with metallicity and SFR (see Section 3.2). These projects will not be feasible without having access to rest-frame wavelengths < 3000Å. % It is also thanks to the wavelength coverage that we will be able to observe the Lyα transition at redshifts z ≃ 1.6, and characterize their properties measuring chemical abundances, ionization parameters and kinematics. Without the NUV we would not be able to observe, simultaneously, the Lyα and MgII λλ2780 transitions, that gives us the possibility to study the escape fraction using other resonant lines. Being able to observe Lyα at such low redshifts gives us a clear advantage over other surveys, as it is since z ≲1.9 that the cosmic SFR density starts to decrease. This wavelength range, together with the large FoV is what makes this project competitive even when compared with other projects carried out at 8-m class telescopes. 7.1.2 Integral Field Spectrograph with a large field of view% Almost all the science cases described in Section 2 rely on obtaining a sufficiently dense coverage of redshifts covering the whole extension of the clusters. The large number of redshifts for an unbiased sample of galaxies will allow to characterize substructure, identify interlopers or measure anisotropy, which in turn will allow us to reduce to the minimum existing biases in the determination of cluster mass profiles. Mapping the outskirts of clusters will also give us access to the calculation of accretion rates and the splashback radius, as well as to study the alignments of galaxies with respect to the filaments. These regions represent more than 90% of the cluster volume and the corresponding projected area in the sky and mapping them with a smaller FoV will extend the duration of the survey to unfeasible timescales. For substructure determination and for the study of galaxy transformations is especially important to observe unbiased samples of galaxies, without having to pre-select galaxies with certain properties, motivating the building of an IFS instead of a multi-object spectrograph covering a larger FoV. TARSIS. Science requirements Document TEC/TAR/007 1.Α - 17/05/2023 55 Furthermore, astronomy is to a large degree still driven by unexpected discoveries and the probability of serendipitous discovery is proportional to the field of view. It is for this that TARSIS, with its 2.8 arcmin x 2.8 arcmin field of view, has an enormous potential to discover. As we mentioned above, the efficiency of the survey is largely determined by the large FoV, and this requirement put some limits to spatial resolution of the instrument, 2.05 ´ 2.05 arcsec2. However, a higher resolving power would not be optimal for medium observing conditions, as a significant fraction of a point-source light would not fit a single spaxel when convolved with the Point Spread Function (PSF), lowering the detection limit of faint emission lines (see Table 3). Although the signal can be increasing by adding spaxels, this would result in a penalty in the final signal-to-noise ratio. This is also the case for extended sources with low surface brightness (i.e., ICL emission and outer parts of galaxies, see Sections 3.1.2 & 3.1.2), where the desired signalto-noise ratio can only be reached after binning the datacubes. Seeing (") (r-band) Seeing (") (320nm) %PSF inside 1x1" SPX %PSF inside 2”x 2” SPX 0.6 0.8 73% 99% 0.8 1.0 57% 96% 1.1 1.5 31% 77% 1.5 2.0 19% 57% Table 3: Percentage of light in a PSF for different seeing fitting inside spaxels of 1 ´ 1" and 2 ´ 2". 8. JUSTIFICATION OF THE REQUESTED OBSERVING TIME Below we describe the observing plan for each cluster and the justification for the exposure time. The total observing time and distribution throughout the 6 years of the survey is described as part of the Operational Concept Document (R.1). 8.1 Description of the sample and the observing plan The CATARSIS survey does not aims to observe a complete sample of clusters with a particular selection function, because its scientific goals are not statistical per se. There are many ongoing and forthcoming surveys that are collecting a very large sample of clusters for that purpose. CATARSIS results, though, will be of vital importance to understand the biases and systematics associated with the different methodologies that will be used in those surveys to infer cosmological parameters. These biases and systematics effects are related with both, local and global properties of the different components, recent history, and projection effects, among others. We have selected a total of 33 high-priority clusters (CTRS01 through CTRS33) that cover a range of these characteristics and that are, therefore, representative of the physical conditions that we expect to find in massive clusters. The clusters are selected in a redshift range from z = 0.15 to z = 0.25 because it optimizes the number of pointing’s to cover a large area of the clusters, restframe bluest wavelength edge (280-320 nm) for studies of stellar populations and chemical evolution and the signal-to-noise ratio achievable with TARSIS on the 3.5m telescope at CAHA. TARSIS. Science requirements Document TEC/TAR/007 1.Α - 17/05/2023 56 To better illustrate the feasibility of the observing plan and describe the strategies described in R.1 (and also in R.2), we have defined a subsample of 16 galaxies that constitute the Golden Sample and for which we are also intensively pursuing complementary observations at other wavelengths. However, the final sample can slightly change as new ancillary data supporting our science goals become available or if we encounter problems during early phases of operation that requires include different targets with specific characteristics. The sample selected is distributed in the sky so to ensure a good coverage from CAHA throughout the entire year with airmasses that do not go above airmass=1.22 (zenith distance < 35º), where the A320nm atmospheric extinction would be less than ~0.2 mag higher than that at the zenith (~0.91 mag). Document R.3 describes our study of the UV attenuation at the CAHA site developed as part of this feasibility study (WP9). The planification of the individual pointings for the Golden Sample are provided as links in the TARSIS webpage (https://guaix.ucm.es/tarsis). These figures include information on the cluster ID, redshift and R200 (arcmin) at the top left corner of the plot (see also Table 4 below), the positions of all square-shaped pointings proposed for observation as part of CATARSIS, together with the positions of cluster galaxies with available spectra and those targets with SDSS magnitude r < 22 mag that we expect to observe with CATARSIS. The position of QSO+AGN in the Million Quasars Catalog (MILLIQUAS; Flesch 2015). ID RA (deg) (J2000) DEC (deg) (J2000) 2 ´ R200 (º) z (1) (3) (4) (5) (6) CTRS01 006.136 +33.207 0.114 0.226 CTRS02 010.961 +20.667 0.1421 0.195 CTRS03 014.001 +26.342 0.2279 0.197 CTRS06 120.237 +36.056 0.1365 0.288 CTRS08 126.938 +34.790 0.1905 0.160 CTRS09 135.175 +20.895 0.0955 0.229 CTRS14 154.265 +39.047 0.2187 0.206 CTRS18 179.342 +33.632 0.1694 0.215 CTRS19 194.813 +27.965 0.1448 0.158 CTRS24 217.827 +35.110 0.1435 0.236 CTRS25 225.085 +21.362 0.2409 0.152 CTRS27 250.082 +46.711 0.1163 0.228 CTRS29 260.037 +27.670 0.2079 0.161 CTRS31 260.618 +32.154 0.0700 0.229 CTRS32 328.400 +17.695 0.2106 0.230 CTRS33 336.175 +18.671 0.1563 0.155 16 clusters Mean values: 0.1734 0.201 Table 4: Mother sample of CATARSIS sample. R200 indicate the radius where the density of the cluster is 200 times the critical density of the Universe. More information, such as the number of paintings to cover each cluster to the first apocenter of galaxy orbits along with the coverage maps, is provided at https://guaix.ucm.es/tarsis. TARSIS. Science requirements Document TEC/TAR/007 1.Α - 17/05/2023 57 In some of the clusters we also include in the online plots TARSIS pointing’s located at the positions of filaments identified by filament-finder algorithms. These observations will be very useful to study the impact of pressure shocks encounter by the galaxies when follow into the cluster and the so-called pre-processing, when they are still no bona-fide cluster members. These observations will also be used to study the alignments between filaments and galaxies. 8.2 Justification of the requested observing time The scientific objectives of CATARSIS require reaching continuum magnitudes mAB,r ~ 22 mag (depending on the cluster galaxy color and redshift) and limiting line fluxes of 1-2 x 10-17 erg s-1 cm-2. Figure 29 shows the distribution of the mean surface brightness in the r-band for a sample of galaxies at redshifts 0.15 < z < 0.25 from SDSS with magnitudes 21.9 < mAB,r < 22.1. Figure 29: Distribution of surface brightness for a sample of galaxies at redshifts 0.15 < z < 0.25 with magnitudes 21.9 < mAB,r < 22.1 mag. We have used the ETC-42 generic exposure time calculator developed at CeSAM (Centre de donneeS Astrophysiques de Marseille) by Gross et al. (2015) together with the total expected throughout of TARSIS and an atmospheric transmission curve of the CAHA observatory provided by Santos Pedraz (priv. comm., from Hayes & Latham 1975). This curve only covers wavelengths above 350nm so we used the transmission curve of La Palma Observatory (King 1985) to cover the range 320-350nm. The atmospheric extinction provided by the King's curve nicely coincide with the values we are obtaining in our program to characterise the UV sky at CAHA (see R.3). These values go from ~0.9 mag/airmass at 320nm down to 0.32 mag/airmass at 394nm. We use the sky emission spectrum obtained at the Paranal Observatory using UVES (Hanuschik 2003) since the one provided by Sánchez et al. (2007) only reached 370nm. Note that we are also running a program at the CAHA 1.23m telescope to monitor the effects of light pollution from LED urban lights at the UV end of the TARSIS wavelength coverage. Figure 30 shows the expected S/N for an extended target at z = 0.2 with a Sa spiral-galaxy template. We show the values at the central wavelength of the blue and the red spectrographs, as a function of the surface brightness of the target in the r-band. Table 5 and