Public Data Access Overview / Data Specifications

This page contains a comprehensive description of the simulation snapshots, group catalogs, merger trees, and supplementary data sets. Differences and additions in TNG with respect to the original Illustris public data release are noted: aspects which have changed, or which are new in TNG, are marked in blue, while those marked in green are identical to Illustris.

Table of Contents

• 1. Snapshots
• Organization
• Contents
• PartType0 (Gas)
• PartType1 (DM)
• PartType3 (Tracers)
• PartType4 (Stars)
• PartType5 (BHs)
• Subboxes
• 2. Group Catalogs
• FoF Halos
• Subfind Subhalos (Galaxies)
• 3. Important Additional Files
• Offsets
• The "simulation.hdf5" File
• 4. Merger Trees
• SubLink
• LHaloTree
• 5. Supplementary Data Catalogs
• (a) Tracer Tracks
• (b) Star Formation Rates
• (c) Stellar Circularities, Angular Momenta, Axis Ratios
• (d) Subhalo Matching Between Runs
• (e) Stellar Projected Sizes
• (f) Blackhole Mergers and Details
• (g) Stellar Assembly
• (h) Subbox Subhalo List
• (i) Molecular and atomic hydrogen (HI+H2) galaxy contents
• (j) Barred Galaxies Properties and Evolution
• (k) SDSS ugriz and UVJ Photometry/Colors with Dust
• (l) SKIRT Synthetic Images and Optical Morphologies
• (m) Disk Galaxy Kinematic Decompositions
• (n) L-Galaxies Semi-Analytical Model
• (o) Lensing Profiles
• (p) Aperture Masses (Total and Stellar)
• (q) Halo Structure
• (r) Galaxy Morphologies (Deep Learning)
• (s) Cosmic Web Distances (Disperse)
• (t) Galaxy Morphologies (Kinematic) and Bar Properties
• (u) LEGA-C/VIMOS Mock Optical Spectra
• (v) JWST-CEERS Mock Galaxy Imaging
• (w) Multi-band (UV to submm) Photometry and SEDs
• (x) Globular Clusters
• (y) Merger History
• (z) MaNGIA Mock MANGA IFU Cubes
• (1) TNG50 Milky Way+Andromeda (MW/M31)-like Galaxies
• (2) HVC-like Cosmological Cloud Catalog
• (3) Zooniverse Cosmological Jellyfish
• (4) iMaNGA Mock MANGA IFU Datacubes
• (5) Synthetic Stellar Light Images with HSC-SSP Realism
• (6) Nearest Neighbors
• (7) TNG50-Skirt Atlas

1. Snapshots

Organization

There are 100 snapshots stored for every run. These include all particles/cells in the whole volume. The complete snapshot listings, spacings and redshifts can be found in the API. Note that, unlike in Illustris, TNG contains two different types of snapshots: 'full' and 'mini'. While both encompass the entire volume, 'mini' snapshots only have a subset of particle fields available (detailed below). In TNG twenty snapshots are full, while the remaining 80 are mini. The 20 full snapshots are:

Every snapshot is stored in a series of "chunks", i.e. more manageable, smaller-size files. The number of chunks per snapshots is different for the different runs:

Note that the snapshot data is not organized according to spatial position. Rather, particles within the snapshot files are sorted according to their group/subgroup memberships, according to the FoF or Subfind algorithms. Within each particle type, the sort order is: GroupNumber, SubgroupNumber, BindingEnergy, where particles belonging to the group but not to any of its subgroups ("fuzz") are included after the last subgroup. The following figure provides a schematic view of the particle organization within a snapshot, for one particle type. Note that the truncation of a snapshot in chunks is arbitrary, thus halos may happen to be stored across multiple, subsequent chunks. Similarly, the different particle types of a halo can be stored in different sets of chunks.

Caption. Schematic diagram of the organization of particle/cell data within the snapshots for a single particle type. Within a type, particle order is determined by a global sort of the following fields in this order: FoF group number, Subfind subhalo number, binding energy, nearest FoF group number. This implies that FOF halos are contiguous, although they can span file chunks. Subfind subhalos are only contiguous within a single group, being separated between groups by an "inner fuzz" of all FOF particles not bound to any subhalo. Here $N_c$ indicates the number of file chunks, $n_F$ the number of FOF groups, and $N_{S,j}$ the number of subhalos in $j^{\rm th}$ FoF group.

Contents

Every HDF5 snapshot contains several groups: "Header", "Parameters", "Configuration", and five "PartTypeX" groups, for the following particle types (the DM only runs have a single PartType1 group):

• PartType0 - GAS
• PartType1 - DM
• PartType2 - (unused)
• PartType3 - TRACERS
• PartType4 - STARS & WIND PARTICLES
• PartType5 - BLACK HOLES

The most important fields of the Header group are given in the following table.

The Parameters and Configuration provide the complete set of parameter file options and run-time configuration options used to run TNG. That is, they encode the fiducial "TNG Galaxy Formation Model". Many will clearly map to Table 1 of Pillepich et al. (2018a), while others deal with more numerical/technical options. In the future, together with the release of the TNG initial conditions and the TNG code base of AREPO, this will enable any of the TNG simulations to be reproduced.

The complete snapshot field listings of the PartTypeX groups, including dimensions, units and descriptions, are given for all the particles types in the following large table. Rows in blue are new or different in some way with respect to original Illustris, while those in green are unchanged.

The general unit system is ${\rm kpc}/h$ for lengths, $10^{10} {\rm M}_\odot/h$ for masses, ${\rm km/s}$ for velocities. The frequently occurring $(10^{10} {\rm M}_\odot/h) / (0.978 {\rm Gyr}/h)$ represents mass-over-time in this unit system, and multiplying by 10.22 converts to ${\rm M}_\odot/{\rm yr}$. Comoving quantities can be converted in the corresponding physical ones by multiplying for the appropriate power of the scale factor $a$. For instance, to convert a length in physical units it is sufficient to multiply it by $a$, volumes need a factor $a^3$, densities $a^{-3}$ and so on. Note that at redshift $z=0$ the scale factor is $a=1$, so that the numerical values of comoving quantities are the same as their physical counterparts.

Tagged Metals

(*) The units of all the entries of GFM_MetalsTagged, except for NSNS, are the same as GFM_Metals: dimensionless mass ratios. If you sum up all the elements of GFM_Metals heavier than Helium, you recover the sum of the three tags SNIa+SNII+AGB. Likewise, the Fe entry of GFM_Metals roughly equals the sum of FeSNIa+FeSNII, modulo the small amount of iron consumed (i.e. negative contribution) by AGB winds. The particular fields are:

• SNIa (0): The total metals ejected by Type Ia supernovae.
• SNII (1): The total metals ejected by Type II supernovae.
• AGB (2): the total metals ejected by stellar winds, which is dominated by AGB stars.
• NSNS (3): the total mass ejected from NS-NS merger events, which are modeled stochastically (i.e. no fractional events) with a DTD scheme similar to that used for SNIa, except with a different $\tau$ value. Note that the units of NSNS are arbitrary. To obtain physical values in units of solar masses, this field must be multiplied by (MyPreferred_NSNS_MassPerEvent/NSNS_MassPerEvent), where MyPreferred_NSNS_MassPerEvent is the amount of mass ejected per NS-NS merger preferred for analysis (e.g. Shen+ 2015 uses here a value of $0.05 M_{\odot}$), and NSNS_MassPerEvent is the value we have set for the simulation, which varies by run. In particular, it is 0.05 for TNG100* and 5000.0 for TNG300* and TNG50*.
• FeSNIa (4): The total iron ejected by Type Ia supernovae alone.
• FeSNII (5): The total iron ejected by Type II supernovae alone.

(^) Although the two shock finder fields exist in subboxes, they are not updated for each subbox output, but only occasionally, so in general one should NOT use these two fields in subbox snapshots.

Subboxes

Separate "subbox" cutouts exist for each baryonic run. These are spatial cutouts of fixed comoving size and fixed comoving coordinates, and the primary benefit is that their time resolution is significantly better than that of the main snapshots. This may be useful for some types of analysis or particular science questions, or for making movies. There are two subboxes for TNG100 (corresponding to the original Illustris subboxes #0 and #2, the latter increased in size), and three subboxes for TNG50 and TNG300. Two notes of caution: first, the time spacing of the subboxes is not uniform in scale factor or redshift, but scales with the time integration hierarchy of the simulation, and is thus variable, with some discrete factor of two jumps at several points during the simulations. Second, the subboxes, unlike the full box, are not periodic.

The subboxes sample different areas of the large boxes, roughly described by the environment column in the following table. The particle fields are all identical to the main snapshots. However, the ordering differs. In particular, particles/cells in the subboxes are not ordered according to their group membership, as no group catalogs are available for these cutouts.

2. Group Catalogs

There is one group catalog associated with each snapshot, which includes both FoF and Subfind objects. The group files are split into a small number of sub-files, just as with the raw snapshots. In TNG, these files are called "fof_subhalo_tab_*", whereas in original Illustris they were called "groups_*" (they are otherwise essentially identical). Every HDF5 group catalog contains the following groups: Header, Group, Subhalo. The IDs of the member particles of each group/subgroup are not stored in the group catalog files. Rather, particles/cells in the snapshot files are ordered according to group membership.

In order to reduce confusion, we adopt the following terminology when referring to different types of objects:

• "Group", "FoF Group", and "FoF Halo" all refer to halos.
• "Subgroup", "Subhalo", and "Subfind Group" all refer to subhalos.
• The first (most massive) subgroup of each halo is the "Primary Subgroup" or "Central Subgroup".
• All other following subgroups within the same halo are "Secondary Subgroups", or "Satellite Subgroups".

FoF Halos

The Group fields are derived with a standard friends-of-friends (FoF) algorithm with linking length $b=0.2$. The FoF algorithm is run on the dark matter particles, and the other types (gas, stars, BHs) are attached to the same groups as their nearest DM particle. The fields for the FoF halo catalog are described in the following table (all fields are float32 unless otherwise specified):

Subfind Subhalos (Galaxies)

The Subhalo fields are derived with the Subfind algorithm, with modifications to add additional baryonic properties to each subhalo entry. Descriptions of all fields in this subhalo catalog are given in the following table. Note that for all mass calculations by type, wind phase cells are counted as gas.

Note: all quantities restricted to some fraction (0.5, 1.0, or 2.0) of the stellar half mass radius are by definition zero if there are no stars in subhalo.

Header

The following table describes the Header group of each groupcat file:

3. Important Additional Files

Offsets

The "offsets" files are simply helpers which facilitate rapid loading of data. If you want to use the helper scripts for working with the actual data files (snapshots or group catalogs) on your local machine, then it is required that you download the offset file(s) for the snapshot(s) you are interested in working with.

If you only work with the web-based tools and API (e.g. only analyze particle-level data using cutouts), then downloading offset files is not required.

Note that in Illustris, offsets were embedded inside the group catalog files for convenience. In TNG however, we have kept offsets as a separate HDF5 file (one per snapshot), which need to be downloaded as required. Most simply, you can think of offsets as describing where in the group catalog files to find a specific halo/subhalo, and where in the snapshot files to find the start of the particles of a given halo/subhalo.

The following table describes the fields in each offsets file.

The "simulation.hdf5" File

Each run has a single file available called simulation.hdf5 which is purely optional, for convenience, and not required by any of the helper scripts or examples. Its purpose is to encapsulate all data of an entire simulation into a single file. They can be downloaded on each simulation page.

To accomplish this, we make advantage of a new feature of the HDF5 library called "virtual datasets". You can think of a virtual dataset as a collection of symbolic links to one or more datasets in other HDF5 file(s), where these symlinks can refer to subsets of a dataset, in either the source or target of the link. The simulation.hdf5 is thus a large collection of "links", which refer to other files which actually contain data. In order to use it, you must therefore also download the necessary files (e.g. of snapshots, group catalogs, or supplementary data catalogs).

What does it let us do? First of all, no more file chunks! The fact that a snapshot or group catalog is split over multiple files is no longer relevant. Loading becomes very simple:

with h5py.File('simulation.hdf5','r') as f:
    gas_cell_mass = f['/Snapshots/99/PartType0/Masses'][()]
    subhalo_size_stars = f['/Groups/99/Subhalos/SubhaloHalfmassRadType'][:,4]

where gas_cell_mass then contains the masses of every gas cell in the entire z=0 snapshot, while subhalo_size_stars contains the stellar half mass radii of all the subhalos in the z=0 group catalog. The simulation.hdf5 also makes loading the particles of a given halo or subhalo trivial:

halo_id = 400
part_type = 1

with h5py.File('simulation.hdf5','r') as f:
    start = f['/Offsets/99/Group/SnapByType'][halo_id, part_type]
    length = f['/Groups/99/Group/GroupLenType'][halo_id, part_type]

    dm_positions = f['/Snapshots/99/PartType1/Coordinates'][start:start+length,:]

where dm_positions would then contain the dark matter coordinates of every DM particle belonging to halo index 400 of the given simulation.

Finally, supplementary data catalogs (either those we provide, or similar computations you have run yourself) can be virtually inserted at any time as datasets in snapshots or group catalogs. This provides a nice, clean way to organize post-processing computations which result in value-added values for halos, subhalos, or individual particles/cells. You can then load such catalogs with the same scripts (and same syntax) as 'original' snapshot/group catalog fields.

There are two requirements to use a simulation.hdf5 file:

• Actual data files must be organized exactly as suggested and described in the example scripts tutorial (i.e., an "output" directory containing a "snapdir_099" subdirectory and a "group_099" subdirectory, along with a "postprocessing" directory containing, among others, an "offsets" subdirectory).
• HDF5 virtual datasets are a new feature, only supported by HDF5 version 1.10.x and later. This is not presently the default version installed on most clusters, which still use 1.8.x. You may have to install a newer version yourself, or request that your system administrator do so. Note! Files created with advanced features of the new HDF5 library series are not backwards compatible. The old 1.8.x series of HDF5 reaches end of life in 2019, so it is a good idea regardless to migrate. In Python, you will also need a fairly new version of h5py (e.g. 2.9.x).

4. Merger Trees

Merger trees have been created for the various Illustris simulations using SubLink (Rodriguez-Gomez+ 2015) and LHaloTree (Springel+ 2005). The LHaloTree are essentially identical to the primary trees of the Millennium and Aquarius simulations, but in HDF5 format. Merger tree formats in TNG are identical to Illustris, with additional fields from the group catalogs also present. In the population average sense the different merger trees give similar results. In more detail, the exact merger history or mass assembly history for any given halo may differ. For any given science goal, one type of tree may be more or less useful, and users are free to use whichever they prefer. These codes are all included in the Sussing Merger Trees comparison project (Srisawat+ 2013).

The following figure shows a schematic of the structure of both the SubLink and LHaloTree merger trees. It is not necessary to understand the complete details of the trees to practically use them. In particular, the only critical links are the 'descendant' (black), 'first progenitor' (green), and 'next progenitor' (red) associations. These are shown for all tree nodes in the diagram. For their exact definitions, see the tables below. Walking back in time following along the main (most massive) progenitor branch consists of following the first progenitor links until they end (value equals -1). Similarly, walking forward in time along the descendants branch consists of following the descendant links until they end (value equals -1), which typically occurs at $z=0$. The full progenitor history, and not just the main branch, requires following both the first and next progenitor links. In this way the user can identify all subhalos at a previous snapshot which have a common descendant. Examples of walking the tree are provided in the example scripts.

Caption. Schematic diagram of the merger tree structure for both SubLink and LHaloTree. Both algorithms connect subhalos across different snapshots in the simulation. Rows indicate discrete snapshots, with time increasing downwards towards redshift zero (the horizontal axis is arbitrary). Green circles represent subhalos (the nodes of the merger tree), while beige boxes indicate the grouping of the subhalos into their parent FoF groups. The most important links are for the descendant (black), first progenitor (green), and next progenitor (red), which are shown for all subhalos. The root descendant (purple), last progenitor (blue), and main leaf progenitor (orange) links exist only for the SubLink trees, and for simplicity these last three link types are shown only for subhalos 5, 7, and 19 (darker striped circles).

The number inside each circle from the figure is the unique ID (within the whole simulation) of the corresponding subhalo, which is assigned in a depth-first fashion. Numbering also indicates the on-disk storage ordering for the SubLink trees, which adopt the approach of Lemson+ (2006). For example, the main progenitor branch (from 5-7 in the example) and the full progenitor tree (from 5-13 in the example) are both contiguous subsets of each merger tree field, whose location and size can be calculated using these links. The ordering within a single tree in the LHaloTree is not guaranteed to follow this scheme.

The 'root descendant' (purple), 'last progenitor' (blue), and 'main leaf progenitor' (orange) links exist only for the SubLink trees. For simplicity, these last three link types are shown only for nodes 5, 7, and 19 (darker striped circles). Using these links is optional, but allows efficient extraction of main progenitor branches, subtrees (i.e., the set containing a subhalo and "all" its progenitors), "forward" descendant branches, and other subsets of the tree. For their full definitions, see the following table.

Each subhalo spans a "subtree" consisting of the subhalo itself and all its progenitors. As an example, the subhalos belonging to the subtree of subhalo 5 are shown in darker green in the figure. Other subhalos not belonging to this subtree are shown in lighter green, and their links are indicated with dashed arrows. In the SubLink trees, the subtree of any subhalo can be extracted easily using the 'last progenitor' pointer. As shown in the figure, since subhalo 13 is the 'last progenitor' of subhalo 5, the subtree of subhalo 5 consists of all subhalos with IDs between 5 and 13. Similarly, the main progenitor branch of any subhalo can be retrieved efficiently using the 'main leaf progenitor' link.

Both SubLink and LHaloTree contain the links 'first subhalo in FoF group' (light brown dotted arrow) and 'next subhalo in FoF group' (dark brown dotted arrow), which connect subhalos that belong to the same FoF group. The FoF groups do not play a direct role in the construction of the merger tree. However, in SubLink, subhalos that belong to the same FoF group are also considered to be part of the same tree. As a result, two otherwise independent trees (based on the progenitor and descendant links) are considered to be the same tree if they are "connected" by a FoF group. This is exemplified in the figure by the FoF group containing subhalos 12, 16, and 20. This FoF group acts as a bridge between the left and right trees.

SubLink

The SubLink algorithm constructs merger trees at the subhalo level. A unique descendant is assigned to each subhalo in three steps (see Rodriguez-Gomez+ 2015). First, descendant candidates are identified for each subhalo as those subhalos in the following snapshot that have common particles with the subhalo in question. Second, each of the descendant candidates is given a score based on a merit function that takes into account the binding energy rank of each particle. Third, the unique descendant of the subhalo in question is the descendant candidate with the highest score. Sometimes the halo finder does not detect a small subhalo that is passing through a larger structure, because the density contrast is not high enough. {\sc SubLink} deals with this issue by allowing some subhalos to skip a snapshot when finding a descendant. Once all descendant connections have been made, the main progenitor of each subhalo is defined as the one with the "most massive history" behind it.

Note that this merger tree comes in two varieties: "dark-matter based" and "baryonic based". The dark-matter based tree is the fiducial choice, and has been publicly released as "SubLink". The baryonic based tree is called "SubLink_gal" and has not been publicly released for simplicity, although it is available upon request. While the two will largely give identical results, depending on scientific application, one might be more useful than the other.

The SubLink merger tree is one large data structure split across several sequential HDF5 files named tree_extended.[fileNum].hdf5, where [fileNum] goes from e.g. 0 to 9 for the Illustris-1 run. These files store the data on a per tree basis, and therefore are completely independent from each other. More specifically, any two subhalos that are connected by any of the pointers described in the SubLink table are guaranteed to belong to the same tree, and, therefore, their data is found in the same file. The following table lists the fields which are present in each file.

LHaloTree

The LHaloTree algorithm is virtually identical to that used for the Millennium simulation, constructing trees based on subhalos instead of main halos, described fully in the supplementary information of Springel+ (2005). In short, to determine the appropriate descendant, the unique IDs that label each particle are tracked between outputs. For a given halo, the algorithm finds all halos in the subsequent output that contain some of its particles. These are then counted in a weighted fashion, giving higher weight to particles that are more tightly bound in the halo under consideration, and the one with the highest count is selected as the descendant. In this way, preference is given to tracking the fate of the inner parts of a structure, which may survive for a long time upon infall into a bigger halo, even though much of the mass in the outer parts can be quickly stripped. To allow for the possibility that halos may temporarily disappear for one snapshot, the process is repeated for Snapshot n to Snapshot n+2. If either there is a descendant found in Snapshot $n + 2$ but none found in Snapshot n+1, or, if the descendant in Snapshot n+1 has several direct progenitors and the descendant in Snapshot n+2 has only one, then a link is made that skips the intervening snapshot.

The LHaloTree merger tree is one large data structure split across several HDF5 files named trees_sf1_99.[chunkNum].hdf5, where [chunkNum] goes from e.g. 0 to 511 for the Illustris-1 run. Within each file there are a number of groups named TreeX, where X is an integer which simply increases from zero to the number of tree groups in that file chunk. Note that a given TreeX group may contain subhalos spanning different FoF groups as well as snapshots, and to efficiently locate a specific subhalo at a specific snapshot (e.g z=0) the offsets can be used. The pair (SubhaloNumber,SnapNum) provides the indexing into the Subfind group catalog. The five other indices for each entry in a TreeX group (e.g. Descendant) index into that same group in the tree file. The following tables describe the fields. First, the Header group:

TreeX Groups:

5. Supplementary Data Catalogs

"Supplementary" or "post-processing" catalogs have been created by various members of the TNG team, extended collaboration, and general public. They typically contain valuable (and often computationally expensive) calculations, which are released here in the hope that they will accelerate the science and productivity of all users of the TNG data. Most often, they are catalogs of physical quantities calculated for halos or subhalos, at one or more snapshots. Each will have a somewhat different structure, so one should carefully read the provided documentation. To use any of these catalogs, one should download the respective HDF5 file(s) and directly load them using e.g. h5py.

(a) Tracer Tracks

The Monte Carlo tracer particles can be difficult to work with, and this dataset aims to enable some simple use cases. In particular, a number of catalogs are available, each of which contains information on all the tracers in the simulation which are contained within (all) FoF halos at redshift zero. Tracers have been re-arranged into "group order", such that you can quickly load the tracers which belong to a given subhalo or halo. Further, their evolution through time has been saved, and these "tracks" through time are immediately available, such that one does not need to load previous snapshots.

A series of files with the common base "tr_all_groups_99_*" indicating that each contains information on all the tracers in the simulation which are contained within FoF halos at snapshot 99 (z=0). The 'meta' file contains the IDs of the tracers (and their parents), which are ordered according to their halo/subhalo membership, exactly following the group-ordering of the particles in the snapshots. Therefore, the order of the ParentIDs list is the same as those IDs appear in the snapshots (although with possible omissions or duplications if a parent has no, or multiple, tracers). The 'meta' file also contains the lengths and offsets, for both halos and subhalos (in exact analogy to GroupLenType, SubhaloLenType, offsets/Group/SnapByType, and offsets/Subhalo/SnapByType). This allows the child tracers of any halo or subhalo (at z=0) to be selected.

Note that tracers are stored separately by parent type (at z=0), either gas, stars (pt4), or bhs.

All other 'data' files (e.g. temp, pos, ...) contain the evolutionary tracks of this property, for the identical set of tracers through time, recorded for all snapshots which contain tracers. For TNG100 these are the 20 full snapshots, while for TNG50 and TNG300 these are all 100 snapshots.

For example, the temperature history of the third tracer of Halo=10 in a BH parent is (python indexing convention pseudocode):

start = meta.hdf5/Halo/TracerOffset/bhs[10]
temp_K = temp.hdf5/temp[start + 2, :]
plot( temp.hdf5/redshifts[:], temp_K )

while the evolution of distance from the halo center, in units of the virial radius, of all tracers in Subhalo=20 in star parents is given by:

start = meta.hdf5/Subhalo/TracerOffset/stars[20]
length = meta.hdf5/Subhalo/TracerLength/stars[20]
rad_rvir = rad_rvir.hdf5/rad_rvir[start:start+length, :]
for i in range(length):
  plot( rad_rvir.hdf5/redshifts[:], rad_rvir[i,:] )

If you want all tracers in a halo/subhalo regardless of their z=0 parent type, then you must extract and combine the three types. If you want a subset of tracers of a given halo/subhalo at z=0 (e.g. only those in some radial shell), you must intersect the set of these IDs with the subset of meta.hdf5/ParentIDs for that halo/subhalo (or equivalently, with the entirety of meta.hdf5/ParentIDs, the only penalty being speed) and use the resulting indices.

For complete definitions of each value, see the following table and Nelson+ (2019b) where they were first used and presented. Citation to this paper is requested if you use these data catalogs. For more information and examples related to the Monte Carlo tracer particles, see Genel+ (2013), Nelson+ (2013), and Nelson+ (2015).

Simulation and snapshot coverage:

• Currently available for the 3 resolution levels of TNG100: TNG100-1, TNG100-2, TNG100-3.
• All tracers in all parent types (gas cells, stars, and black holes), in all FoF halos at $z=0$.
• For each, time tracks have twenty entries, corresponding to the twenty "full" snapshots in TNG100.

The contents of the 'meta' file are:

For the data files, all values which are computed with respect to evolving halo properties use the Sublink tree. Tracers are associated with the central subhalo of the FoF halo they belong to at z=0, and the main progenitor branch (MPB) of that subhalo is used backwards in time. If the MPB is untracked at a snapshot, these values (rad_rvir,vrad,angmom) are set to NaN at that snapshot. Values with respect to the subhalo center use SubhaloPos (maximally bound particle position), while those with respect to the subhalo velocity uses the mass-weighted mean velocity of all star particles in the subhalo (if the subhalo has no stars, these values are then recorded as NaN).

The structure of the data files is:

Important note! These files are very large. For TNG100, the one-dimensional data files (e.g. temp, entr, vrad) are ~300 GB each. Therefore, functionality has been added to the data access API to enable the retrieval of tracer tracks on a per-halo or per-subhalo basis. In particular, an API request for a halo/subhalo cutout, of particles/cells from the snapshot can specity tracer_tracks as the 'particle type', together with any of the dataset names above (namely, TracerIDs, ParentIDs, temp, entr, and so on). For example, whereas a normal cutout request for the positions of all gas cells of halo ID 1000 in TNG100-1 at redshift zero would be:

https://www.tng-project.org/api/TNG100-1/snapshots/99/halos/1000/cutout.hdf5?gas=Coordinates

A cutout request for the temperature tracks of all the tracers belonging to that halo would be:

https://www.tng-project.org/api/TNG100-1/snapshots/99/halos/1000/cutout.hdf5?tracer_tracks=temp

(b) Star Formation Rates

A catalog of time-averaged, rather than instantaneous, star formation rates of galaxies. To do so the SFRs are derived the stellar particles actually produced across different time spans, using their initial mass at birth. This is in contrast to all values in the group/subhalo catalogs, which are instantaneous values (measured from the gas cells). The time-averaged methodology, across a given period of Myr or Gyr, better reflects the values obtained via various observational tracers. For complete definitions on the calculation of each value, see the following table, Donnari+ (2019), and Pillepich+ (2019) where they were first presented. Citation to these two papers is requested if you use these data catalogs.

Simulation and snapshot coverage:

• Available for all baryonic simulations: TNG50-1, TNG50-2, TNG50-3, TNG50-4, TNG100-1, TNG100-2, TNG100-3, TNG300-1, TNG300-2, TNG300-3, Illustris-1, Illustris-2, Illustris-3.
• All snapshots/redshifts.
• Restricted to subhalos with a reasonable number of stellar particles (>~ 100).

For all measurements, time intervals {X} exist for X = 10, 50, 100, 200, 1000 (Myr).

(c) Stellar Circularities, Angular Momenta, Axis Ratios

A catalog of circularities, angular momenta and axis ratios of the stellar component of galaxies. For complete definitions on the calculation of each value, see the following table and Genel+ (2015), where they were first presented. Citation to that paper is requested if you use these data catalogs.

The first four quantities are calculated after alignment with the angular momentum vector of the stars within 10 times the stellar half-mass radius, and measure the quantities inside that radius. The Circ* fields are based on the distribution of the circularity parameter $\epsilon$. First the system is rotated such that the z-axis is aligned with the angular momentum vector as described above. Then, for every stellar particle with specific angular momentum $J_z$ we calculate $ \epsilon = \frac{ J_z }{ J(E) } $ where $J(E)$ is the maximum angular momentum of the stellar particles at positions between 50 before and 50 after the particle in question in a list where the stellar particles are sorted by their binding energy ($=U_{\rm grav} + v^2$).

Simulation and snapshot coverage:

• Available for all baryonic simulations: TNG50-1, TNG50-2, TNG50-3, TNG50-4, TNG100-1, TNG100-2, TNG100-3, TNG300-1, TNG300-2, TNG300-3, Illustris-1 (all snapshots/redshifts).
• Restricted to subhalos with stellar mass $M^\star > 3.4 \times 10^8 M_\odot$ (TNG50/TNG100/Illustris) or $M^\star > 10^{10} M_\odot$ (TNG300), measured within the usual twice stellar half mass radius, and at least 100 stars.

Note: For original Illustris (only), the "SpecificAngMom" and "Circ*" fields are available for snapshots 13 through 135, while the "*MassTensor*" fields are available for snapshots 38 through 135.

Note: In addition to the values measured within $10 R_E$, the "SpecificAngMom" and "Circ*" fields are also computed including all stars in the subhalo. These are available as the _allstars datasets.

(d) Subhalo Matching Between Runs

The possibility exists to cross-match subhalos between:

• (i) matched baryonic and dark matter only runs (e.g. TNG100-1 and TNG100-1-Dark) runs, at the same resolution.
• (ii) two baryonic realizations of the same box at different resolutions (e.g. TNG100-1 and TNG100-2)
• (iii) two simulations of the same box with different physical models (e.g. TNG100-1 and Illustris-1)

All baryonic runs have complete matching catalogs to their dark matter only counterparts, the first option (i) above.

These are data products from: Rodriguez-Gomez+ (2015; for SubLink), and Nelson+ (2015; for LHaloTree). Citation to these papers is requested if you use this data.

Simulation and snapshot coverage:

• All baryonic runs at all resolutions, all snapshots, all subhalos:
• TNG50-1, TNG50-2, TNG50-3, TNG50-4.
• TNG100-1, TNG100-2, TNG100-3.
• TNG300-1, TNG300-2, TNG300-3.

Data format: a single file subhalo_matching_to_dark.hdf5 for each simulation. Each contains one hundred groups named Snapshot_N. Within each group are two arrays, each with the same size, equal to the number of subhalos at the corresponding snapshot. The two arrays provide the results of two different matching algorithms, and should for the most part provide the same answer. Each array gives integer indices, whose value is the corresponding index of the matched subhalo in the DM only run. If no suitable match exists, a value of -1 is present for that subhalo index.

• The first array SubhaloIndexDark_LHaloTree is based on the LHaloTree matching algorithm. The matching is bidirectional, i.e. TNG <-> DMO. In each case, the best subhalo candidate is chosen as that with the largest number of matching DM particles ($\alpha=0$). Only if the candidate in each direction agrees (bijective), then these matches are saved.
• The second array SubhaloIndexDark_SubLink is based on the SubLink weighting algorithm. The direction of the matching is TNG -> DMO, i.e. for each subhalo in the baryonic physics box a best match is found in the DMO run.

The second two types (ii) and (iii) of matching catalogs (between different resolution levels, and between TNG100 and original Illustris), will be available soon.

(e) Stellar Projected Sizes

A catalog of stellar "sizes", mostly half-light radii, all in 2D projection. For more details of the modeling see Genel+ (2018), where they were first presented. Citation to that paper is requested if you use these data catalogs.

Simulation and snapshot coverage:

• Available for all baryonic simulations: TNG50-1, TNG100-1 and TNG300-1.
• Eight snapshots: z=0, 0.5, 1.0, 1.5, 2.0, 3.0, 4.0, 5.0.
• Restricted to subhalos with stellar mass $M^\star > 3 \times 10^8 M_\odot$ measured within the usual twice stellar half mass radius, and at least 100 stars.

There is one file for every snapshot, containing a single dataset /Rhalf/, which is a 10 x 5 x Nsubs shaped array. Namely, for each subhalo there are 10x5 different values of half-something radii, as follows:

• Along the dimension of length 10, the values correspond to different somethings, in order, as follows: half-mass, half-star-formation-rate, half-light in the 8 bands of the “stellar photometrics”, namely U, B, V, K, g, r, i, z. Note that the impact of dust attenuation/scattering is not taken into account.
• Along the dimension of length 5, the values correspond to projections along various lines of sight, in order, as follows: face-on, largest edge-on, smallest edge-on, random edge-on, box z-axis, where:
• The face-on direction is set according to the moment of inertia tensor of the star-formation distribution within $2 R_{1/2,\star}$ (unless there are fewer than 50 star-forming cells, in which case it is according to the moment of inertia tensor of the stellar mass distribution within $R_{1/2,\star}$.
• All three edge-on directions are from lines of sights that are perpendicular to the face-on direction. The largest and smallest edge-on directions mean the particular such lines of sight that produce the largest/smallest, respectively, projected radii for the star-formation distribution within $2 R_{1/2,\star}$. The random edge-on direction is indeed at a random line of sight within the "edge-on plane".
• For a "random" angle, choose the projection along the box z-axis, which is of course "random" with respect to each and every individual galaxy, since there is no preferred direction to the simulation.

All units are physical kpc.

Note: a value of -1 indicates that the subhalo falls outside of the selection described above, and no data is available for this object.

(f) Blackhole Mergers and Details

This catalog is comprised of two data sets: 'mergers' and 'details'. The mergers file contains a record of each BH-BH merger in the simulation. The details file contain properties of each BH at significantly higher time resolution than the snapshots. In both cases the information is extracted from output files separate from the snapshots, and so represents additional data which is otherwise not available from the snapshots alone. Here we distinguish the two BHs which participate in the merger as 'in' and 'out' BH. The difference, which during the simulation is chosen randomly by the code, is which BH ID number persists along with the remnant after the merger. The 'out' BH survives after the merger, increased in mass by that of the 'in' BH--which no longer exists after the merger event.

There exists one blackhole_mergers.hdf5 and one blackhole_details.hdf5 per baryonic run. All datasets corresponding to physical values are in code units, with masses in units of $10^{10} M_\odot / h$, mass accretion rates in units of $10.22 M_\odot / \rm{yr}$, gas densities in units of $(10^{10} M_\odot / h) / (\rm{ckpc} / h)^3$, and gas sound speeds in units of $\rm{km / s}$. They were generated with the illustris_blackholes code. Citation recommmended to Kelley et al. (2016) and Blecha et al. (2016) as appropriate.

Simulation and snapshot coverage:

• Available for all TNG baryonic runs, as well as all original Illustris baryonic runs.
• Black hole mergers: TNG50-1, TNG50-2, TNG50-3, TNG50-4, TNG100-1, TNG100-2, TNG100-3, TNG300-1, TNG300-2, TNG300-3.
• Black hole details: TNG50-1, TNG50-2, TNG50-3, TNG50-4, TNG100-1, TNG100-2, TNG100-3, TNG300-1, TNG300-2, TNG300-3.
• IMPORTANT NOTE: For both the mergers and details files, there are a few small missing data gaps. In general these files are fragile and may contain inconsistencies. In TNG100-1, this includes redshift periods [0.96 - 0.98], [1.18 - 1.20], [1.27 - 1.40], [2.06 - 2.48]. For Illustris-1, data is missing in the redshift range from approximately z = 0.14 to z = 0.38, where roughly half of the BH data is missing, except at the higher-z end of the range, where nearly all of the mergers are missing around snapshots 110-111. All other runs are unaffected. This information was lost to data corruption and cannot be recovered, and appropriate care should be taken in analysis of this data.

Blackhole mergers:

Blackhole details:

(g) Stellar Assembly

This large catalog contains information about the stellar assembly of all galaxies across all snapshots, focusing on in situ vs. ex situ stellar mass growth, the contribution from star formation before or after infall, the role of major and/or minor mergers and flyby encounters. There exists one stellar_assembly.hdf5 file per baryonic run. All datasets are masses, and are given in code units, that is, $10^{10} M_\odot / h$. There is one group per snapshot, and within each group, all datasets have the same size, corresponding to exactly one entry per Subfind subhalo. Citation recommmended to Rodriguez-Gomez et al. (2015), Rodriguez-Gomez et al. (2016a), and/or Rodriguez-Gomez et al. (2016b) as appropriate. Further information and usage examples are available in these same papers.

Simulation and snapshot coverage:

• Available for all baryonic simulations:
• TNG50-1, TNG50-2, TNG50-3, TNG50-4, TNG100-1, TNG100-2, TNG100-3, TNG300-1, TNG300-2, TNG300-3, Illustris-1, Illustris-2, Illustris-3 (all snapshots/redshifts).
• All subhalos with nonzero stellar mass.

The 'galaxy' quantities above should satisfy the following invariants (up to rounding errors):

• StellarMassInSitu + StellarMassExSitu == StellarMassTotal
• StellarMassBeforeInfall + StellarMassAfterInfall + StellarMassFormedOutsideGalaxies == StellarMassExSitu
• StellarMassFromCompletedMergers + StellarMassFromOngoingMergers + StellarMassFromFlybys + StellarMassFormedOutsideGalaxies == StellarMassExSitu

The above descriptions make reference to two values computed for every star particle in the simulation, derived as follows (see references for more details):

• AccretionOrigin: A value which can take the following integer values: 0, 1, and 2 for ex situ stellar particles that were accreted from completed mergers (i.e., when the subhalo in which the stellar particle formed has already merged with the current subhalo), ongoing mergers (i.e., when the subhalo in which the stellar particle formed has not yet merged with the current subhalo, but will do so at a later snapshot in the simulation), and flybys (i.e., when the subhalo in which the stellar particle formed has not merged with the current subhalo, and will not do so at any future snapshot in the simulation), respectively; and -1 if not applicable (i.e., if the particle was formed in situ or if it was formed outside of any subhalo). NOTE: towards the end of the simulation, it becomes impossible to distinguish a flyby from an ongoing merger. Therefore, cases (1) and (2) are usually considered as being part of the same category: "stripped from surviving galaxies".
• MergerMassRatio: The stellar mass ratio of the merger in which a given ex-situ stellar particle was accreted (if applicable). The mass ratio is measured at the time when the secondary progenitor reaches its maximum stellar mass. NOTE: this quantity was calculated also in the case of flybys, without a merger actually happening.

A small caveat: the "subhalo switching" problem" can result in some galaxies having (spurious) ex situ fractions very close to 1 (say, if a satellite suddenly becomes a central, then most of its newly assigned mass will appear as ex situ). The number of galaxies this affects is negligible, but still noticeable in e.g. a scatter plot.

(h) Subbox Subhalo List

This catalog describes the time-evolving intersection of the subhalos/merger trees with the subboxes. Allows the selection of objects of interest, which can then be traced at high time resolution using the snapshots of a given subbox. This enables science requiring high time resolution, and also visualizations of galaxy-scale evolution. This data was created in Nelson+ (2019b) and citation to that work is requested if you use this supplementary catalog.

Simulation and snapshot coverage:

• Available for all subboxes of all TNG baryonic runs (see download page for a given run for links).

Each simulation contains one file per subbox named subbox{N}_{snap}.hdf5, corresponding to subbox number {N} and the subhalo selection of all {nSubhalosFullSnap} at snapshot {snap} of the full box. Each contains the following fields:

Note that {nSubhalos} corresponds to the subset of subhalos which are ever inside the subbox. The 'SubLink' merger tree is used for all MPBs.

(i) Molecular and atomic hydrogen (HI+H2) galaxy contents

This catalog contains post-processed modeling of atomic (HI) and molecular (H2) hydrogen, on a per subhalo (i.e. per galaxy) basis. The total mass of each component is available, as well as pre-computed radial profiles. Note that all results sum, by construction, over the gas cells gravitationally bound to each subhalo only. Citation recommmended to Diemer et al. (2018) and Diemer et al. (2019) as appropriate.

Simulation and snapshot coverage:

• Available for: TNG50-1, TNG100-1, TNG100-2, TNG300-1, and Illustris-1.
• For each, redshifts: z=0, z=0.5, z=1, z=1.5, and z=2 (five snapshots per run).

In many of the datasets, the placeholder {model}_{type} appears. Here, {model} refers to a model for the HI/H2 transition, and can be one of:

• L08 - Leroy et al. 2008
• GK11 - Gnedin & Kravtsov 2011
• K13 - Krumholz 2013
• GD14 - Gnedin & Draine 2014
• S14 - Sternberg et al. 2014

Each model was computed on a cell-by-cell basis (the "volumetric" method, {type} = vol) and in 2D projection (the "map" method, {type} = map). For the L08 model, the cell-by-cell method was found to be unphysical and the L08_vol data are not included in the release for that reason.

Due to the fundamentally different nature of the volumetric and projected calculations, the radial profiles are provided in three versions, denoted as {proj} below. Profiles of projected quantities can only be expressed in projected, 2D radii (they carry the suffix {proj} = map in the profile names listed below). The profiles of volumetric quantities are given in two versions, namely as a 3D profile (suffix {proj} = 3d) and projected onto the same plane as the maps from which projected quantities were generated (suffix {proj} = 2d). When integrating the profiles to get total quantities, one should multiply the suffix {proj} = map and suffix {proj} = 2d profiles with the bin area and the suffix {proj} = 3d profiles with the bin volume. When multiplying the profiles with each other (for example, to obtain the total HI or H2 mass profiles), one should always multiply profiles of the same projection type.

The galaxy selection was designed to be complete both in stellar and in gas mass, i.e., a galaxy needs to either exceed the minimum stellar OR the minimum gas mass. For the original Illustris and TNG100, those limits were both set to $2 \times 10^8 M_\odot$. Please note that this is a rather aggressive limit, including galaxies that are resolved by only a few hundred particles. Depending on the application, it may well be wise to make more conservative mass cuts. For TNG300, the stellar mass cut is $5\times 10^{10}M_{\odot }$, the gas mass cut is $5\times 10^{9}M_{\odot }$. The stellar mass cut was chosen to be relatively high to reduce the otherwise very large dataset, and because that stellar mass range is well covered by TNG100.

In the listings below, Ngal denotes the number of subhalos (galaxies) in the file, and Nbin the number of radial bins for the profiles.

In addition, the config/ group contains a number of metadata attributes giving details of the analysis procedure:

(j) Barred Galaxies Properties and Evolution

Two catalogs are available which describe the existence, and properties, of galatic bars. They are from Rosas-Guevara+ (2020) and Zhao+ (2020), respectively.

The first catalog contains the properties of galaxy bars, and their evolution in time, of the disk galaxy sample analyzed in Rosas-Guevara+ (2020), to which citation is requested if you use this data.

Simulation and snapshot coverage:

• Available for: TNG100-1, redshift zero.
• Computed only for massive disk galaxies, defined as those with stellar mass M^* > 10^10.4 Msun, and with stellar (D/T + B/T) > 0.7, as determined by the kinematic circularity decomposition method of Genel+15.
• There are 270 galaxies which satisfy these criteria.

The first dimension of each array corresponds to a particular galaxy, and the second dimension to the 100 time snapshots. Each bar-related property is therefore given along the main progenitor branch of each galaxy. The only exception is the Bartype field with a length equal to the number of galaxies at z=0. In general, values equal to NaN mean that the calculation was not performed at this particular snapshot, and thus should be ignored. The available fields are:

This catalog was extended to the TNG50-1 simulation, as analyzed in Rosas-Guevara+ (2022), to which citation is requested if you use this data.

Simulation and snapshot coverage:

• Available for: TNG50-1 (six snapshots), z=0, 0.5, 1, 2, 3, 4.
• Computed only for massive disk galaxies, defined as those with stellar mass M^* > 10^10.0 Msun, and with stellar (D/T + B/T) > 1/2.
• There are 1062 galaxies (across all snapshots) which satisfy these criteria.

The available fields are:

The second catalog contains measurements of bar properties of the galaxy sample analyzed in Zhao+ (2020), to which citation is requested if you use this data.

Simulation and snapshot coverage:

• Available for: TNG100-1, redshift zero.
• Relatively high-mass, disk galaxies only, defined as those with stellar mass $M_\star > 10^{10} \rm{M}_{\rm sun}$ (within 30 kpc), and with the κ_rot parameter >= 0.5, measuring the fraction of kinetic energy in ordered rotation, as defined by Sales+10.
• There are 1179 galaxies which satisfy this criterion.

The available fields are:

(k) SDSS ugriz and UVJ Photometry/Colors with Dust

This catalog contains synthetic stellar photometry (i.e. colors) including the effects of dust obscuration. These correspond to the fiducial dust model of Nelson+ (2018a) (i.e. Model C), to which citation is requested if you use this data. Two separate catalogs are available: one for SDSS ugriz (rest-frame) bands, and one for UVJ (rest-frame). The latter was used in the analysis of Donnari+ (2019).

For SDSS ugriz: Simulation and snapshot coverage:

• Available for: TNG100-1.
• Available for: TNG300-1.
• For each, redshifts: z=0, 0.1, 0.2, 0.3, 0.4, 0.5, 0.7, 1.0, 1.5, 2.0, 3.0 (one file per snapshot).
• All subhalos.

For UVJ: Simulation and snapshot coverage:

• Available for: TNG100-1.
• Available for: TNG300-1.
• For each, redshifts: z=0, 0.75, 1.0, 1.75, 2.0, 2.2 (one file per snapshot).
• All subhalos.

In the single HDF5 file per snapshot, there are two datasets: subhaloIDs gives the list of subhalo indices which have been computed (this is redundant, and this array equals [0,1,2,...,nSubhalos-1] exactly.), and either Subhalo_StellarPhot_p07c_cf00dust_res_conv_ns1_rad30pkpc with a shape of [nSubhalos,8,12] (for ugriz), or Subhalo_StellarPhot_UVJ_p07c_cf00dust_res_conv_z_30pkpc with a shape of [nSubhalos,3] (for UVJ).

The dimensions 8 or 3 corresponds to photometric bands, in order: sdss_u, sdss_g, sdss_r, sdss_i, sdss_z, wfc_acs_f606w, des_y, jwst_f150w or u, v, 2mass_j. Note that the wfc, des, and jwst bands were for testing only and should not be used. The high redshift SDSS files have only 5 photometric bands saved, which are (in order): sdss_u, sdss_g, sdss_r, sdss_i, sdss_z.

For the SDSS ugriz catalogs, the dimension 12 corresponds to twelve different projection directions (i.e. observer view angles), since the dust attenuation mode is view-dependent. In general, one can simply take the first entry for each subhalo, or a random entry for each subhalo. For the UVJ catalogs, only one projection direction is computed (along the z-axis of the box).

The units of these datasets are always AB magnitudes, and the bands are always rest-frame. Check carefully if the magnitudes are apparent (usually the case), or absolute (e.g. at z=0). Only stars within a 3D radial aperture of 30 physical kpc are included.

Additional information is available in the header attributes of the HDF5 file.

(l) SKIRT Synthetic Images and Optical Morphologies

This catalog contains synthetic images and the corresponding morphological measurements, including Gini-M20, CAS and MID statistics, as well as 2D Sersic fits. The full description of the synthetic images and measurements can be found in Rodriguez-Gomez+ (2019), to which citation is requested if you use this data. Full details are not repeated, for which the user is directed to the paper.

Synthetic images and measurements are available separately for: "pogs" and "sdss", designed to match Pan-STARRS and SDSS, respectively.

Simulation and snapshot coverage:

• Available for: TNG50-1, TNG100-1, and Illustris-1.
• For each, redshifts: z=0, z=0.05.
• Restricted to subhalos with (total) stellar mass M^* > 10^9.5 solar masses.

There are two ways to download this data.

• The entire dataset (for one survey/snapshot combination) can be downloaded as a single tar file (above).
• The FITS and PNG images can be downloaded separately for each subhalo for which they are available through the API, similar to the original Illustris FITS files. For example, /api/TNG100-1/snapshots/99/subhalos/397866/skirt/broadband_sdss.fits downloads the broadband FITS file matched to the SDSS survey, for the central subhalo of FoF group #500 in TNG100-1 at redshift zero.

In the single tar collection, the files "morphs_i.hdf5" and "morphs_g.hdf5" (and "morphs_r.hdf5" for SDSS only) contain statmorph morphological measurements in the observed i-band, g-band, (and r-band), respectively (using Pan-STARRS or SDSS filter curves, depending on the dataset), and "subfind_ids.txt" contains the associated subhalo IDs. For a description of the measurements, see Section 4 of Rodriguez-Gomez+ (2019). In general, only morphological measurements with flag == 0 are reliable. If interested in parameters derived from the Sérsic fits, then flag_sersic == 0 should be imposed as well. In addition, only measurements with S/N > 2.5 (included in the catalog as sn_per_pixel) should be trusted. See Section 5 of the paper, for more details.

The idealized synthetic images themselves are available in FITS format. The dimensions of each image are NxNx4, where N is the number of pixels in each dimension and the 4 layers correspond to the observed g,r,i,z broadband filters of Pan-STARRS or SDSS. Note that the SDSS z-band data should be used with caution (they are reasonable for morphologies, but not for magnitudes and colors), since the long-wavelength tail of the SDSS z filter curve extends beyond the wavelength range of the SKIRT runs (clipped at 0.95 microns in the rest frame).

These synthetic broadband images are "idealized", meaning that some realism may need to be added before analyzing them. Most importantly, they should be convolved with an adequate PSF (for most purposes, a 2D Gaussian with the same FWHM as the observations should suffice) and background shot noise should be added. In general, Gaussian noise with sigma_sky ~ 1/20 (1/10) for Pan-STARRS (SDSS) represents a good compromise between observational realism and detection strength. Note that the image units are electrons/s/pixel. For more details, see Section 3.4 of the paper.

In addition to the FITS files, PNG images are also available. These are analogous to Fig. 4 of Rodriguez-Gomez+ (2019), such that different panels show different morphological diagnostics. These figures can be used to quickly inspect the morphology of a given galaxy.

Notes: The z=0 images and measurements were created by assuming that the galaxies are actually located at z=0.0485 (this determines the pixel scale, etc.). Galaxy sizes (e.g. "rhalf", "sersic_rhalf", "rpetro", "rmax") are given in pixels. To convert to simulation units (ckpc/h), note that the scale of each pixel at z=0.0485 corresponds to 0.174 ckpc/h for Pan-STARRS and 0.276 ckpc/h for SDSS.

2022 addition: KiDS Survey Mock Images. A new, third set of images were generated with the same procedure as in Rodriguez-Gomez et al. (2019), except that:

1. the field-of-view is fixed for all images (does not scale with galaxy size), and
2. neighboring galaxies (from the same FoF group) were included.

The image settings are consistent with the Kilo-Degree Survey (KiDS). In particular, each FITS file has 4 layers corresponding to OmegaCAM g,r,i,z filters.

Simulation and snapshot coverage:

• Available for: TNG50-1.
• Redshift: z=0.0337 (snapshot 96) only.
• Restricted to subhalos with (total) stellar mass M^* > 10^8.0 solar masses.

There are three projections for each galaxy (xy, yz, zx). The pixel scale: 0.2 arcsec/pixel, corresponding to 138.9 pc (physical) per pixel at z = 0.0337. The dimensions of each image are 240x240 pixels (33.34 kpc per side). Note that the most massive galaxies may extend beyond the FoV.

The image units are ADU/s, just like in the real survey, and the zeropoint is zero, i.e. mag = -2.5 * log10(data). The images are "idealized", i.e. they have not been convolved with a PSF and they do not include background or shot noise. See Section 2.4 from Guzmán-Ortega et al. (2022) for information on how to apply realism. Complete details are available in that paper, to which citation is requested if you use this dataset.

(m) Disk Galaxy Kinematic Decompositions

This catalog contains mass fractions of galactic disk structures which have been kinematically identified and decomposed by applying the "auto-GMM" (Gaussian mixture model) technique. The full description of the synthetic images and measurements can be found in Du+ (2019) and Du+ (2020), to which citation is requested if you use this data.

Simulation and snapshot coverage:

• Available for: TNG100-1, redshift zero and TNG50-1, redshift zero only.
• Computed only for massive "disk galaxies", defined as those subhalos with a high relative fraction of their kinetic energy in ordered rotation (K_rot > 0.5, measured within 30 kpc), and high stellar mass (M^* > 10¹⁰ Msun for TNG100-1, and M^* > 10⁹ Msun for TNG50-1).
• There are 3931 subhalos in TNG100-1 (2831 in TNG50-1) which satisfy this criterion.

The classification 1 (cls1) decomposes each galaxy into a cold disk, warm disk, bugle, and halo. The classification 2 (cls2) isolates an additional structure with moderate rotation, but compact morphology, namely the disky bulge that is the counterpart of the so-called pseudo bulge in observations (see Du et al. 2020). The supplementary catalog file has the following fields:

Here {method} can have the values class1,class2, corresponding to the two classification techniques, and {aperture} can have the values 1re,3re,allstars, corresponding to using stars within 1 times r_e, 3 times r_e, or in the entire subhalo (where r_e is the 3D stellar half-mass radius).

Associated PNG or PDF images, showing the decomposed structures, can be downloaded here: PNG .tar.gz image set, or PDF .tar.gz image set. These archives contain the following directories:

• "total_level": images of all stars.
• "SD_level": images of the spheroidal and disky structures.
• "structures_level_cls1/2": images of the kinematic structures defined in Du+2020. cls1 and cls2 correspond to classifications 1 and 2, respectively.
• "components_level_cls1/2": images of the multiple Gaussian components from the auto-GMM, that correspond to sub-structures.

(n) L-Galaxies Semi-Analytical Model

This catalog contains the results of the L-Galaxies semi-analytical model run on the TNG simulation volumes. Specifically, of the latest public version of this model, as described in Henriques+2015 (H15). The SAM itself is run on the merger trees of the dark matter only analog boxes of TNG100-1 and TNG300-1.

This catalog was produced in Ayromlou+ (2020), where these results were first presented, and to which citation is requested if you use this data.

Simulation and snapshot coverage:

• Available for: TNG100-1.
• Available for: TNG300-1.
• All snapshots, and all (L-Galaxies) galaxies, covering all TNG subhalos.

The outcome of L-Galaxies is a population of galaxies which reside within the halos and subhalos of TNG100-1-Dark or TNG300-1-Dark, respectively. These have been cross-matched to the subhalos of the baryonic TNG100-1 and TNG300-1 runs (as well as the subhalos in the corresponding -Dark runs). As a result, the galaxy properties predicted by L-Galaxies can be directly compared with those in TNG, on an object by object basis.

There is one HDF5 file per TNG snapshot, with the following datasets in the /Galaxy/ group:

Note: For all units, ckpc stands for kpc length scale in comoving units. Multiply by the scale factor $a=1/(1+z)$ to convert to physical units. Additional information is available in the header attributes of the HDF5 file.

For each of the three Mag* datasets, all magnitudes are rest-frame, absolute (AB). The twenty entires correspond to (in order):

As an example, the following (python) code shows how this L-Galaxies catalog can be matched and compared to TNG galaxies from the group catalogs:

filePath = "sims.TNG/L75n1820TNG/postprocessing/LGalaxies/LGalaxies_099.hdf5"

with h5py.File(filePath,'r') as f:
  mstar = f['Galaxy/StellarMass'][()]
  match_ids = f['Galaxy/SubhaloIndex_TNG'][()]

w = np.where(match_ids >= 0)

mstar_matched = np.zeros( mstar_tng.shape, dtype=mstar.dtype )
mstar_matched.fill(np.nan)

mstar_matched[match_ids[w]] = mstar[w]

The resulting array mstar_matched has the same size as mstar_tng (which should be loaded from the TNG subhalo catalog at this snapshot, e.g. SubhaloMassInRadType[i,4]), and the value mstar_matched[i] can be compared directly to mstar_tng[i].

(o) Lensing Profiles

This catalog contains lensing profiles of halos/subhalos, as described in Renneby+20 to which citation is requested if you use this data.

Simulation and snapshot coverage:

• Available for: TNG100-1, TNG100-1-Dark, TNG300-1, TNG300-1-Dark.
• Redshift zero (snapshot 99) only, all subhalos above a stellar mass limit (see below).

Lensing profiles have been computed for all galaxies with $M_\star \geq 10^{8.2} h^{-2} M_\odot$ in the baryonic runs, and their matched gravity-only counterparts in the dark matter only runs. In total there are profiles for 417,535 galaxies for TNG300-1 and 334,606 matched substructures in TNG300-1-DMO (matching rate of ~80%). For TNG100 there are 35,270 galaxies in TNG100-1 and 27,088 galaxies in TNG100-1-DMO (~77% matched). The DMO matches are made with the LHaloTree-algorithm.

We have measured the signal radially for field projections along the x, y and z axes in 40 log10-equidistant bins between 30 kpc/h and 3 Mpc/h, and have separate "core terms" where we store all particle/cell counts below 30 kpc/h.

There is one HDF5 file per simulation/snapshot, with the following datasets:

Finally, we provide a supplementary .zip file for download. It contains several items:

• an example selection for TNG300 in ”tng300_example_selection_galaxies_mstar_10p11_10p11p2_Msun.hdf5” and an example compute script in ”compute_delta_sigma_signal_example_script.py”.
• Lensing profiles columnated in .txt-files where the first column holds the radius r [h^-1 Mpc], the second column the model Delta Sigma [h M_sun pc^-2], the third column the upper bootstrapped 95 % model Delta Sigma [h M_sun pc^-2], and the fourth column the lower bootstrapped 95 % model Delta Sigma [h M_sun pc^-2]. They are stored in folders referring to the reference of the dataset we are attempting to model.
• Baryonic deformations of the lensing profiles are stored in subdirectories ”BaryonicEffects” for the van Uitert+16 and Velliscig+17 datasets. Note that we use matched subhaloes only and restrict ourselves to centrals.
• The corresponding SAM-profiles are found in subdirectories for the different datasets under the folder ”SAMs”. The title of the file holds the parameter settings the model was run on. Note that the Guo+11 model parameter values for the Henriques+15 model is referred to as ”guo11”, whereas the original Guo+11 model is referred to as ”guo11_makefile_switch”.

(p) Aperture Masses (Total and Stellar)

Measurements of 3D total and stellar masses within physical apertures, accounting for the resolution elements that are gravitationally bound to a subhalo according to the SUBFIND algorithm within 100 (physical) kpc, 30 pkpc, 10 pkpc, and 5 pkpc. For complete definitions on the calculation of each value, see the following table and Engler+ (2021a) where they were first presented. Citation to this paper is requested if you use these data catalogs.

Simulation and snapshot coverage:

• Available for all TNG baryonic simulations: TNG50-1, TNG50-2, TNG50-3, TNG50-4, TNG100-1, TNG100-2, TNG100-3, TNG300-1, TNG300-2, TNG300-3.
• All full snapshots between z=0 and z=1.5 (i.e. 99, 91, 84, 78, 72, 67, 59, 50, 40).
• Restricted to subhalos with at least one star (for TNG50 and TNG100), or at least 1000 stars (for TNG300).

Note: these are original values directly from the simulation. No "resolution rescalings" have been applied.

For all measurements, apertures {X} exist for X = 5, 10, 30, 100 (pkpc).

(q) Halo Structure

Measurements of halo structural properties such as concentrations, shapes, formation times, and so on. All are computed using the dark matter particles that both (i) belong to the central/primary Subfind subhalo, and; (ii) are within R200c of the halo. For complete definitions of each value, see the following table and Anbajagane+ (2021) where they were presented. Citation to this paper is requested if you use these data catalogs.

Simulation and snapshot coverage:

• Available for the highest resolution run of each TNG box, both baryonic and dark-matter only: TNG50-1, TNG100-1, TNG300-1, TNG50-1-Dark, TNG100-1-Dark, TNG300-1-Dark.
• All twenty full snapshots (i.e. $z=0, 0.1, 0.2, 0.3, 0.4, 0.5, 0.7, 1.0, 1.5, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12$).
• Restricted to halos with total masses greater than $10^{8.8} M_\odot$ (for TNG50), $10^{9.8} M_\odot$ (for TNG100), and $10^{10.8} M_\odot$ (for TNG300).

Note: one file per snapshot. Be careful to use only entries with GroupFlag == 1.

(r) Galaxy Morphologies (Deep Learning)

Probabilities that galaxies have certain visual-like morphologies, as determined by a machine learning algorithm. In the case of TNG50, classifications are provided at high redshift and are CANDELS-like, i.e. based on synthetic H-band galaxy images at CANDELS depth and resolution. In the case of TNG100, classifications are provided at low redshift, and are SDSS-like, based on synthetic r-band galaxy images with SDSS resolution and noise characteristics. For complete definitions on the calculation of each value, see the following table, Huertas-Company+ (2019), and Varma+ (2021), where they were first presented and used. Citation to these papers is requested if you use these data catalogs.

Simulation and snapshot coverage:

• Available for TNG50-1 (z=0.5, 1, 1.5, 2, 2.5, 3), and TNG100-1 (z=0, 0.05).
• Restricted to subhalos with $M_\star > 10^9 M_\odot$ (for TNG50 classifications), or $M_\star > 10^{9.5} M_\odot$ (for TNG100, i.e. 12,468 galaxies at z=0.05).

For TNG100 (SDSS-like):

For TNG50 (CANDELS-like):

Note: a Snapshot_N group exists for the above-mentioned snapshots (two for TNG100-1, six for TNG50-1).

(s) Cosmic Web Distances (Disperse)

A post-processing identification of the topology of the cosmic web, using the DisPerSE code. This method provides an automatic identification of topological structures such as peaks, voids, walls, and in particular filaments. To do so, DisPerSE uses a set of discrete points (e.g. galaxies) to estimate a density field, and then identify the cosmic web. By definition critical points in this density field are nodes, and the unique integral lines between them are filaments: every filament starts and ends at a node, and saddle points are minima along the filaments.

For this catalog, the points used to define the cosmic web are galaxies. Specifically, subhalos with a minimum stellar mass of $10^{8.5} M_\odot \rm{h}^{-1}$, which is designed to be roughly comparable to what you might be able to recover from an observational galaxy survey. The only other parameter of DisPerSE is the "persistence", which defines the robustness of each pair of critical points, in terms of the peak relative to background. For this catalog a persistence of $\sigma = 4$ has been chosen.

The physical properties presented in this catalog are cosmic web distances, between subhalos in the simulation and the nearest cosmic web structures of a given type. This allows for the identification, for example, of galaxies as a function of distance away from the centers of filaments, void galaxies, and so on.

For more details, see the DisPerSE paper Sousbie (2011). This catalog was developed as part of the work of Duckworth+ (2020a) and Duckworth+ (2020b). Citation to these three papers is suggested if you use these data catalogs. The codes to run DisPerSE on TNG and to derive quantities of interest are available on github.

Simulation and snapshot coverage:

• Available for the highest resolution run of TNG100 and TNG300 currently: TNG100-1, TNG300-1.
• Several full snapshots between redshift two and zero (i.e. $z=0, 0.1, 0.2, 0.3, 0.5, 1.0, 1.5, 2$).
• All subhalos.

Note: in the above, N_subhalos always refers to the number of subhalos in the group catalog at a particular snapshot. All distances are in code units.

(t) Galaxy Morphologies (Kinematic) and Bar Properties

This catalog contains (i) kinematic decompositions of the stars of galaxies into different morphological components, and (ii) bar properties. Galaxies are decomposed in five morpho/kinematical components, namely a thin/cold disc, a thick/warm disc, a pseudo-bulge, a bulge and a halo, after an analysis of the stellar kinematics with the MORDOR code. In addition, a detailed catalogue of stellar bars is given, along with bar properties, based on a fourier analysis of the galactic stellar surface density. For complete definitions on the calculation of each value, see Zana+ (2022), where they were first presented and used. Citation to this paper is requested if you use these data catalogs.

Simulation and snapshot coverage:

• Available for TNG50-1 (all redshifts).
• Restricted to subhalos with $M_\star \gtrsim 10^9 M_\odot$ (corresponding to a minimum of $10^4$ star particles).

This catalogue is one file per simulation, containing many groups Snapshot_N, one per snapshot. The following datasets are included:

Note: for TNG50-1, the available snapshot numbers N range from 6 to 99.

Note: if no bar structure is found, the related entries are set equal to -1.

(u) LEGA-C/VIMOS Mock Optical Spectra

This catalog contains mock (i.e. synthetic) optical spectra of galaxies at intermediate redshift, designed to match the instrumental properties and setup of the LEGA-C Survey. This spectroscopic survey obtained deep, high-resolution spectra of thousands of galaxies with $M_\star > 10^{10} \rm{M}_\odot$ from $z=0.6$ to $z=1$, using the VLT/VIMOS multi-object slit spectograph.

These synthetic spectra have been designed to cover TNG galaxies over the same redshift and mass ranges, such that LEGA-C matched samples can be constructed. The spectra have been created to exactly mimic the instrumental and survey characteristics, such as spectral resolution, noise, seeing, and so on. Each extends over a rest-frame wavelength range of $\sim 3700$ Angstrom to $\sim 5200$ Angstrom. For every TNG galaxy, 10 realizations are provided, by matching to 10 different random LEGA-C galaxies of similar mass. In addition, the original 'idealized' spectra are also included, one per TNG galaxy.

Complete details on the methodology are presented in Wu+ (2021), where they were first analyzed. Citation to that paper, as well as to Nelson+ (2018) where the mock spectra construction was developed, is requested if you use this data catalog.

Simulation and snapshot coverage:

• Available for TNG100-1 (five snapshots, z=0.6, 0.7, 0.8, 0.9, and 1.0).
• Restricted to subhalos with $10^{10.3} < M_{\star} / \rm{M}_\odot < 10^{11.5}$ (stellar masses measured within 30 pkpc).

This catalogue is one single file, containing five groups named Snapshot_N, one per snapshot. The following datasets are included in each snapshot group:

Note: for TNG100-1, the available snapshot numbers N are 50, 53, 56, 59, and 63. The corresponding values of {N_gal} are 3058, 3223, 3371, 3503, and 3624.

Note: the dimension of size 10 in the above datasets corresponds to 10 realizations of mock spectra for each TNG galaxy (see paper for details).

Note: For all spectra, {N_wave} is a constant 6166 wavelength points.

(v) JWST-CEERS Mock Galaxy Imaging

This catalog contains synthetic images for JWST-like NIRCam and MIRI observations of high-redshift galaxies. The full description of the synthetic images and measurements can be found in Costantin+ (2022), to which citation is requested if you use this data. For full details, please see the paper.

Simulation and snapshot coverage:

• Available for: TNG50-1.
• Redshifts: z=3, z=4, z=5, and z=6.
• Restricted to subhalos with stellar mass $M_\star / \rm{M}_\odot > 10^{9.0}$ as well as total half-mass radius $R_{\rm half} \gtrsim 1\, \rm{arcsec}$.
• For each galaxy, there are 20 view configurations (5 inclinations: i=[0, 45, 90, 135, 180], and 4 azimuths: a=[0, 90, 180, 270]).

Each dataset (for one instrument configuration and redshift) can be downloaded as a single tar.gz file (below).

• NIRCam SW: 0.031 arcsec/px: z=3, z=4, z=5, z=6
• NIRCam LW: 0.063 arcsec/px: z=3, z=4, z=5, z=6
• MIRI: 0.110 arcsec/px: z=3, z=4, z=5, z=6
• Highly-resolved "ideal" TNG50 0.010 arcsec/px: z=3, z=4, z=5, z=6

As an additional dataset, simulated imaging data suited for CEERS observations (F200W and F356W filters) are provided. The instrumental noise effects of JWST/NIRCam were simulated using the mirage tool (v2.2.1). Data simulated with mirage were run through the JWST calibration pipeline (v1.4.6), mimicking the data reduction strategy to be used for in-flight data, and are provided at nominal and half-nominal angular resolution (see detailed information). Morphological catalogs based on Costantin+2023 are also provided (coming soon).

• NIRCam F200W (0.031 arcsec/px and 0.015 arcsec/px): z=3, z=4, z=5, z=6
• NIRCam F356W (0.063 arcsec/px and 0.030 arcsec/px): z=3, z=4, z=5, z=6

The synthetic images are available as datacubes in FITS format. Each dataset contains the same 4 redshift bins (galaxies from four simulation snapshots). The ‘EXTn’ keyword in the main header can be used to identify the different filters available (n=1-12 for NIRCam SW, n=1-15 for NIRCam LW, n=1-9 for MIRI, and n=1-36 for the resolved TNG50 version). Catalogs containing the main properties of each synthetic image are also available.

(w) Multi-band (UV to submm) Photometry and SEDs

This catalog contains mock (i.e. synthetic) multi-wavelength photometry and spectral energy distributions (SEDs), across a broad wavelength range, from the UV to the submillimeter. These are highly realistic values, obtained with the SKIRT dust radiative transfer code, and with a calibration against DustPedia data. These band magnitudes/fluxes and SEDs are available for TNG50 galaxies at low redshift: $z=0$ and $z=0.1$ in particular.

Complete details on the methodology are presented in Trčka+ (2022) and Gebek+ (2024), where they were created and first analyzed. Citation to these two papers is requested if you use this data catalog.

Simulation and snapshot coverage:

• Available for TNG50-1, TNG50-2, and TNG100-1 (in all cases, for two snapshots, z=0.0 and z=0.1).
• Restricted to subhalos with $M_{\star} > 10^{8} \rm{M}_\odot$ for TNG50 (or $M_{\star} > 10^{8.5} \rm{M}_\odot$ for TNG100), stellar masses measured within twice the stellar half mass radius.
• This corresponds to 7375 (5669) [30712] galaxies at $z=0$ and 7302 (5665) [30364] galaxies at $z=0.1$ for TNG50-1 (TNG50-2) [TNG100-1].

This catalogue is one single file per simulation, containing groups named Snapshot_N, one per snapshot. The following datasets are included for each snapshot:

Note: {N_apertures} is always 4, and these correspond to (in order): 10 kpc, 30 kpc, twice the stellar half mass radius, and five times the stellar half mass radius. (Also listed in the HDF5 attributes).

Note: {N_orientations} is always 3, and these correspond to (in order): edge-on, face-on, and random. (Also listed in the HDF5 attributes).

Note: {N_bands} is always 53, and these correspond to (in order): GALEX_FUV, GALEX_NUV, SDSS_u, SDSS_g, SDSS_r, SDSS_i, SDSS_z, TwoMASS_J, TwoMASS_H, TwoMASS_Ks, UKIDSS_Z, UKIDSS_Y, UKIDSS_J, UKIDSS_H, UKIDSS_K, Johnson_U, Johnson_B, Johnson_V, Johnson_R, Johnson_I, Johnson_J, Johnson_M, WISE_W1, WISE_W2, WISE_W3, WISE_W4, IRAS_12, IRAS_25, IRAS_60, IRAS_100, IRAC_I1, IRAC_I2, IRAC_I3, IRAC_I4, MIPS_24, MIPS_70, MIPS_160, PACS_70, PACS_100, PACS_160, SPIRE_250, SPIRE_350, SPIRE_500, SCUBA2_450, SCUBA2_850, ALMA_10, ALMA_9, ALMA_8, ALMA_7, ALMA_6, PLANCK_857, PLANCK_545, PLANCK_353. (Also listed in the HDF5 attributes). For the values at redshift zero, all instruments were placed at a distance of 20 Mpc.

Note: {N_wave} is always 387, and the values are given by the SEDs_wave dataset.

Finally, we provide a supplementary .zip file for download. This contains: (i) a Jupyter notebook which shows examples of loading and exploring the data, and (ii) the SKIRT .ski configuration file used to perform the radiative transfer runs.

(x) Globular Clusters

This catalog contains a dataset of globular clusters for high-mass halos. The existence, and properties, of globular clusters (GCs) has been inferred in post-processing with a 'tagging' technique. At the time when a galaxy infalls (i.e. accretes) into a more massive host, its GC population is created. Each GC is attached to (i.e. represented by) a single dark matter particle, which sets its subsequent gravitational dynamics and location.

Complete details on the methodology and physical modeling assumptions are presented in Doppel+ (2020), where this GC catalog was created and first analyzed. Citation to that paper is requested if you use this data. Additional details are available in Ramos+ (2015), Ramos-Almendares+ (2018), and Ramos-Almendares+ (2020), which describe previous versions and the evolution of the GC tagging model.

This catalog specifically contains the "realistic GCs" from Doppel+ (2020), which is one realization of the GC population, subsampled from a larger ensemble of candidate GC populations. It has been designed to roughly reproduce realistic GC properties at redshift zero. Additional realizations, as well as the entire set of GC candidates, can be made available upon request.

Simulation and snapshot coverage:

• Available for TNG50-1, redshift zero (snapshot 99).
• Restricted to halos with $M_{200c} > 5 \times 10^{12} \rm{M}_\odot$. More specifically, GCs are present within subhalos with stellar masses $M_{\star} \geq 10^{6} \rm{M}_\odot$ at $z=0$ which have additionally reached a maximum stellar mass of at least $5 \times 10^{6} \rm{M}_\odot$ in their lifetime, and had at least 100 dark matter particles and non-zero stellar mass at infall.
• This corresponds to the subhalos of 39 massive host halos at $z=0$, and a total of 196,606 GCs.

This catalogue is one file, which includes the following datasets:

(y) Merger History

Catalogs containing information and statistics about the merging history of all subhalos (i.e. galaxies), across all time. Here mergers are split into three categories: major (stellar mass ratio > 1/4), minor (stellar mass ratio between 1/10 and 1/4), and all (any stellar mass ratio) mergers. The mass ratio is always based on the stellar masses of the two merging galaxies at the time when the secondary reached its maximum stellar mass.

To avoid spurious flyby and re-merger events, mergers are only included when both galaxies can be tracked back to a time when each of them belonged to a different FoF group. Furthermore, to clean the mergers from events released to non-cosmological subhalos (i.e. "clumps"), a cleaning procedure has been applied as follows. First, all subhalos with SubhaloFlag == False are ignored. Second, the secondary is ignored if it did not exist for more than 2 snapshots.

For complete definitions on the calculation of each value, see the following table, Rodriguez-Gomez (2017), and Eisert et al. (2022). Citation to these two papers is requested if you use these data catalogs.

Simulation and snapshot coverage:

• Available for all TNG baryonic simulations: TNG50-1, TNG50-2, TNG50-3, TNG50-4, TNG100-1, TNG100-2, TNG100-3, TNG300-1, TNG300-2, TNG300-3.
• All snapshots: one HDF5 file per snapshot.
• All subhalos.

Major mergers (stellar mass ratio > 1/4), minor mergers (1/10 < stellar mass ratio < 1/4), and all mergers (any stellar mass ratio):

Note: in all cases above, {time} can be any of six options: "Last250Myr", "Last500Myr", "LastGyr", "SinceRedshiftOne", and "SinceRedshiftTwo".

Note: if a snapshot number is undefined, i.e. there is no major merger for SnapNumLastMajorMerger, the value is set to -1.

There are also additional merger statistics that attempt to quantify the cumulative effects of all the mergers that a galaxy has ever had. These are plausibly more meaningful quantities than the "standard" merger statistics described above (e.g. the time since the last major merger, etc.). As usual, the properties of the secondary progenitor (as well as the merger mass ratio) are measured at the time when it reached its maximum stellar mass, while the time of the merger corresponds to the snapshot in which the two merger tree branches join. As shown in Rodriguez-Gomez et al. (2017), these quantities are strongly correlated with the fraction of accreted stars in a galaxy. Available quantities are:

Finally, there are additional merger statistics, with analog meanings as above, unless otherwise stated:

Note: in all cases above, {time} can be any of four options: "Last2Gyr", "Last5Gyr", "Last8Gyr", and "SinceRedshift5".

Finally, in the case of TNG50-1 catalogs (only), each file contains an additional group named "WithConstraint". In this group, a second version of all the above datasets can be found, where an additional constraint has been applied to secondaries. Specifically, secondaries are ignored if there were not more than 50 stellar particles in at least one snapshot in its lifetime. These merger statistics exclude very small galaxies, and may be useful. They reproduce the published results of Eisert+ (2022) and Sotillo-Ramos+ (2022).

(z) MaNGIA Mock MANGA IFU Cubes

This catalog contains 10,000 MaNGA-like (IFU datacube) mocks from TNG50 galaxies for stellar population analysis. To enable a statistical comparison with the observations, the sample of simulated galaxies has been chosen to match the MaNGA selection criterion based on stellar mass, radius and redshift. The mock datacubes emulate the instrumental characteristics of the observed survey, and have been analyzed with the same tools to extract stellar populations properties. The final products are comparable the latest release of the analysis performed with pyPipe3D on the MaNGA survey (Sánchez+ 2022).

The full description of the sample, mock procedure, and data products can be found in Sarmiento+ (2023), to which citation is requested if you use this data.

The cubes where produced using the SSP template MaStar_CB19.slog_1_5 (Sánchez+ 2022) assuming a dust screen model and later analyzed with pyPipe3D. The code to produce the mocks is based on Ibarra-Medel+ (2019). For this work, we use an updated version of this code, available here.

Simulation and snapshot coverage:

• Available for: TNG50-1.
• Sample: Selection of 10,000 galaxies from TNG50-1 snapshots 87 through 98 (inclusive), chosen to match MaNGA properties.

Bulk download of the mocks, for all galaxies, is available. They can be downloaded as two tar.gz files:

• MaNGIA_catalog.fits (1 MB)- list of the 10,000 TNG50 galaxies used to produce the mocks (and the corresponding snapshots).
• MaNGIA_maps.tar.gz (35 GB) - stacked 2D property maps (stellar mass, stellar kinematics, intrinsic and assigned ages and metallicities) derived directly from the simulation with the same spatial resolution as the cube mocks. Also includes the relevant pyPipe3D outputs: stellar absorption indices, "SFH", and "SSP". The SFH extension contains the recovered weights of the linear combination of stellar populations, while the SSP extension has the main parameters derived from the analysis of the stellar populations (LW and MW ages, metallicities, dust attenuation and stellar kinematics properties).
• MaNGIA_rss.tar.gz (345 GB) - row-stacked spectra. Files containing the list of emulated fiber spectra corresponding to each galaxy. The 3D-datacubes can be re-built from the row-stacked spectra (RSS) files using the provided code.

In the "maps" and "rss" .tar.gz files, there are one or more .fits files per subhalo. They are named with the format ""TNG50-{snap}-{subhaloID}-{view}-{numfibers}.fits" where {snap} is the snapshot number, {subhaloID} is the Subhalo ID, {view} is an integer from 0-5 (see paper), and {numfibers} is the number of fibers in the bundle (19, 37, 61, 91, or 127, see paper). Note: the majority of subhalos are only viewed once, so they have only one corresponding .fits file with {view} == 0.

(1) TNG50 Milky Way+Andromeda (MW/M31)-like Galaxies

The TNG50 Milky Way + Andromeda sample, as presented in Pillepich+ (2023), contains detailed information and special snapshot cutouts for 198 MW/M31-like galaxies. Three data products are available: (i) a catalog of the sample, including many physical properties of each MW/M31-like galaxy, (ii) a catalog of the satellite galaxies surrounding each MW/M31-like central, and (iii) individual snapshot cutouts of the stars, dark matter, gas, and supermassive black holes which reside in, and around, each galaxy, at $z=0$ as well as following the main progenitor for all redshifts up to $z=7$.

See the TNG50 Milky Way+Andromeda Sample and Data Release page.

(2) HVC-like Cosmological Cloud Catalog

This is the Cosmological Cloud Catalog (CCC), a library of physical properties of cool clouds in the circumgalactic medium of Milky Way-like galaxies from TNG50. These objects are possible High Velocity Cloud (HVC) and/or Intemediate Velocity Cloud (IVC) analogs. For complete definitions on the calculation of each value, see the following table and Ramesh+ (2023b), where they were first presented. Citation to that paper is requested if you use this catalog.

These clouds can be visualized in the Ramesh+ (2023b) Infinite Gallery.

Simulation and snapshot coverage:

• Available for TNG50-1 at z=0 (snapshot 99).
• Restricted to a subset of 132 galaxies from the MW/M31-like sample, corresponding to those with $10^{10.5} \rm{M}_\odot < M^\star < 10^{10.9} \rm{M}_\odot$.

The single HDF5 file consists of various datasets. Each dataset is an array of length 119895, and each entry corresponds to a single cloud. Clouds here are defined as contigous sets of Voronoi cells that are cold ($T ≤ 10^{4.5}\,\rm{K}$). Refer to Ramesh+(2023b) for more details. The "Interface" fields refer to the immediate shell of gas cells surrounding each cloud, i.e. the closest/transition region. The "Background" fields refer to the shell of gas cells surrounding the interface, i.e. a more distant/background region. See this website for more information.

(3) Zooniverse Cosmological Jellyfish

This catalog contains the complete results of the Citizen Science Project Cosmological Jellyfish on Zooniverse.org. In this project a large number of satellite galaxies were visually inspected to determine if they exhibit jellyfish-like morphological features, namely: extended tails of gas consistent with ram pressure stripping (RPS). The catalog includes the merged results of both Phase 1 (June 2021) and Phase 2 (August-December 2021). It contains two sets of scores: the 'raw' scores, based on the votes of at least 20 Inspectors (all scores receive equal weights), and the 'adjusted' scores, found after implementing the Inspector-weighting scheme described in Zinger et al. (2023).

There are three parts to this catalog: "flags and scores", "viewing angle comparison", and "branches". The first two come from Zinger et al. (2023), while the last comes from Rohr et al. (2023). Citation to these two papers is requested if you use this data. In addition, please include the following statement in the Acknowledgements: "This publication uses data generated via the \href{https://www.zooniverse.org/}{Zooniverse.org} platform, development of which is funded by generous support, including a Global Impact Award from Google, and by a grant from the Alfred P. Sloan Foundation."

Simulation and snapshot coverage:

• Available for the highest resolution runs of TNG100 and TNG50 currently: TNG100-1, TNG50-1.
• Only satellites were inspected.
• Satellite stellar mass: $\mathrm{M}_{\ast}>10^{8.3}\rm{M}_\odot$ (TNG50) or $\mathrm{M}_{\ast}>10^{9.5}\rm{M}_\odot$ (TNG100).
• Satellite gas fraction: $f_{\mathrm{gas}}>0.01$.
• Snapshots: all full snapshots up to $z=2$ (i.e. 33, 40, 50, 69, 67, 72, 78, 84, 91, 99), plus all snapshots up to $z=0.5$ for TNG50 only (i.e. 68-98 inclusive).

Flags and scores: in the catalog file there is a unique group labeled Snapshot_NNN for each snapshot number. It contains a number of datasets, each has a size equal to the number of inspected satellites at that snapshot:

Viewing angle comparison: in the catalog file there are two groups named Snapshot_NNN_ViewingAngle for snapshots 67 and 99 only. Here we chose a subset of objects (8762) for which we generated images from a preferred vantage point for jellyfish identification, such that the velocity vector is in the plane of the image. These images were presented for classification as separate objects so as not to introduce any bias to the classification process. The results of this experiment -- based on images from preferred instead of random directions -- are available in these groups. Both contain a number of datasets, each has a size equal to the number of inspected satellites at that snapshot:

Branches: the evolutionary tracks or branches of the galaxies inspected as part of the Zooniverse Cosmological Jellyfish project. While there are 53610 and 28094 inspected galaxy images for TNG50 and TNG100, many of these galaxies were inspected at different instances of cosmic time, resulting in 5023 and 9052 unique galaxy branches in TNG50 and TNG100, respectively. We store information about these galaxies for all snapshots, from snapshot 99 to 0 (in the same order as the merger trees). A given galaxy may not exist in the merger trees (for this catalog Sublink_gal) at every snapshot. In these cases, all fields except for SnapNum have a value -1. Each dataset has two dimensions: the first corresponds to the branch number (of size 5023 or 9052), while the second always has a size of 100, corresponding to the 100 snapshots:

(4) iMaNGA Mock MANGA IFU Datacubes

This catalog contains ~1,000 mock IFU datacubes recreating SDSS-IV/MaNGA-like observations. The galaxies are drawn from TNG50, selected to mimic the MaNGA-Primary sample selection. The IFU cubes are in .fits format. They are created using the MaStar stellar population models, plus Mappings-III models for young star-forming regions, using the SKIRT dust radiative transfer code.

The full description of the sample, mock procedure, and data products can be found in Nanni+ (2023), to which citation is requested if you use this data.

Simulation and snapshot coverage:

• Available for: TNG50-1.
• Sample: Selection of ~1,000 galaxies from TNG50-1 snapshots 88 through 98 (inclusive), chosen to match MaNGA properties.

The mock datacubes can be individually downloaded for each subhalo for which they are available. The API page for each (included) subhalo gives the download path, and the full list of available subhalos and the iMaNGA downloads are also available below.

• iMaNGA Datacubes - (~10 MB to ~3 GB each) list of the TNG50-1 galaxies available, each with a link to download the .fits mock IFU datacube.
• iMaNGA_catalog.hdf5 (1 MB) - catalog with additional information, including sample IDs (see table below).
• iMaNGA_catalog.fits (1 MB) - same catalog, also available in .fits format (see table below).
• iMaNGA_VAC.fits (11 GB) - value-added catalog (VAC) in .fits format (see second table below).

The contents of the iMaNGA VAC are:

Note: For HDU12, 10 is the maximum number of SSPs combined by Firefly in the fitting process.

(5) Synthetic Stellar Light Images with HSC-SSP Realism

This catalog contains synthetic near-UV/optical/near-IR images made using dust radiative transfer post-processing with SKIRT. The images include dust radiative transfer, and have been 'injected' into real Hyper Suprime-Cam Subaru Strategic Program (HSC-SSP, Aihara+ 2022) backgrounds.

The full description of the sample, mock procedure, and data products can be found in Bottrell+ (2023). Citation to that work, as well as to Eisert+ (in prep) is requested if you use this data.

Simulation and snapshot coverage:

• Available for: TNG50-1 and TNG100-1.
• Redshifts: $0.1 \leq z \leq 0.4$ (snapshots 72-91).
• Subhalos: total subhalo stellar masses $\mathrm{M}_{\star}\geq10^9 \mathrm{M}_{\odot}$ for TNG50-1; $\mathrm{M}_{\star}\geq10^{10} \mathrm{M}_{\odot}$ for TNG100-1, excluding SubhaloFlag=0 objects.
• Cameras: 4 viewing angles v0-v3 (tetrahedron; where v3 is the box z-axis).

The catalog includes images in noise-free, high-resolution format (idealized) and statistical injections into the HSC-SSP 3rd public data release wide layer (pdr3_wide). All images for a given galaxy / camera angle are contained in a FITS file with comprehensive headers detailing information about the images.

Idealized images:

• Filters: CLAUDS u-band; CFHT MegaCam r-band; HSC grizy-bands.
• FoV size: 50-500 kpc (see Bottrell+)
• Pixel scale: 100 pc

Survey-realistic images:

• Filters: HSC-SSP grizy-bands.
• FoV size: 50-500 kpc (see Bottrell+)
• Pixel scale: 0.167 arcseconds
• Survey extras: HSC-SSP variance maps, mask images, and PSF reconstructions

The FITS files can be downloaded separately for each subhalo for which they are available through the API. For example,

• TNG50-1 FITS Files (~10 MB to ~1 GB each) - list of the TNG50-1 galaxies available, each with a link to download the .fits mock.
• TNG100-1 FITS Files (~10 MB to ~1 GB each) - list of the TNG100-1 galaxies available, each with a link to download the .fits mock.
• /api/TNG100-1/snapshots/91/subhalos/388890/skirt/skirt_images_hsc_idealized_v0.fits downloads the idealized mock (v0) for the central subhalo of FoF #600 in TNG100-1 at $z=0.1$.
• /api/TNG50-1/snapshots/91/subhalos/503737/skirt/skirt_images_hsc_realistic_v3.fits downloads the hsc_realistic mock (v3) for the central subhalo of FoF #200 in TNG50-1 at $z=0.1$.

(6) Nearest Neighbors

A catalog of the nearest neighbours of galaxies. The catalog is constructed by measuring Euclidean distances between the galaxies in the simulation box with stellar masses above a minimum threshold. The minimum stellar mass threshold for the nearest neighbours ($M_{\ast \rm,min}^{\rm nn}$) depends on the resolution of the simulation run, with $M_{\ast \rm,min}^{\rm nn} = 10^7 \ \rm M_\odot$ for TNG50-1, $M_{\ast \rm,min}^{\rm nn} = 10^{8.2} \ \rm M_\odot$ for TNG100-1 and $M_{\ast \rm,min}^{\rm nn} = 10^{9.1} \ \rm M_\odot$ for TNG300-1. For each galaxy, the catalogue stores information about its ten nearest neighbours within 15 Mpc and the total number of neighbours within a fixed spherical aperture centred on the target galaxy. For further information, see Flores-Freitas+ (2024). Citation to that paper is requested if you use these data catalogs.

In the datasets, the placeholders {i} and {r} appears. Here, {i} refers to the index of the $i$-th neighbour, and can be any integer from 1 to 10. While {r} refers to the radius (in Mpc) of the spherical aperture used to count galaxies around a target galaxy, and can be the integers 1, 2 or 5.

Simulation and snapshot coverage:

• Available for: TNG50-1, TNG100-1, TNG300-1 (all snapshots/redshifts).
• Restricted to subhalos with stellar mass $M_\ast \geq 10^7 \ \rm M_\odot$ in TNG50-1, $M_\ast > 10^{8.2} \ \rm M_\odot$ in TNG100-1 and $M_\ast \geq 10^{9.1} \ \rm M_\odot$ in TNG300-1.

(7) TNG50-SKIRT Atlas

A database of images in 18 UV to near-infrared broadband filters. The images were generated with the SKIRT radiative transfer code and account for different stellar populations and absorption and scattering by interstellar dust. The images have a high spatial resolution (100 pc) and a wide field of view (160 kpc). Each galaxy is observed from 5 observing positions, spread on the unit sphere in optimal arrangement. In addition to the dust-obscured images, we also release dust-free images and physical parameter property maps with matching characteristics. For full details, see Baes et al. (2024). Additional images in the Euclid bands are also generated; they are presented in Kovacic et al. (in prep). Citations to these papers are appreciated if you use these images.

Simulation and snapshot coverage:

• Available for: TNG50-1 at z=0 only (the link lists all available files for download).
• Restricted to the 1154 subhalos with (total) stellar mass M^* > 10^9.8 solar masses.

There are two datasets:

• A simple text-file catalog containing the subhalo IDs, and a few basic properties, for the sample.
• The individual FITS files, available as one .tar.gz file per subhalo, available through the API (10MB to 2GB each).
• For example, /api/TNG50-1/snapshots/99/subhalos/503987/skirt/skirt_atlas.tar.gz downloads the images for the central subhalo of HaloID 150 in TNG50-1 at $z=0$.

The synthetic images are available in FITS format, one file per band. All FITS for a given subhalo are combined into one .tar.gz file for download. The dimensions of each image are 1600x1600 pixels. These images are 'idealized', i.e. they do not contain instrumental effects such as PSF smoothing or noise. This must be added by the user, if desired.

(8) ... coming soon?