Data Products

Halo catalogs and Merger Trees

The folder Rockstar/CAMELS-SAM/LH_???/Rockstar contains the CAMELS-SAM Rockstar halo catalogs for each simulation. The format of the catalogs is the default of the code, and we refer the user to the Rockstar documentation for further details. Due to storage limitations, we have kept just the out_??.list files and all ROCKSTAR configuration files that were used to run it. Additional files are stored on long-term magnetic tape; please contact the CAMELS team to discuss accessing them.

We also release the merger trees constructed through ConsistentTrees, tree_?_?_?.dat’ for each subvolume created in the simulation within ``Rockstar/CAMELS-SAM/LH_???/ConsistentTrees`. We refer the reader to the consistent trees documentation for details on the format of the files. The SC-SAM can be rerun with new parameters upon these trees.

The Santa Cruz SAM for galaxy formation

We use the Santa Cruz semi-analytic model (SAM) for galaxy formation (Somerville et al. 2008, 2015, 2021) to robustly populate our large N-body simulations with galaxies. SAMs can be thought of an alternative to full hydrodynamic simulations, which apply simplified empirical recipes for physical processes of galaxy formation and evolution within dark matter halo ‘merger trees’ (e.g. from ConsistentTrees our of ROCKSTAR). Like numerical simulations, SAMs include free parameters that are calibrated to galaxy observations (e.g. Yung et al. 2019 for the local universe). The SC-SAM pipeline used in this work is nearly identical to that applied to IllustrisTNG300 in Gabrielpillai et al. (2021); see their work for complete justification and verification for the fiducial set up.

A note: the SC-SAM that was run for CAMELS-SAM did not store the averaged SFR properties for the galaxies, due to our particular needs and priorities, so all those will be equal to zero.

Click to expand/collapse this table describing the SC-SAM *galaxy* outputs at each redshift

Field	Shape	Units	Description
GalpropBirthHaloID	N	–	ID of host halo from Haloprop
GalpropHaloIndex	N	–	Index of host halo in Haloprop (recommend usage: when loading whole subvolumes)
GalpropHaloIndex_Snapshot	N	–	Index of host halo in Haloprop (recommended usage: when loading snapshots)
GalpropMBH	N	10^9 M☉	Black hole mass
GalpropMH2	N	10^9 M☉	Molecular hydrogen cold gas mass
GalpropMHI	N	10^9 M☉	Neutral hydrogen cold gas mass
GalpropMHII	N	10^9 M☉	Singly ionized hydrogen cold gas mass
GalpropMaccdot	N	M☉ / yr	accretion rate onto black hole
GalpropMaccdot_radio	N	M☉ / yr	accretion rate onto black hole in radio mode
GalpropMbulge	N	10^9 M☉	Stellar mass of the bulge
GalpropMcold	N	10^9 M☉	Total cold gas mass
GalpropMvir	N	10^9 M☉	Total dark matter halo mass
GalpropMstar	N	10^9 M☉	Total stellar mass
GalpropMstar_merge	N	10^9 M☉	Stellar mass entering via mergers
GalpropMstrip	N	10^9 M☉	Stripped mass of DM sub-halo |
GalpropMu_merger	N	–	mass ratio of last merger (see S08)
GalpropOutflowRate_Mass	N	M☉ / yr	rate of outflowing gas from ISM
GalpropOutflowRate_Metal	N	M☉ / yr	rate of outflowing metals from ISM
GalpropPos	N, 3	cMpc	x, y, and z coordinates
GalpropRbulge	N	kpc	3D half-mass Radius of bulge
GalpropRdisk	N	kpc	3D half-mass Radius of disk
GalpropRedshift	N	–	redshift
GalpropRfric	N	Mpc	distance from halo center
GalpropRhalo	N	Mpc	halo virial radius
GalpropRootHaloID	N	–	ID of root halo at z = 0 from Haloprop
GalpropSatType	N	–	0 = central
GalpropSfr	N	M☉ / yr	instantaneous SFR
GalpropSfrave100myr	N	M☉ / yr	SFR averaged over 100 Myr
GalpropSfrave1gyr	N	M☉ / yr	SFR averaged over 1 Gyr
GalpropSfrave20myr	N	M☉ / yr	SFR averaged over 20 Myr
GalpropSigmaBulge	N	km / s	velocity dispersion of bulge
GalpropTmerger	N	Gyr	time since last merger
GalpropTmerger_major	N	Gyr	time since last major merger
GalpropTsat	N	Gyr	time since galaxy became a satellite in this halo
GalpropVdisk	N	km / s	rotation velocity of disk
GalpropVel	N, 3	km / s	x, y, and z components of velocity
GalpropZcold	N	Z☉ * M☉	metal mass in cold gas
GalpropZstar	N	Z☉ * M☉	metal mass in stars

Click to expand/collapse this table describing the SC-SAM *halo* outputs at each redshift

Field	Shape	Units	Description
HalopropC_nfw	N	–	NFW concentration parameter for DM halo
HalopropHaloID	N	–	Halo ID given by Consistent-Trees
HalopropIndex	N	–	Index of halo in the file (recommended usage: when loading whole subvolumes)
HalopropIndex_Snapshot	N	–	Index of halo in file (recommended usage: when loading snapshots)
HalopropMaccdot_metal	N	M☉ / Z☉ / yr	accretion rate of metals into the halo
HalopropMaccdot_pristine	N	M☉ / yr	accretion rate of pristine gas into the halo
HalopropMaccdot_radio	N	M☉ / yr	accretion rate onto the BH in radio mode
HalopropMaccdot_reaccreate	N	M☉ / yr	accretion rate of “recycled” gas
HalopropMaccdot_reaccreate_metal	N	M☉ / Z☉ / yr	accretion rate of “recycled” metals
HalopropMass_ejected	N	10^9 M☉	total gas mass in “ejected” reservoir
HalopropMcooldot	N	10^9 M☉ / yr	rate of gas cooling/accretion from halo into ISM
HalopropMdot_eject	N	M☉ / yr	rate of ejection of gas from halo
HalopropMdot_eject_metal	N	M☉ / Z☉ / yr	rate of ejection of metals from halo
HalopropMetal_eject	N	Z☉ / yr	total mass of metals in “ejected” reservoir
HalopropMhot	N	10^9 M☉	mass of hot (CGM) gas in halo
HalopropMstar_diffuse	N	10^9 M☉	mass of stars in a diffuse stellar halo (from disrupted satellites)
HalopropMvir	N	10^9 M☉	halo virial mass
HalopropRedshift	N	–	redshift
HalopropRockstarHaloID	N	–	Halo ID from the Rockstar run
HalopropRootHaloID	N	–	Halo ID of the root halo for this merger tree
HalopropSnapNum	N	–	snapshot file number
HalopropSpin	N	N	spin of DM halo
HalopropSubfindID_DMO	N	–	Subfind index in TNG DMO simulation of bijective match
HalopropSubfindID_FP	N	–	Subfind index in TNG FP simulation of bijective match
HalopropZhot	N	10^9 M☉ / Z☉	metal mass in hot halo (CGM)

Identifying each simulation’s parameters

Within each of these data product directories and within each simulation’s folder exists the file CosmoAstro_params.txt. This file includes the exact values of the cosmological and astrophysical parameters that created each CAMELS-SAM N-body simulation and halo/galaxy catalogs. There are six numbers listed, which are:

1. \(\Omega_{M}\)

   The true value of \(\Omega_{M}\) of the given simulation.
2. \(\sigma_{8}\)

   The true value of \(\sigma_{8}\) of the given simulation.
3. \(A_{\rm SN1} \times 1.7\) for SC-SAM

   The value given to the SC-SAM for the \(\epsilon_{\rm SN0}\) free parameter, which is equal to the prefactor \(A_{\rm SN1}\) times 1.7 (the best-fit fiducial value for this SC-SAM free amplitude parameter for local observations.) \(A_{\rm SN1}\) was generated log-uniformly between 0.25 and 4.0.
4. \(A_{\rm SN2} + 3.0\) for SC-SAM

   The value given to the SC-SAM for the \(\alpha_{\rm rh}\) free parameter, which is equal to the prefactor \(A_{\rm SN2}\) plus 3.0 (the best-fit fiducial value for this SC-SAM free power law slope for local observations.) \(A_{\rm SN2}\) was generated uniformly between -2.0 and 2.0.
5. \(A_{\rm AGN} \times 0.002\) for SC-SAM

   The value given to the SC-SAM for the \(\kappa_{\rm radio}\) free parameter, which is equal to the prefactor \(A_{\rm AGN}\) times 0.002 (the best-fit fiducial value for this SC-SAM free amplitude parameter for local observations.) \(A_{\rm AGN}\) was generated log-uniformly between 0.25 and 4.0.
6. 0.005

   This is an unused parameter in CAMELS-SAM for \(\epsilon_{\rm wind, QSO}\) from the SC-SAM. This was originally meant to be the analogous to \(A_{\rm AGN2}\) in CAMELS, but the iteration of the SC-SAM used in this work showed little to no response to varying this parameter beyond its fiducial value.