DESCQA (v1): LSST DESC Quality Assurance for Galaxy Catalogs

Latest Run |  All Runs |  GitHub

Please wait while the content is being generated.

This is DESCQA v1. You can also visit the latest version, DESCQA v2.

The DESCQA framework executes validation tests on mock galaxy catalogs. These tests and catalogs are contributed by LSST DESC collaborators. See the DESCQA paper for more information. Full details about the catalogs and tests, and how to contribute, are available here (collaborators only). The source code of DESCQA is hosted in this GitHub repo.

Below is a summary matrix of Run 2017-11-27. To view the results for an individual catalog, click one of the catalog names at the top of the table. To view the results for an individual test, click one of the test names along the left of the table.

You can also view the latest full run and a list of all runs.


2017-11-27

full run
Run initiated by yymao at 2017/11/27 16:12:34 PT
This run took 10.8 minute(s).
Versions used: GCRCatalogs: 1.0.0 | DESCQA: 1.0.0
[ Test prefix: CLEAR | Color | HMF | SMF | SMHM | WpRp ]
[ Catalog prefix: CLEAR | CAM | Galacticus | MBII | SAG | SHAM | iHOD ]
  SAG
Color_SDSS NOT QUITE
0.0703
HMF_Tinker PASSED
0.567
SMF_LiWhite NOT QUITE
1
SMF_MBII NOT QUITE
1
SMHM_MBII NOT QUITE
1
WpRp_MBII PASSED
0.762
WpRp_SDSS NOT QUITE
0.987
ValidationDescription
Color_SDSSFor each of the mock catalogs, we calculate the distributions of <i>M_u-M_g</i>, <i>M_g-M_r</i>, <i>M_r-M_i</i> and <i>M_i-M_z</i> colors, where the magnitudes are k-corrected absolute magnitudes, and compare with SDSS colors. The SDSS dataset includes <i>ugriz</i> photometry and spectroscopic redshifts from the SDSS main galaxy sample (Gunn98, York2000). SDSS galaxies in the redshift range of 0.06<z<0.09 are used for this comparison.
HMF_TinkerThe mass distribution of halos is one of the essential components of precision cosmology, and occupies a central place in the paradigm of structure formation. There are two common ways to define halos in a simulation. One is based on identifying overdense regions above a certain threshold. The other method, the FOF algorithm, is based on finding neighbors of particles and neighbors of neighbors as defined by a given separation distance. In DESCQA, we calculate the halo mass function from each catalog, and compare it against some well-established analytic fits in the literature. We assume Poisson error bars. We use the Bhattacharya et al. 2001 fit for the FOF halos, and Tinker et al. 2008 fit for the case of SO halos.
SMF_LiWhiteWe calculate the stellar-mass density as a function of the total stellar mass for each galaxy. Stellar masses are defined as the mass locked up in long-lived stars and stellar remnants (the most common definition). For the SAM models, the total stellar mass is the sum of the disk and spheroid components. The densities are derived from the number counts of galaxies in each stellar mass bin, divided by the simulation volume. These densities are compared with the data from Li and White 2009.
SMF_MBIIWe calculate the stellar-mass density as a function of the total stellar mass for each galaxy. Stellar masses are defined as the mass locked up in long-lived stars and stellar remnants (the most common definition). For the SAM models, the total stellar mass is the sum of the disk and spheroid components. The densities are derived from the number counts of galaxies in each stellar mass bin, divided by the simulation volume. These densities are compared with the data from the MassiveBlackII simulation.
SMHM_MBIIMean stellar mass as a function of halo mass for host halos.
WpRp_MBIIFor each of the mock catalogs, we calculate the projected two-point correlation function, w_p(r_p), in the thin-plane approximation. We use the catalog at one single epoch and then add redshift space distortion along one spatial axis (z-axis). We then calculate the projected pair counts, with a projection depth of 80 Mpc/h. We assume periodic boundary conditions for all three spatial axes. We estimate the sample variance of w_p(r_p) using the jackknife technique.
WpRp_SDSSFor each of the mock catalogs, we calculate the projected two-point correlation function, w<sub>p</sub>(r<sub>p</sub>), in the thin-plane approximation. We use the catalog at one single epoch and then add redshift space distortion along one spatial axis (z-axis). We then calculate the projected pair counts, with a projection depth of 80 Mpc/h. We assume periodic boundary conditions for all three spatial axes. We estimate the sample variance of w<sub>p</sub>(r<sub>p</sub>) using the jackknife technique.
CatalogDescription
CAM_LiWhiteThis catalog is based on the dark matter only (DMO) version of the MB-2 simulation, and has been tuned to reproduce the Li & White stellar mass function. The galaxy catalog was created using the conditional abundance matching technique described in Hearin et al. 2014.
CAM_MBIIThis catalog is based on the dark matter only (DMO) version of the MB-2 simulation, and has been tuned to reproduce the Li & White stellar mass function. The galaxy catalog was created using the conditional abundance matching technique described in Hearin et al. 2014.
GalacticusThis catalog is based on the dark matter only (DMO) version of the MB-2 simulation, and employs a semi-analytic model Galacticus. Galacticus models the baryonic physics of galaxy growth within an evolving, merging hierarchy of dark matter halos. Baryonic processes (including gas cooling and inflow, star formation, stellar and AGN feedback, and galaxy merging) are described by a collection of differential equations which are integrated to a specified tolerance along each branch of the merger tree, plus instantaneous transformations associated with merger events. The result is a catalog of galaxies at all redshifts including both physical properties (stellar masses, sizes, morphologies) and observational properties (e.g. luminosities in any specified bandpass. The parameters of the model are constrained through either particle swarm optimization or MCMC techniques to match a wide variety of data on the galaxy population.
MBIIThe hydrodynamical counterpart of the MB-2 simulation.
SAGThis catalog is based on the dark matter only (DMO) version of the MB-2 simulation, and employs a semi-analytic approach to galaxy formation. This particular approach is based on the model developed by Springel et al. 2001}, which, as is usual with semi-analytic models, combines merger trees extracted from a dark matter cosmological simulation with a set of coupled differential equations for the baryonic processes taking place within these merger trees as time evolves.
SHAM_LiWhiteThis catalog is based on the dark matter only (DMO) version of the MB-2 simulation, and has been tuned to reproduce the Li & White stellar mass function. The abundance matching technique, also known as subhalo abundance matching (SHAM), is a generic scheme to connect one galaxy property (e.g., stellar mass or luminosity) with one halo property (e.g., halo mass) by assuming an approximately monotonic relation between these two properties. The two properties are matched at the same cumulative number density, and the resulting galaxy catalog, by explicit construction, preserves the input stellar mass (or luminosity) function.
SHAM_MBIIThis catalog is similar to SHAM Li White but has been tuned to the stellar mass function measured from a hydro-simulation, MassiveBlackII.
iHODThis catalog is based on the dark matter only (DMO) version of the MB-2 simulation, and has been tuned to reproduce the spatial clustering and the galaxy-galaxy lensing observed in SDSS. The iHOD model aims to provide a probabilistic mapping between halos and galaxies, assuming that the enormous diversity in the individual galaxy assembly histories inside similar halos would reduce to a stochastic scatter about the mean galaxy-to-halo connection by virtue of the central limit theorem. Building on the global HOD parameterization of Zu and Mandelbaum (2015, 2016, 2017), the iHOD formalism was developed to solve the mapping between galaxy properties (stellar mass and color) and halo mass, using the spatial clustering and the galaxy-galaxy lensing of galaxies in SDSS. Compared to the traditional HOD methods, iHOD can include ~80% more galaxies while taking into account the stellar mass incompleteness of galaxy samples in a self-consistent fashion.