Value Added Catalog of physical properties of more than 1.3 million galaxies from the DESI Survey

TL;DR

The paper delivers a comprehensive value-added catalog of physical properties for over 1.3 million DESI galaxies, derived via energy-balance SED fitting with CIGALE and explicit AGN templates. It demonstrates how including optical grz and MIR WISE photometry improves stellar mass and AGN decomposition, and it analyzes systematic dependencies on IMF, SSP, SFH, metallicity, and dust/AGN models. Cross-validation with COSMOS, DEVILS, and SDSS-based catalogs confirms generally good agreement for stellar masses and SFRs while highlighting caveats tied to photometric depth and AGN modeling. The VAC enables large-scale studies of galaxy demographics, the main sequence, and AGN host-galaxy connections across $0 < z \lesssim 6$, with practical quality flags to tailor samples to specific science goals. The work lays a foundation for DESI DR1 and future data releases, supporting robust statistical analyses of galaxy evolution and AGN activity.

Abstract

Aims. We present an extensive catalog of the physical properties of more than a million galaxies within the Dark Energy Spectroscopic Instrument (DESI), one of the largest spectroscopic surveys to date. Spanning over a full variety of target types, including emission line galaxies and luminous red galaxies as well as quasars, our survey encompasses an unprecedented range of spectroscopic redshifts, stretching from 0 to 6. Methods. The physical properties, such as stellar masses and star formation rates, are derived via the CIGALE spectral energy distribution (SED) fitting code accounting for the contribution coming from active galactic nuclei (AGN). Based on the modeling of the optical-mid-infrared (grz complemented by WISE photometry) SEDs, we study galaxy properties with respect to their location on the main sequence. Results. We revise the dependence of stellar mass estimates on model choices and availability of the WISE photometry. The WISE information is mandatory to minimize the misclassification of star-forming galaxies as AGN. The lack of WISE bands in SED fits leads to elevated AGN fractions for 68% of star-forming galaxies identified using emission line diagnostic diagram but does not significantly affect their stellar mass nor star formation estimates.

Figures (20)

• Figure 1: Redshift distribution of the DESI EDR sources with reliable redshift and photometry estimates (see Sect. \ref{['sec:data']}). The catalog includes all galaxies and quasars observed within the DESI target classes: BGS, LRG, ELG, QSO, SCND, and MWS. The redshift distribution of MWS targets is scaled up by a factor of 10. Some of the sources are targeted by multiple target classes. The numbers shown in the figure also contain duplicates, so this total is higher than the number of unique targets.
• Figure 2: Distribution of the $\rm \chi^2_r$ for the DESI EDR SED fits. The mean and standard deviation of $\rm \chi^2_r$ are reported in the legend. The dashed line corresponds to the mean $\rm \chi^2_r$. At least $\sim 88\%$ of the sample is characterized by robust fits (defined as $\rm \chi^2_r < 5$):
• Figure 3: Distribution of the SED-derived properties: rest-frame g-r color (top panel), stellar mass ($\rm log(M_{star}/M_{\odot})$; middle panel), and star formation rate (SFR; bottom panel) of 1,286,124 of DESI EDR galaxies. The MWS target class is scaled up by a factor of ten.
• Figure 4: UVJ diagram for BGS galaxies observed at redshift $\rm 0<z<0.6$ color-coded according to their sSFR estimates. Red and blue galaxies are selected by the cut defined by Whitaker2012 shown in black.
• Figure 5: Stellar mass (left) and SFR (right) distributions for red (in red) and blue (in blue) BGS galaxies observed at redshift $\rm 0<z<0.6$ with means and standard deviations reported in the legend. Red and blue galaxies are selected by the cut defined by Whitaker2012 in the UVJ diagram (see Fig. \ref{['fig:UVJ']}).