Photometric stellar masses for galaxies in DESI Legacy Imaging Surveys
In many areas of extragalactic astrophysics, we need to convert the luminosity of a galaxy into its stellar mass. In this work, we aim to find a simple and effective formula to estimate the stellar mass from the images of galaxies delivered by the currently popular DESI Legacy Imaging Surveys. This survey provides an unsurpassed combination of a deep imaging with an extensive sky coverage in up to four photometric bands. We calibrated the sought formula against a sample of local galaxies observed by the Spitzer Survey of Stellar Structure in Galaxies (S$^4$G) that was directly dedicated to measure the stellar masses. For the absolute magnitudes $M_g$ and $M_r$ of a galaxy in the Legacy Surveys $g$ and $r$ bands, we find that the stellar masses can be estimated as $0.673M_g - 1.108M_r + 0.996$ with the scatter of 25\%. Employing more complex functions does not improve the estimate appreciably, even after including the galaxy ellipticity, Sérsic index, or the magnitudes in different Legacy Surveys bands. Generally, measurements in $r$ band were the most helpful ones, while adding $z$-band measurements did not improve the mass estimate much. We provide a Python-based script \texttt{photomass\_ls.py} to automatically download images of any galaxy from the Legacy Surveys database, create image masks, generate GALFIT input files with well-assessed initial values, perform the GALFIT photometry, and calculate the stellar mass estimate. Additionally, we tuned another version of the formula to the magnitudes provided by the Siena Galaxy Atlas 2020 (SGA-2020) with a scatter of 29\%. For both\,--\,our default and SGA-2020 formula, we offer two alternatives derived from different calibrations of S$^4$G masses that were based on different methods and assumptions.