2025RAA....25c5001S

Accurate determinations of metallicity for large, complete stellar samples are essential for advancing various studies of the Milky Way. In this paper, we present a data-driven algorithm that leverages photometric data from the KiDS and the VIKING surveys to estimate stellar absolute magnitudes, effective temperatures, and metallicities. The algorithm is trained and validated using spectroscopic data from LAMOST, SEGUE, APOGEE, and GALAH, as well as a catalog of very metal-poor stars from the literature, and Gaia EDR3 data. This approach enables us to estimate metallicities, effective temperatures, and g-band absolute magnitudes for approximately 0.8 million stars in the KiDS data set. The photometric metallicity estimates exhibit an uncertainty of around 0.28 dex when compared to spectroscopic studies, within the metallicity range of -2 dex to 0.5 dex. The photometric effective temperature estimates have an uncertainty of around 149 K, while the uncertainty in the absolute magnitudes is approximately 0.36 mag. The metallicity estimates are reliable for values down to about -2 dex. This catalog represents a valuable resource for studying the structure and chemical properties of the Milky Way, offering an extensive data set for future investigations into Galactic formation and evolution.