We present a method for identifying and characterizing ultra metal-poor (UMP) stars using combined GALEX near-ultraviolet (NUV) and Gaia broadband photometry. The aim of this work is to test whether the NUV − G and BP − RP color plane provides a metallicity-sensitive diagnostic at the lowest metallicities. We construct synthetic color-color grids by computing GALEX NUV and Gaia G, BP , and RP photometry from MARCS model flux distributions modified to higher-resolution using 1D-LTE synthetic spectra. Our grids span stellar parameters typical of FGK-type stars: 4500 K ≤ Teff ≤ 7000 K, 1.0 ≤ log(g) ≤ 5.0, and −5.00 ≤ Fe/H ≤ −1.50. We compare these grids to a compiled literature sample of ≈ 500 known metal-poor stars. Our initial results show that the synthetic grids reproduce the overall morphology of the observed distribution and preserve metallicity ordering, with the strongest agreement for dwarf stars and larger residuals among giants. Our method recovers 39% known stars with Fe/H ≤ −3.5 for which GALEX photometry exists, including 6/14 dwarfs and 5/14 giants. We get a combined enrichment factor of 5.70, with giants at 4.50 and dwarfs at 7.71. These preliminary results demonstrate metallicity sensitivity in the NUV-Gaia color-color plane and motivate calibration for UMP candidate selection in a full Gaia-GALEX crossmatch
Charvu et al. (2026) studied this question.