EROSE: An algorithm searching for resolved Local Group satellites. The impact of combining different photometric bands on the detectability of Milky Way satellites in the LSST

Abstract

Systematic searches for ultra-faint Milky Way satellite galaxies typically combine local density estimators with colour-magnitude cuts, which are usually performed in 2D and have rarely, if ever, been extended to higher-dimensional spaces. These studies generally rely on $gri$ photometry, overlooking the shallower but metal-sensitive $u$ band. Entering the era of LSST, our study explores different $ugri$ combinations to determine which provides the best performance. We injected 36,000 dwarf galaxies into the simulated data from the LSST Data Challenge 2 (DC2) and attempted to recover them using EROSE, a new fast multi-band search algorithm designed to be easily applied to any survey and combination of photometric bands. We find that a multi-colour-magnitude space combining all three $gri$ photometric bands slightly outperforms, on average, the more commonly used colour-magnitude spaces combining only $gr$ or $ri$. The highest recovery fraction is achieved using a colour-magnitude space combining the $u$ and $g$ bands, recovering, on average, $64.1^{+6.3}_{-6.0}\%$ of the injected satellites. This improvement is particularly evident for the most extended systems, for which the additional colour information provided by the $u$ band improves our ability to isolate metal-poor member stars from contaminating foreground populations. We estimate that the LSST will discover $\sim 50$ new Milky Way dwarf galaxies, more than doubling the currently known number of satellites within its footprint. Combining the LSST photometric bands with a space-based star/galaxy separation could further improve these results, enabling the discovery of satellites at the low-mass end of the Milky Way's satellite galaxy luminosity function and providing valuable constraints on dark matter models.