Information content of BP/RP spectra in Gaia DR3

Gaia Data Release 3 has provided the astronomical community with the largest stellar spectroscopic survey to date (> 220 million sources). The low resolution (R similar to 50) blue photometer (BP) and red photometer (RP) spectra will allow for the estimation of stellar atmospheric parameters such as effective temperature, surface gravity, and metallicity. We create mock Gaia BP/RP spectra and use Fisher information matrices to probe the resolution limit of stellar parameter measurements using BP/RP spectra. The best-case scenario uncertainties that this analysis provides are then used to produce a mock-observed stellar population in order to evaluate the false positive rate (FPR) of identifying extremely metal-poor stars. We conclude that the community will be able to confidently identify metal-poor stars at magnitudes brighter than G = 16 using BP/RP spectra. At fainter magnitudes true detections will start to be overwhelmed by false positives. When adopting the commonly-used G < 14 limit for metal-poor star searches, we find a FPR for the low-metallicity regimes [Fe/H] < -2, -2.5, and -3 of just 14 per cent, 33 per cent, and 56 per cent respectively, offering the potential for significant improvements on previous targeting campaigns. Additionally, we explore the chemical sensitivity obtainable directly from BP/RP spectra for carbon and a-elements. We find an absolute carbon abundance uncertainty of sigma(A(C)) < 1 dex for carbon-enriched metal-poor (CEMP) stars, indicating the potential to identify a CEMP stellar population for follow-up confirmation with higher resolution spectroscopy. Finally, we find that large uncertainties in alpha-element abundance measurements using BP/RP spectra means that efficiently obtaining these abundances will be challenging.