Description
Until now, galaxy properties and chemical evolution in the lower metallicity regime have remained relatively uncharted, with considerable scatter. For example, chemical abundances like N/O ratios exhibit a scatter of 0.1–0.3 dex at an oxygen abundance of 12+log(O/H) < 7.69 (~0.1 $Z_{\odot}$), likely driven by stochastic chemical enrichment. Additionally, recent JWST observations have revealed a potentially significant increase of ionising photons produced at higher redshift than previously assumed. This new finding highlights the need for more representative samples to improve statistical constraints in this regime.
Extremely metal-poor dwarf galaxies (XMPs) in the local Universe are unique archaeological sites, preserving near-pristine gas and often regarded as the “living fossils” of the earliest galaxies. These relics serve as ideal laboratories for studying the chemical evolution and enrichment of early galaxies, while also enhancing statistical constraints in the lower metallicity regime. Furthermore, they provide a robust probe of primordial element abundances, such as Helium, offering valuable constraints on the Standard Model. However, XMPs are exceedingly rare; despite extensive searches over the past two decades, only a few hundred have been identified among millions of galaxy samples.
This talk will highlight the use of deep learning pipelines for effective identifying and characterising these galaxies in existing surveys. In our pilot study, we confirmed the metal-poor nature of 45 XMPs using strong line diagnostics. We will present these new discoveries from our ongoing search, and discuss their significance in understanding chemical evolution, ionisation states, and probing primordial helium abundance.
