0 Assigned ID: 7454 ----- Attempting Submission 1 (Thu Apr 16 16:26:24 GMT 2026) ----- Metal-poor massive stars in early galaxies are currently thought to be significant drivers of cosmic reionization based on early results from JWST. However, interpretations of observations at high-redshift are based solely on stellar models developed for higher-metallicity stars that have been extrapolated to the very metal-poor regime. Since the winds and ionizing fluxes of O- and early B-type stars can change dramatically at low metallicity, the existing spectral templates used to predict the UV and ionizing spectra of young, high-redshift galaxies may be substantially biased. We propose to analyze archival HST/COS FUV spectroscopy of 27 metal-poor massive stars in 5 extremely metal-poor dwarf galaxies to produce new ionizing spectral templates for stellar populations at Z < 10% Zsun. We will leverage the ULLYSES spectral library and data processing tools to provide a thorough analysis across spectral types and luminosities on the upper HR diagram. We will fit detailed stellar atmosphere models to the FUV spectroscopy, producing the first data-driven spectral templates for very metal-poor massive stellar populations. These will inform the development of a grid of predicted spectral templates, which will be distributed to the community. Our results will provide the first empirically-constrained ionizing spectral templates for massive stars at very low metallicity, which will inform the next generation of stellar population synthesis models. Erba is an expert in the observation and interpretation of massive star UV and optical spectroscopy (e.g., Erba et al. 2021a 2021b, 2024), particularly with respect to the winds, atmospheres, and magnetospheres of O- and early B-type stars. She has a strong background in the reduction and analysis of HST/COS data, and has collected ground-based optical data that will be used to provide fundamental stellar properties for the targets in this program. Erba has previously worked with the UV spectra of metal-poor massive stars, and has developed stellar atmosphere models to assess the properties of massive star winds. Telford is an expert in extremely metal-poor massive stars and has experience analyzing their FUV+optical spectra with the PoWR atmosphere code (Telford+2021, 2024). She is also the PI of multiple HST/COS programs (GO-16767, 16920, 17491). Chisholm is a co-PI of CLASSY and PI of multiple HST programs targeting FUV observations of metal-poor O stars (COS, STIS). Chisholm has an extensive background in stellar population synthesis modeling at low metallicity (e.g., Chisholm+2019) and in observations of feedback-driven outflows (e.g., Chisholm+2017). Hawcroft has deep expertise in using atmosphere modeling to interpret observations of O-stars, and particularly in inferring stellar wind properties from observed FUV wind profiles at low metallicity (Hawcroft+ 2024). Leitherer is a world leader in stellar population synthesis modeling and the creator of Starburst99 (Leitherer+ 1999). He is also the PI of a HST archival program to incorporate state-of-the-art stellar models into Starburst99, drawing on the rich ULLYSES dataset in the 20-50% SMC and LMC (AR-16623). McQuinn is an expert in the resolved stellar populations and chemical evolution of nearby, metal-poor galaxies (e.g., McQuinn+2015) and has extensive experience using DOLPHOT (Dolphin+2000) to measure PSF photometry of resolved stars in nearby galaxies. Sander is an expert in stellar atmosphere development and co-developer of the Potsdam Wolf-Rayet (PoWR) atmosphere code that will be used to analyze the data observed by this proposal. Ramachandran has in-depth expertise in quantitative spectroscopy of hot, massive stars, in particular to model low-metallicity OB stars in sub-solar metallicity environments (Ramachandran+2018, 2019, 2021). Our team is well-positioned to deliver a high-quality and timely analysis of the proposed observations. Stellar Physics and Stellar Types Stellar Populations and the Interstellar Medium 0 false true false false false false false SUPPORTED