Errors When Constraining Hot Blackbody Parameters with Optical Photometry

Abstract

Abstract Measuring blackbody parameters for objects hotter than a few 104 K with optical data alone is common in many astrophysical studies. However this process is prone to large errors because at those temperatures the optical bands are mostly sampling the Rayleigh–Jeans tail of the spectrum. Here we quantify these errors by simulating different blackbodies, sampling them in various bands with realistic measurement errors, and refitting them to blackbodies using two different methods and two different priors. We find that when using only optical data, log-uniform priors perform better than uniform priors. Still, measured temperatures of blackbodies above ∼35,000 K can be wrong by ∼10,000 K, and only lower limits can be obtained for temperatures of blackbodies hotter than ∼50,000 K. Bolometric luminosities estimated from optical-only blackbody fits can be wrong by factors of 3–5. When adding space-based ultraviolet data, these errors shrink significantly. For when such data are not available, we provide plots and tables of the distributions of true temperatures that can result in various measured temperatures. It is important to take these distributions into account as systematic uncertainties when fitting hot blackbodies with optical data alone.