Errors When Constraining Hot Blackbody Parameters with Optical Photometry

Measuring blackbody parameters for objects hotter than a few 10^4K with optical data alone is common in many astrophysical studies. However this process prone to large errors because at those temperatures the bands are mostly sampling Rayleigh-Jeans tail of spectrum. Here we quantify these by simulating different blackbodies, them various realistic measurement errors, and re-fitting blackbodies using two methods priors. We find that when only data, log-uniform priors perform better uniform Still, measured above ~35,000K can be wrong ~10,000K, lower limits obtained ~50,000K. Bolometric luminosities estimated from optical-only fits factors 3-5. When adding space-based ultraviolet shrink significantly. For such not available, provide plots tables distributions true result temperatures. It important take into account as systematic uncertainties fitting hot alone.