UCLA

Abstract We analyzed 82,548 carefully curated observations of 4420 asteroids with Wide-field Infrared Survey Explorer (WISE) four-band data to produce estimates of diameters and infrared emissivities. We also used these diameter values in conjunction with absolute visual magnitudes to infer estimates of visible-band geometric albedos. We provide solutions to 131 asteroids not analyzed by the NEOWISE team and to 1778 asteroids not analyzed with four-band data by the NEOWISE team. Our process differs from the NEOWISE analysis in that it uses an accurate solar flux, integrates the flux with actual bandpass responses, obeys Kirchhoff’s law, and does not force emissivity values in all four bands to an arbitrary value of 0.9. We used a regularized model-fitting algorithm that yields improved fits to the data. Our results more closely match stellar-occultation diameter estimates than the NEOWISE results by a factor of ∼2. Using 24 high-quality stellar-occultation results as a benchmark, we found that the median error of four-infrared-band diameter estimates in a carefully curated data set is 9.3%. Our results also suggest the presence of a size-dependent bias in the NEOWISE diameter estimates, which may pollute estimates of asteroid size distributions and slightly inflate impact-hazard risk calculations. For more than 90% of asteroids in this sample, the primary source of error on the albedo estimate is the error on absolute visual magnitude.