UCLA

Abstract: We used the reparameterized Near-Earth Asteroid Thermal Model to model observations of a curated set of over 4000 asteroids from the Wide-field Infrared Survey Explorer in two wavelength bands (W2–3 or W3–4) and compared the results to previous results from all four wavelength bands (W1–4). This comparison was done with the goal of elucidating unique aspects of modeling two-band observations so that any potential biases or shortcomings for planned two-band surveys (e.g., the NASA Near-Earth Object Surveyor Mission) can be anticipated and quantified. The W2–3 two-band fits usually yielded slightly smaller diameters than the four-band fits, with a median diameter difference of −10%, with the 5% and 95% quantiles of the distribution at −32% and −1.5%, respectively. We conducted similar comparisons for W3–4, in part because the longest wavelength bands are expected to provide the best two-band results. We found that the W3–4 two-band diameters are slightly larger than the four-band results, with a median diameter difference of 11% and the 5% and 95% quantiles of the distribution at −2.1% and 26%, respectively. The diameter uncertainty, obtained with bootstrap analysis, is larger by 30% and 35% (median values) for the W2–3 and W3–4 fits, respectively, than for the corresponding four-band fits. Using 23 high-quality stellar occultation diameters as a benchmark, we found that the median errors of W2–3 and W3–4 diameter estimates are −15% and +12%, respectively, whereas the median error of the four-band fits is 9.3%. Although the W2–3 and W3–4 diameters appear to have greater systematic errors and uncertainties than their four-band counterparts, two-band estimates remain useful because they improve upon diameter estimates obtained from visible photometry alone.