The New MISR Research Aerosol Retrieval Algorithm: A Multi-Angle, Multi-Spectral, Bounded-Variable Least Squares Retrieval of Aerosol Particle Properties over Both Land and Water

J.A. Limbacher, R.A. Kahn, and J. Lee

Atmosph. Meas. Tech., 2022 (in press)

Summary

Since well before the MISR instrument was launched on NASA’s Terra satellite, the MISR Research Aerosol retrieval algorithm (RA) was used to assess likely instrument sensitivity to particle microphysical properties and to explore options for squeezing as much information from the multi-angle, multi-spectral data as possible. This paper represents the latest iteration in a progression of advances pioneered with the MISR RA. The newest development takes advantage of ever-increasing computer speed and storage capabilities, as well as the availability of mature products from other satellite instruments, specifically the MODIS/MAIAC land surface retrieval data.

This new RA version incorporates aerosol retrievals over land and water using both retrieved and prescribed surfaces. When aerosol loading is high, performing a coupled surface-atmosphere retrieval is difficult, because the aerosol obscures the surface. In these circumstances, a prescribed surface yields better aerosol retrieval results – the MODIS/MAIAC surface reflectance models are adopted over land, and a fixed set of reflectance values are applied over water. Errors in these assumed surface representations have relatively little effect on the aerosol solutions under these circumstances, because the surface contribution to the top-of-atmosphere signal is small. Coversely, at low aerosol loading, the combined surface-atmosphere retrieval approach generally works well, whereas small errors in a prescribed surface can alias the aerosol retrieval significantly. This paper presents the algorithms associated with these approaches, and provides validation of the results against coincident AERONET surface-based sun photometer spectral optical depth measurements. The aerosol loading criteria used to select a prescribed-surface solution, a retrieved-surface solution, or some combination of the two, are also developed.