Exploring urban surface effects for GOES-R aerosol retrieval

The GOES-R ABI algorithm for Aerosol over Land retrieval is based on the heritage MODIS approach where surface spectral ratios between Blue/SWIR and Red/SWIR bands are determined a-priori from Landsat based measurements or more recently from direct surface reflectance retrievals using AERONET radiometers. However, while the ABI bands are close to the MODIS Aerosol over Land bands, spectral response differences in both band center and width require alternate approaches. For example, ABI prelaunch algorithms rely on vegetation responses from the ASTER Spectral library convolved over the SRS of the ABI channels which are driven by vegetation species. For urban areas, these results are often not representative and can lead to extreme biases (i.e overestimates) as observed for MODIS aerosol retrievals. In this presentation, we use results from the Santa Barbara hyperspectral surface reflection measurement campaign as well as spectral ground reflection outputs from AERONET-based Surface Reflectance Validation Network (ASRVN) to explore urban spectral signatures for built environments and assess the errors that can be expected from the GOES-R ABI algorithm due to the urbanization.