Atmospheric and Forest Decoupling of Passive Microwave Brightness Temperature Observations Over Snow-Covered Terrain in North America

This study addresses two significant sources of uncertainty prevalent in snow water equivalent (SWE) retrievals derived from advanced microwave scanning radiometer (AMSR-E) passive microwave (PMW) brightness temperature (Tb) observations at 18.7 and 36.5 GHz. Namely, atmospheric and overlying forest effects are decoupled from the original AMSR-E PMW Tb observations using relatively simple, physically based radiative transfer models. Comparisons against independent Tb measurements collected during airborne PMWTb surveys highlight the effectiveness of the proposed AMSR-E atmospheric decoupling procedure. The atmospheric contribution to Tb ranges from 1 to 3 K depending on the frequency and polarization measured as well as meteorologic conditions at the time of AMSR-E overpass. It is further shown that forest decoupling should be conducted as a function of both land cover type and snow cover class. The exponential decay relationship between the forest structure parameter, namely, MODIS-derived leaf area index (LAI) and forest transmissivity, is fitted across snow-covered terrain in North America. The fitted exponential function can be utilized during forest decoupling activities for evergreen needle-leaved forest and woody savanna regions, but remains uncertain in other forest types due to a sparsity of snow-covered areas. By removing forest-related Tb contributions from the original AMSR-E observations, the results suggest that Tb spectral difference between 18.7 and 36.5 GHz, in general, increases across thinly vegetated to heavily vegetated regions, which can be beneficial when applied to traditional SWE retrieval algorithms.