ADEOS-II/GLI snow/ice products — Part III: Retrieved results

Abstract

Spatial and temporal distributions of snow grain size, mass fraction of snow impurity, and snow surface temperature around the northern polar region were retrieved together with snow and sea ice cover extents from radiance data observed by the Global Imager (GLI), an optical sensor onboard Japan's Advanced Earth Observing Satellite-II (ADEOS-II). The sensor acquired global radiance data from April 2 to October 24 in 2003. Two grain sizes, one for the shallow (0–20 cm) layer and the other for the top surface (0–2 cm) layer, were retrieved using channels with different light penetration depths. Snow and sea ice related products derived from the Advanced Microwave Scanning Radiometer (AMSR) and the Moderate Resolution Imaging Spectroradiometer (MODIS) were used for comparison. Retrieved seasonal variations of snow and sea ice cover extents were consistent with the MODIS and AMSR products. GLI derived snow surface temperatures were also correlated with the MODIS surface temperatures and exhibited latitudinal dependence over the northern hemisphere. Retrieved snow impurities were considered to be partly inaccurate over the Arctic sea ice in April and over inland areas of the continents possibly due to thin snow and/or dense vegetation. However, the snow cover around the northern Canadian Arctic tundra and Archipelago regions and the Greenland ice sheet in April to June was found to be very clean with a soot-equivalent impurity fraction of around 0.05 ppmw or less. Thus, these areas can be used as reference sites for monitoring pollution levels of the Arctic snow cover. Retrieved snow grain radii in the shallow layer seemed erroneous over thin snow covers, but indicated drastic spatial and seasonal evolution of the snow cover over the northern hemisphere with radius ranging from 50 μm to over 1000 μm synchronized with variations in surface temperature. The spatial variability of the top surface grain radius was partly different from that of the grain size in the shallow layer suggesting the possibility of detecting vertical inhomogeneity of the top 10 cm of the snow cover.