IN035-0001
Enhancement of snow and ice albedo performance in VIIRS global surface albedo products

Tuesday, 15 December 2020
Poster
Jingjing Peng, University of Maryland College Park, College Park, MD, United States, Yunyue Yu, NOAA, STAR, College Park, MD, United States, Aolin Jia, University of Maryland, College Park, College Park, MD, United States, Dongdong Wang, University of Maryland College Park, Department of Geographical Sciences, College Park, MD, United States and Liang Shunlin, University of Maryland College Park, Geographical Sciences, College Park, MD, United States
Abstract:
Snow, as well as ice, shows larger albedo than snow-free land surface, which is the basis of their feedback with albedo and causing further interaction with climate. The single-observation-dependence nature and the feasibility of the generic look-up table (LUT) facilitate the VIIRS albedo to capture temporal new snow. However, when apparent snow omission in input happens over the Greenland or Antarctic region, albedo would be underestimated due to using the bare soil LUT on permanent snow surface. Thus, the VIIRS snow albedo suffers from certain uncertainty sources including the input data accuracy and the climatology discontinuity, which has been improved in the recent mitigations.

The albedo climatology, which is the multiple-year albedo average for each pixel, is used for filtering VIIRS clear-sky retrieval results to get the gap-free surface albedo. In the new version we have reduced the temporal discontinuity in sea-ice and snow surface by value interpolation at polar night and smoothing the albedo time series. Antarctic sea-ice climatology is newly added from collection of multi-year sea-ice historical data, and calculate the yearly mean, standard deviation.

Validation results has proved the influence of LUT type selection on bright-surface albedo, as the output reveals lower-than-normal albedo value over Antarctic region when bare-soil LUT is used for the permanent snow-covered surface. The snow omission has influenced the completeness and accuracy of albedo retrieving due to the distinct difference between snow and non-snow surface anisotropy. An attempt is applying the snow mask from Interactive Multisensor Snow and Ice Mapping System (IMS), a recognized snow mask product derived from multiple satellite source. One advantage of IMS is the less impact from clouds. A gap free snow mask would assist VIIRS albedo algorithm in selecting appropriate LUT and climatology, so that the VIIRS algorithm can estimate snow albedo using the single observation with comparable accuracy as MODIS does using multi-day dataset. This provided the advantage of VIIRS albedo over fresh/new snow or temporal snow when insufficient snow observations accumulated for BRDF modeling.

Previous Abstract | Next Abstract >>