Examination of the impacts of vegetation on the correlation between snow water equivalent and passive microwave brightness temperature

• Published in: Remote sensing of environment 2017-05, Vol.193, p.244-256
• Main Authors: Cai, Shanshan, Li, Dongyue, Durand, Michael, Margulis, Steven A.
• Format: Article
• Language: English
• Subjects: Basins; Brightness; Brightness temperature; Correlation; Critical components; Density; Elevation; Equivalence; Forests; Hydrologic cycle; Hydrology; Masks; Microwaves; MODIS; Mountain regions; Mountainous areas; Passive microwave radiometry; Radiometry; Rainfall; Remote sensing; Remote sensing of snow; Runoff; Sensitivity; Sensitivity analysis; Sierra Nevada; Snow; Snow accumulation; Snow water equivalent; Terbium; Time series; Treeline; Vegetation; Vertical polarization; Water resources
• ISSN: 0034-4257; 1879-0704
• DOI: 10.1016/j.rse.2017.03.006

Snow accumulation, ablation, and runoff in mountainous areas are critical components of the hydrologic cycle, but are poorly known. Passive microwave (PM) measurements are sensitive to snow water equivalent (SWE), even in mountain regions, but vegetation masks the microwave signals and reduces this sensitivity. This study examines how the PM snow signal is affected by the forest density in fourteen basins in the Sierra Nevada, USA, and in a series of sixteen subsets of the Kern basin that have varied vegetation density. 36.5GHz vertical polarization brightness temperature (Tb) time series for each basin were produced from the spaceborne AMSR-E operational period (water year (WY) 2003 to WY2011). For each basin, the coefficient of determination (R2) between the annual minimum Tb and the concurrent SWE was calculated to evaluate the sensitivity of the PM to SWE. The relationship between the R2 values and the forest density was then analyzed to assess how vegetation affect the SWE information in the observed Tb. Mean forest coverage from MODIS was used to represent forest density. The R2 between the annual minimum Tb and concurrent SWE was >0.6 for three of the basins. Consistent with previous studies, WY2006 demonstrated anomalous Tb values for many basins, apparently due to anomalous warm winter rainfall. Excluding WY2006, R2 is significantly higher in all basins: eight of fourteen basins have R2 >0.6. For basins with average elevation >2500m, SWE correlates well with Tb. The R2 decreases monotonically with decreasing elevation. Basin elevation and forest cover are highly correlated in the Sierra; a basin elevation of 2500m generally coincides with forest cover of 20%. A total of 42% of Sierra Nevada has