Snow cover in the Tibetan Plateau is highly variable in space and time and plays a key role in ecological processes of this cold-desert ecosystem. Resolution of passive microwave data is too low for regional-scale estimates of snow cover on the Tibetan Plateau, requiring an alternate data source. Optically derived snow indices allow for more accurate quantification of snow cover using higher-resolution datasets subject to the constraint of cloud cover. This paper introduces a new optical snow index and assesses four optically derived MODIS snow indices using Landsat-based validation scenes: MODIS Snow-Covered Area and Grain Size (MODSCAG), Relative Multiple Endmember Spectral Mixture Analysis (RMESMA), Relative Spectral Mixture Analysis (RSMA) and the normalized-difference snow index (NDSI). Pearson correlation coefficients were positively correlated with the validation datasets for all four optical snow indices, suggesting each provides a good measure of total snow extent. At the 95% confidence level, linear least-squares regression showed that MODSCAG and RMESMA had accuracy comparable to validation scenes. Fusion of optical snow indices with passive microwave products, which provide snow depth and snow water equivalent, has the potential to contribute to hydrologic and energy-balance modeling in the Tibetan Plateau.

Dust deposition to snow can have regionally important climatic and hydrologic impacts resulting from direct reduction of surface albedo and indirectly from the initiation of snow albedo feedbacks. Modeling the radiative impacts of dust deposited in snow requires knowledge of the optical properties of both components. Here we present an inversion technique to retrieve the effective optical properties of dust deposited in mountain snow cover from measurements of hemispherical dust reflectance and particle size distributions using radiative transfer modeling. First, modeled reflectance is produced from single scattering properties modeled with Mie theory for a specified grain size distribution over a range of values for the imaginary part of the complex refractive index (k = 0.00001–0.1). Then, a multi-step look-up table process is employed to retrieve kλ and single scattering optical properties by matching measured to modeled reflectance across the shortwave and near infrared. The real part of the complex refractive index, n, for dust aerosols ranges between 1.5 and 1.6 and a sensitivity analysis shows the method is relatively insensitive to the choice of n within this range, 1.525 was used here. Using the values retrieved by this method to update dust optical properties in a snow + aerosol radiative transfer model reduces errors in springtime albedo modeling by 50–70%.