Interannual Variability of Dust-Mass Loading and Composition of Dust Deposited on Snow Cover in the San Juan Mountains, CO, USA: Insights into Effects on Snow Melt

Dust deposited on snow cover (DOS) in the American West can enhance snow-melt rates and advance the timing of melting, which together can result in earlier-than-normal runoff and overall smaller late-season water supplies. Understanding DOS properties and how they affect the absorption of solar radiation can lead to improved snow-melt models by accounting for important dust components. Here, we report on the interannual variability of DOS-mass loading, particle size, organic matter, and iron mineralogy, and their correspondences to laboratory-measured reflectance of samples from the Swamp Angel Study Plot in the San Juan Mountains, Colorado, USA. Samples were collected near the end of spring from water year (WY) 2011-2016, when dust layers deposited throughout the year had merged into one layer at the snow surface. Dust-mass loading on snow during WY2011-WY2016 ranged from 2-64 g/m², mostly as particles with median sizes of 13-33 micrometers. Average reflectance values of DOS varied little across total (0.4 to 2.50 mm) and visible (0.4 to 0.7 mm) wavelengths at 0.30-0.45 and 0.19-0.27, respectively. Reflectance values lacked correspondence to particle size. Total reflectance values inversely corresponded to concentrations of (1) organic matter content (4-20 weight %; r² = 0.71) that included forms of black carbon and locally derived material such as pollen, and (2) magnetite (0.05 to 0.13 weight %; r² = 0.62). Magnetite may be a surrogate for related dark, light-absorbing minerals. Concentrations of crystalline ferric oxide minerals (hematite+goethite) based on magnetic properties at room-temperature did not show inverse association to visible reflectance values. These ferric oxide measures, however, did not account for the amounts of nano-sized ferric oxides known to exist in these samples. Quantification of such nano-sized particles is required to evaluate their possible effects on visible reflectance. Nonetheless, our results emphasize that reflectance values of year-end DOS layers at this site do not appear to be highly sensitive to variations in some measured DOS properties. These preliminary results cannot be broadly applied to other DOS sites in the American West on the basis of previous and ongoing studies.