Accounting for interannual variability in dust accelerated snowmelt in process-based hydrologic prediction, Rocky Mountains, USA

Abstract

Seasonal mountain snowmelt is an important contributor to surface water resources and groundwater recharge in the midlatitudes, making forecasting of snowmelt timing and duration critical for accurate hydrologic prediction. Net solar radiation, controlled primarily by snow albedo, is the main driver of snowmelt in most snow covered environments. Lowering of snow albedo from episodic dust deposition has been shown to be an important control on snowmelt patterns in the Rocky Mountains of the Western United States. Here, we compare and contrast trends in dust impacted albedo over the previous two decades with a focus on two regions: 1) the Colorado Rockies, headwaters of the Colorado River, which recieves dust from the southern Colorado Plateau and 2) the Wasatch Mountains (UT), headwaters of the Great Salt Lake, which recieves dust from the Great Basin. Results show that while snow darkening occurs every year, the magnitude of impact is spatially and temporally variable, and there are no emerging relationships that indicate when 'high-impact' dust years will occur. To account for spatial and interannual variability in dust impacted net solar radiation in hydrologic prediction we developed a spatially distributed process-based snowmelt model that incorporates near-real time snow albedo from remote sensing and incoming solar radiation from numerical weather prediction. The model improves simulated timing of snowmelt initiation and duration in all years, even those with lower dust impacts, demonstrating the importance of accurate snow albedo in snowmelt modeling.