Using MODIS snow cover area and SSMI derived snow water equivalent to assess the value of using snow depletion curves predict river discharge and flood potential over the Red River Basin.

Snow depletion curves relate satellite-derived snow cover area (SCA) to snow water equivalent (SWE) to represent sub-grid SWE and snowmelt processes for hydrological modeling purposes. These curves are advantageous as satellite derived SCA provides a well-validated and high-resolution spatial representation of snow extent which can be used to fill in the sparse in-situ SWE observations. However, the reliability of these curves is based on an assumption that areal snow cover is fairly consistent from year-to-year and is largely determined by the physical attributes of the region. Critically, these depletion curves cannot capture anomalous patterns of accumulation and melt, nor are they designed to handle situations when new snow occurs on partially depleted surfaces. The Red River Basin (RRB) region, located in the northern US Great Plains, may or may not be well suited for depletion curve input to runoff estimation due to wind redistribution and storm patterns. We investigate the suitability of SWE depletion curves in the RRB by deriving SWE depletion curves from MODIS SCA data and the SSMI passive-microwave derived SWE. The results are evaluated against SNODAS modeled SWE data and basin-aggregated discharge observations. We further assess the benefits of leveraging the high-resolution MODIS SCA dataset as a means of disaggregating the coarser resolution passive microwave SWE dataset to better characterize the spatial distribution of SWE in the RRB.