Timing of Snow Melt and Refreeze Events in the Northern United States (2003—Present) from Passive Microwave Satellite Observations

Information about snow melt and refreeze is important for estimating spring runoff timing, and for informing models of snowpack stratigraphy and metamorphism. Snowpack that contains liquid water emits more microwave radiation than if it were completely frozen. Previous studies have shown that brightness temperature (T_b) observations from space-based passive microwave radiometers are sensitive to snow phase changes. The difference between twice-daily T_b observations, called the diurnal amplitude variation (DAV), has been used to identify snow melt and refreeze events, where diurnal phase changes due to nighttime freezing and daytime melting result in high DAV values. Here, the enhanced DAV method of Tuttle and Jacobs (2019, Wat. Resour. Res.) is applied to identify melt and refreeze events across the contiguous United States and southern Canada from 2003 to present, using T_b from Advanced Microwave Scanning Radiometer (AMSR) satellite instruments, as well as near-surface air temperature (T_a). Individual melt and refreeze events are detected as large deviations from the approximately linear relationship between 12-hourly T_b change (ΔT_b) and Ta change (ΔTa) for frozen snow, using clustering techniques. The spatiotemporal properties of the identified melt and refreeze events within the study region, including frequency and timing, are examined. Some limitations of the method are also highlighted, including the need for observations from frozen snowpack and the challenge of detecting events beneath vegetation canopy.