Evaluation of satellite-based observations for capturing early winter snowmelt

Abstract:

Abstract

Over the past fifty years global climate change has altered various environmental processes. Due to global climate change early snowmelt is occurring more frequently throughout much of the world. The increasing frequency of these events is a relatively new phenomena and it is challenging the effectiveness of current water resource management and flood forecasting best practices. Early snowmelt events are caused by a brief period of unusually high air temperature, high humidity, or rain-on-snow. This research focuses on the detection of rain-on-snow events using remote sensing approaches to identify the frequency, extent and magnitude of these events. Early snow melt events, driven by rainfall with the presence of snow, are identified from The Flood Observatory archives. Passive microwave data from the AMSR-E and SSM/I instruments are compared with MODIS imagery and field observations to assess the microwave products’ reliability in capturing these events. Early melt detection algorithms that use passive microwave retrievals for northern latitude areas, primarily Alaska and Canada, were evaluated in the continental United States. These algorithms failed to capture mid latitude early snow melt events primarily due to climatological differences between northern and mid latitude areas. This research developed an alternative algorithm using the passive microwave signature that reflects the inherent characteristics of mid latitude rain-on-snow events. The two algorithms are compared for their relative value in detecting mid latitude rain-on-snow events. Performance is linked to climatological signatures of observed rain-on-snow events.