Diagnosing the drivers of rain on snow events in Alaska using dynamical downscaling

Abstract:

Rain on snow (ROS) events are fairly rare in Alaska but have broad impacts ranging from economic losses to hazardous driving conditions to difficult caribou foraging due to ice formation on the snow. While rare, these events have recently increased in frequency in Alaska and may continue to increase under the projected warming climate. Dynamically downscaled data are now available for Alaska based on historical reanalysis for 1979-2013, while CMIP5 historical and future scenario downscaling are in progress. These new data offer a detailed, gridded product of rain and snowfall not previously possible in the spatially and temporally coarser reanalysis and GCM output currently available. Preliminary analysis shows that the dynamical downscaled data can identify extreme ROS events in Interior Alaska. The ROS events in the dynamically downscaled data are analyzed against observations and the ERA-Interim reanalysis data used to force the historical downscaling simulations. Additionally, the synoptic atmospheric circulations conditions that correspond to major ROS events in various regions of Alaska are identified with Self-Organizing Map (SOM) analysis. Such analysis is beneficial for operational forecasters with the National Weather Service and for diagnosing the mechanisms of change in future climate projections.