An Evaluation of Precipitation Forecasts during the Olympex field Campaign of 2015-16

Although substantial progress has been made, the accurate prediction of precipitation in complex terrain continues to be a challenge for numerical models.  The microphysical complexity, sharp gradients, and strong forcing in mountainous regions provides a potent testbed of current microphysical parameterization schemes.  This study evaluates the accuracy of WRF and GFS precipitation forecasts during the OLYMPEX field campaign of 2015-16.   Model precipitation is analyzed both spatially (e.g., windward versus leeward) and temporally (hourly through monthly) to determine systematic biases.   The WRF model will evaluated at a variety of resolutions (36, 12, 4, and 1.3 km grid spacing), while the GFS evaluaton is limited to its native resolution (13 km).  Initial results indicate significant overprediction by the GFS over the windward slopes and crest levels.    In contrast, the higher-resolution WRF simulations are more skillful, but do possess a tendency for underprediction along the windward slopes of coastal terrain. Nonphysical behavior of GFS precipitation fields is noted, particularly over terrain peaks and over the nearshore waters of the Pacific Ocean.  Finally, precipitation forecasts in WRF using several microphysical schemes are evaluated across multiple resolutions for the November 2015-April 2016 time period.