Optimizing Strategies for an Observation-Nudging-Based Four-Dimensional Data Assimilation Forecast Approach with WRF-ARW

The Weather Research and Forecasting model is an open-source numerical weather prediction model with numerous features, such as the four-dimensional data assimilation system, an option to nudge forecasts to observations in order to improve forecast quality and performance. This process can act as a dynamical initialization prior to the forecast period to reduce errors that result from interpolation across the grid. For data void regions, assimilating observations from a temporary network could vastly improve forecasts in the region. However, users must specify how much influence each observation will get in the model and over how large an area each observation will influence. These weighting factors will have a great effect on the forecast quality. This research attempts to determine the optimal setting of the radius of influence for upper air and surface observations within the assimilation system. A case study over Yuma, AZ, is examined in which a low-pressure system produced dynamic and orographic precipitation across the region. Observations from the U.S. Air Force in Yuma will be assimilated into Weather Research and Forecasting for the 06 coordinated universal time forecast for an entire 24-h forecast cycle. Results suggest upper air observations have a greater impact on forecast accuracy given a large radius of influence.