The role of the Global Systems Division (GSD) of ESRL is to provide real-time numerical model support to complement the NAM, RUC, and GFS services provided by the National Weather Service. In particular, HMT provides a platform to test the quality and value of multiple high-resolution mesoscale models (i.e., various configurations of the WRF model) and ensemble postprocessing methods to generate unbiased quantitative precipitation forecasts (QPF) and well-calibrated probabilities of precipitation (PQPF) in excess of multiple thresholds.
This winter marks the second season of GSD participation in the HMT project. In the first season, also conducted in the ARB, the model forecasts and verifying precipitation observations from several IOPs were used to retrospectively test various combinations of physical parameterizations, intializations, and lateral boundary sources to optimize the ensemble makeup. This was also the process for computing the weights on the ensemble members which were used in real time for this winter's operations.
This talk focuses on ensemble modeling strategies for specific regional phenomena, using the HMT experience along with the results from other fields projects to guide the development of tailored numerical predictions systems.