Hybrid Ensemble Variational Analysis of Polarimetric Radar Data to Improve Microphysical Parameterization and Short-term Weather Prediction

点击翻译按钮获取中文摘要

This project seeks to study the best possible ways to utilize polarimetric radar data (PRD) to improve understanding and prediction of severe weather. The United States’ national weather radar network was recently upgraded to dual-polarization capability, which provides detailed, 4D, real-time data about the observed precipitation particles, such as their shape, phase, and amount. This information is often poorly represented in numerical weather prediction (NWP) models, which can negatively impact their forecasts. However, the expected benefits of incorporating this observed polarimetric radar information into NWP models to improve their forecasts have not yet been realized. This project seeks to advance our understanding of, and improve the prediction of, severe weather and microphysical characterization in NWP models by exploring the application of advanced storm-scale data assimilation techniques to PRD. Such improvements will help realize the benefits of the existing upgrade to the radar network and provide more timely severe weather information to the public as storm-scale NWP models are increasingly incorporated into the warning decision process.The newly available PRD from the WSR-88D radar network are arguably the best source of data for storm-scale weather quantification and forecasts because PRD contain rich information about hydrometeor microphysics, including the size, shape, phase, and composition of precipitating particles, and can be used to characterize the microphysics and radar signatures of severe weather event precursors. Hydrometeor classification and the retrieval of hydrometeor particle size distributions from PRD are performed to better diagnose microphysical states and their evolution. Further, PRD can be directly assimilated into NWP models to improve model initialization and to produce more realistic analyses and forecasts. Specific goals of this work include: (1) development of accurate and efficient parameterized PRD forward operators for hydrometeors; (2) quantification of observation errors that include both measurement and forward operator errors; (3) observation-based retrievals of hydrometeor particle size distributions; (4) simulation of severe storms using convective-scale NWP models with advanced microphysics parameterization schemes under different environmental conditions, and their comparison with PRD for real cases; and (5) assimilation of PRD into NWP models using a hybrid ensemble variational data assimilation routine for optimal model initialization and better prediction of severe weather.This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.