Seminário: Recent Progresses with Convective-Scale Data Assimilation and Ensemble Forecasting at the Center for Analysis and Prediction of Storms

Data: 
02/03/2015 - 14:00
Local: 
Auditório IAG (Rua do Matão, 1226, Cidade Universitária)


O Departamento de Ciências Atmosféricas realiza seminário extra no dia 2 de março (segunda-feira), às 14:00, no Auditório IAG (bloco G). O palestrante é o Prof. Ming Xue (Universidade de Oklahoma, USA), convidado do LabHidro, coordenado pelo Prof. Dr. Augusto José Pereira Filho.

Recent Progresses with Convective-Scale Data Assimilation and Ensemble Forecasting at the Center for Analysis and Prediction of Storms
 
Ming Xue
Center for Analysis and Prediction of Storms and School of Meteorology
University of Oklahoma, Norman OK 73072, USA
 
The assimilation of high-resolution observations is critical for improving numerical prediction of severe weather while the inherent uncertainties with the initial conditions and numerical models demand probabilistic forecasting information that can be derived from ensemble forecasts. Among the observations, radar is the only platform that has the spatial and temporal resolutions needed to capture internal structures of convective storms. 
 
The Center for Analysis and Prediction of Storms (CAPS) at the University of Oklahoma was established over 20 years ago to tackle the grand challenge problem of numerical prediction of thunderstorms. Over the years, CAPS has been developing and applyingcutting-edge methods and tools for assimilating Doppler weather radar data, and for the prediction of convective weather. This presentation will try to highlight research progresses at CAPS for recent years, including data assimilation algorithms and systemsbased on 3DVAR/cloud analysis, ensemble Kalman filters, 3D and 4D ensemble-variational (EnVar) hybrid methods. Example results applying these algorithms to a regional operational model, and to the analysis and prediction of mesoscale convective systems, tornadic supercell storms, tropical cyclones, and to realtime continental-scale severe weather prediction will be presented. The ability of such systems in dealing with multi-moment microphysics, attenuated reflectivity data, and in reproducing polarimetric radar variables and signatures will also be discussed. Results for severe weather and quantitative precipitation forecasting using a multi-model convection-resolving ensemble forecasting system that CAPS developed over the continental United States will also be presented.