5A.5 Recent developments in mesoscale data assimilation with WRF/DART

Snyder, Chris, National Center for Atmospheric Research

WRF/DART provides an ensemble Kalman filter for WRF, implemented via the Data Assimilation Research Testbed (DART). Because the ensemble Kalman filter requires minimal assumptions about the forecast covariances, WRF/DART is well suited to the mesoscale, where balance conditions are uncertain and flow dependent. Recent results for assimilation of surface observations and from extended periods of real-time, cycling data assimilation will be described. Because forecast models exhibit numerous deficiencies at the mesoscale, accounting for model error in the ensemble forecasts is a crucial research issue and a comparison of stochastic-backscatter and multi-physics approaches will be outlined