Applying a new integrated mass-flux adjustment filter in rapid update cycling of convective-scale data assimilation for the COSMO model (v5.07)
Yuefei Zeng, Alberto de Lózar, Tijana Janjić, Axel Seifert
Abstract
Abstract. A new integrated mass-flux adjustment filter is introduced, which uses the analyzed integrated mass-flux divergence field to correct the analyzed wind field. The filter has been examined in twin experiments with rapid update cycling using an idealized setup for convective-scale radar data assimilation. It is found that the new filter slightly reduces the accuracy of background and analysis states; however, it preserves the main structure of cold pools and primary mesocyclone properties of supercells. More importantly, it considerably diminishes spurious mass-flux divergence and the high surface pressure tendency, and it thus results in more dynamically balanced analysis states. For the ensuing 3 h forecasts, the experiment that employs the filter becomes more skillful after 1 h. These preliminary results show that the filter is a promising tool to alleviate the imbalance problem caused by data assimilation, especially for convective-scale applications.