The Influence Of Microphysical Scheme And The Parameters On EnSHF Data Assimilation

Ensemble square root filter is a kind of deterministic data assimilation method. It has the advantage of ensemble methods, such as that the error convariances are flow dependent and it is combined with ensemble forecast. At the same time, it has the advantage of traditional ensemble kalman filter methods, such as eliminating sample errors introduced by adding perturbations to the observations and improving the efficiency of the assimilation. Because of these advantages, EnSRF has become the hotspot of the data assimilation in the field of research and application.The study is based on the WRF-EnSRF assimilation system, combining microphysical parameterization scheme and perturbating parametesr for the error introduced by the process of constructing ensemble members. The improved system considers the influence of microphysical parameterization scheme and its parameter on the assimilation performance, and use the improved scheme to a series of comparison tests of assimilating simulated radar data and real radar data. The experiment of simulated radar data examines the influence of the new scheme on the assimilation system and verifies the feasibility of the scheme applied to assimilation system; the experiment of real radar data tests the performance of the improved scheme in the presence of many actual complex factoes, including observation errors and model errors, further tests the performance of the new system in the actual cases. And the main conclusions are as follows:(1) Using the single microphysical parameterization scheme and perturbating parameters, during the ideal storm tests, the scheme improves the assimilation performance. When the range of parameter perturbation is smaller, the assimilation performance is better. During the tests of assimilating real radar data, and using the optimal range of parameter perturbation, the improved scheme shows the most obvious advantages.(2) Using multi-scheme and perturbating parameters, and using the optimal range of parameter perturbation, the ideal storm tests make best assimilation performance. During the tests of assimilating real radar data, the scheme improves the assimilation performance to some extent, and has no obvious advantages compared with single microphysical parameterization scheme and parameter perturbation. But for lack of a priori in the actual numerical weather prediction, the improved scheme may be the better choice.(3) In WRF-EnSRF assimilation system, whether using a single microphysical parameterization scheme and perturbating parameters or using multi-scheme and perturbating parameters, both of them can improve the assimilation performance to some extent. Studies have also show that sometimes a single microphysical parameterization scheme and parameter perturbation has more obvious advantages, and thus to improve the performance of EnSRF assimilation system, only perturbating parameters in microphysical parameterization scheme is more convenient and feasible.