Technical Research Of Four-dimensional Ensemble Variational Assimilation Of Doppler Radar Data Based On Awinter Heavy Rainstorm In South China

The key technique of assimilating Doppler radar data with the Proper Othogonal Decomposition-based four-dimensional ensemble variational(POD-4DEnVar)assimilationmethod is studied by taking a heavy rainstorm occuring in South China as an example with the WRF model.This paper is divided into three parts.The first part of this paper analyzes the circulationand mechanism of the heavy rainfall,and carries out the numerical simulation experiment with the WRF model.The second part realizes the radar data assimilation based on the POD-4DEnVar method,and analyzes the reasons of the difference of the assimilation and simulation between the POD-4DEnVar and 3DVar method from the perspective of the development mechnisam of the heavy rainstorm.Also,the impact of assimilating different radar data on the simulation of the heavy rainstorm is further dicussed.The third part deals with the influence of different physical ensemble methods on the four-dimensional ensemble variational assimilation.The main conclusions are as follows:The circulation of this heavy rainfall occuring in South China is a typical north high south low circulation situlation in500hPa.The southern trough influences South China and the South China Sea becomes the main source of water vapor of the heavy rainstorm.The occurence of the heavy rainstorm is closely related to the cold air from the high latitudes in Eastern Europe.And the southward motion of the cold air triggers the frontogenesis which centers in 700hPa.The secondary circulation stimulated by the frontogenesis provides the dynamic conditions and is the mechanism of the development of the heavy rainstorm.The simulation of the precipitation with the POD-4DEnVar method ismuch closer to the observed one than that with the 3DVar method,because the two kinds of the assimilation methods have different adjustment to the initial conditions.The 3DVar method adjusts the initial wind field while the POD-4DEnVar method adjusts the water vapor condition to achieve the purpose of improving the simulation results.In addition,the precipitation simulation will be different when assimilating different radar data.However,it is worth noting that the initial wind fields and the water vapor conditions of these assimilation experiments are in similar conditions.The further study shows that it is the increase of the variable deviations and the difference total energy that leads to the significant difference in precipitation simulation among these experiments.Also,the rapid growth of the difference total energy is preceeded by that of the deviations of the variables and the precipitation,and the regions where the deviations of the variablesgrow the most evidently is the area where the gradient of difference total energy is large.The sensitivity of the POD-4DEnVar method to the parameterization schemes is investigated,and it is found that the choice of cumulus and microphysics schemes for ensemble samples significantly impacts simulated precipitation intensity,and that Doppler radar data assimilation using POD-4DEnVar is sensitive to the parameterization scheme used for ensemble samples.To obtain stable precipitation simulations,three methods are designed based on the physical ensemble technique,and it is found that the application of the physical ensemble technique is an effective method to improve the forecasting accuracy of the heavy rainstorm for the POD-4DEnVar method based on the initial ensemble samples.