Numerical simulation analysis on conventional data assimilation for a rainstorm in Chongqing

A synoptic overview of the heavy rainfall event occurred in Chongqing during 31st August to 2nd September was conducted using conventional observations and hourly precipitation data. We also used Weather Forecast Research (WRF) model and its assimilation system 3DVAR to assimilate the conventional radiosonde data into NCEP/NCAR reanalysis to generate new initial fields. The simulations of precipitation's spatial distributions with and without assimilation and the effect on simulating the mesoscale system's structure after assimilating radiosonde data were compared. The results indicates that the heavy rainfall event was the total consequence of southwest vortex, shear line and LLJ developed simultaneously in Chongqing against the large-scale circulation background of divergence in higher level caused by South Asia High and jets at 200 hPa, approaching the trough at 500 hPa and the westward move of the subtropical high. By comparing the simulations with and without assimilation, we found that both simulations could reproduce the large-scale circulation features while the simulation of precipitation area and intensity were improved when assimilated radiosonde data, especially significant for precipitation above the heavy rain magnitude. By analyzing the increment fields with and without assimilation at the initial moment, we also found that large-scale systems (South Asia High, subtropical High), mesoscale systems (vortex and jets) and moisture transport were all strengthened after assimilation, thus can support more favorable conditions for convective system's development and maintenance. Based on the analysis of heavy rainfall area's vertical structure during the most intensified rainfall period, the simulations of divergence, vorticity, vertical velocity as well as intensity and altitude of atmosphere's thermal structure were enhanced in various degree after assimilation, implying that simulations after assimilating radiosonde data can improve the distribution characteristics of initial fields, and have an essential effect on simulating mesoscale convective system's structure.