A Research On Different Radar Data Assimilation Schemes Of A Meiyu Front Precipitation

Doppler radar is an effective tool that can monitor weather conditions in real time and make short-term forecasts.Assimilating Doppler radar data to numerical model weather forecasting systems can effectively improve its forecasting effect.This paper uses the WRF3.9.1 model and its assimilation system WRFDA3.9.1 to study the precipitation process of Meiyu front in the Jianghuai region on June 26-28,2015,and uses NCEP / NCAR FNL reanalysis data as well as Nanjing Long Wangshan SA weather radar to make research on the assimilation experiments of doppler radar data and its improving effect on precipitation forecast.The main conclusions are as follows:(1)In the three-dimensional variational assimilation(3DVAR)experiment,the parameter adjustment of the background error covariance matrix has a significant effect on improving the assimilation prediction effect.Reasonable variance scale and feature length can effectively extract the information of the background error covariance matrix,thereby providing a better initial condition.(2)Under the condition of the same parameters,assimilating the radar reflectivity factor and the radial wind at the same time has the best effect.Assimilating the reflectivity factor and the radial wind alone can also improve the forecasting effect to a certain extent.In addition,the higher the frequency of assimilating radar data,the more obvious the effect will be.(3)In hybrid assimilation experiments,reasonable weight coefficients and localized radii can effectively reduce or avoid errors introduced by assimilation calculations by a limited set of members,and have a certain promoting effect on improving the effect of precipitation prediction.(4)Compared with 3DVAR,the assimilation effect of hybrid assimilation is more obvious in the improvement of precipitation forecast,especially in the heavy rain,rainstorm,large rainstorm,and extraordinary rainstorm levels.(5)A dual radar data assimilation experiment was performed after joining the SA weather radar in Hefei.The results of hybrid assimilation experiments of dual radar data prove that the system can not only achieve dual radar data processing,but more radar data can provide more accurate and effective storm scale information,thereby improve the prediction effect to certain degree.(6)Because the system's RCV method of generating ensemble set members has a certain randomness,the forecast effects are not always stable in the hybrid assimilation of dual radar data.However,to a certain extent,it can also reduce the problem of larger precipitation in the model itself;from the point of view of the precipitation score,it has also improved.