The Research On Optimal Configuration Of Radiosonde Observation System In 3DVAR Data Assimilation And Its Forecast

The sounding data is the most basic routine observation in data assimilation.Because of its high data quality,sounding data plays an important role in the effective improvement of analysis fields and forecast results.However,accessing to sounding data twice a day limits the accuracy of small and medium-sized analysis fields of atmospheric conditions to go further because of the low spatial resolution,the small quantities and the uneven distributions of the radiosonde stations.After the L-band radar-digital sounding system in our country has been updated,the sounding system could obtain the sounding data 4 times a day with high vertically spatial resolution and second-level and minute-level time resolution.At present,China is carrying out optimization and adjustment of the space-time configuration of the observation system.In order to accurately understand the impact of space-time configuration optimization of sounding stations on China's numerical forecasting quality,and to make full use of the effects of sounding observations on numerical forecasting quality,in this paper,the WRF mesoscale numerical model is used to analyze the sensitivity of sounding data to the assimilation analysis and forecast quality through the 06 UTC intensive sounding data and 12 UTC conventional sounding data and six kinds of resolutions of simulated radiosonde observations by observation system simulated experiment.The main findings are as follows:(1)Compared with assimilating 12 UTC conventional sounding observation,the impro-vement of the rainstorm forecast effects on afternoon is superior when assimilation 06 UTC intensive sounding data.To be specific,prediction grades of heavy rain and rainstorm by assimilating 06 UTC intensive sounding data are more accurate than by assimilating 12 UTC conventional sounding data.The root mean square error of prediction fields such as geopotential height,temperature and wind field is more effective at higher levels.The assimilation of the intensive sounding data at 06 UTC contributes more to the forecast quality of upper-level jet stream and lower-level water vapor flux divergence.Continuous 30 days tests further confirm that assimilating 06 UTC intensive sounding data is of positive signifi-cance for both analysis fields and numerical forecasting.(2)The spatial configuration scheme of radiosonde observation has a significant impact on the assimilation analysis and forecast quality.From the point of numerical forecasting model,the radiosonde observation configuration matched with the model horizontal resolution can maximize the analysis and forecast quality.Low-resolution radiosonde observations have low contribution to analysis and forecasting because of less observation information,while over-high-resolution radiosonde observations are difficult to obtain better analysis and forecast quality due to insufficient absorption capacity of the forecasting model.(3)The simulated sounding observation configuration with a spatial resolution of 110 km has the most significant contribution to the analysis and forecasting.The analysis error of assimilating radiosonde observation data and the corresponding forecast quality of this configuration scheme are obviously better than other schemes.And assimilating of simulated sounding observation data with a resolution of 110 km can improve the reproduction of small and medium-scale weather systems,and have significant positive contributions to numerical predictions of precipitation intensity,falling area,etc.,and further improve the forecast quality of small and medium-scale weather systems during heavy rain weather in China.