Simulated Geostationary Orbit Microwave Data Assimilation Experiments Based On WRFDA-4DVar

Geostationary orbit microwave(GEO-MW)radiometer atmospheric detection can monitor the three-dimensional temperature and humidity structure of the atmosphere all-time and all-weather.The use of observational data with high spatial and temporal resolution for numerical assimilation is expected to improve weather forecasts for severe convective weather such as typhoons and heavy rains,which is of great significance for improving the level of numerical weather forecast.However,because the high spatial resolution detection technology of GEO-MW meteorological satellite is difficult to achieve,there is currently no GEO-MW meteorological satellite in orbit in the world.China is concentrating its efforts on key technologies,and may be the first to launch a GEO-MW meteorological satellite.It will work with optical detection satellite in orbit to form a high-orbit weather satellite system with complete detection methods.As so far,because there is no actual GEO-MW data,it is difficult to evaluate the practical application effect of GEO-MW observation methods.Therefore,the assimilation experiment of simulated GEO-MW data is carried out in advance,which can explore the way of using GEO-MW data,evaluate its improvement effect on typhoon weather forecast,and accumulate experience for the practical application of GEO-MW data in the future.This thesis is based on the WRF numerical prediction model and the WRFDA-4DVar system to carry out the 4DVar assimilation experiment of the simulated GEO-MW data.The 4DVar assimilation method can assimilate the observation data at multiple moments and extract the dynamic information of the observation data distributed over time.Therefore,it is very suitable for assimilating the high-frequency GEO-MW data.This thesis selects No.1506 Typhoon "Noul",No.1601 Typhoon "Nepartak" and No.1617 Typhoon "Megi" as cases to simulate 50?60GHz and 183 GHz GEO-MW data to carry out4 Dvar assimilation experiments.Finally,the 72 h path and intensity prediction results of the typhoon are used to evaluate the assimilation effect of the GEO-MW data and analyze the impact of the three factors of data time frequency,system noise and resolution on the assimilation performance.The experimental results show that assimilating the atmospheric temperature band at50?60GHz data can significantly improve the typhoon path forecast.Compared with the average path deviation of the control experiment,the forecast accuracy of typhoons "Nepartak","Megi" and "Noul" increased by 25.35%,27.45% and 5.11% respectively.After assimilating the data of the atmospheric humidity frequency band at 183 GHz,the forecast accuracy of the average intensity of typhoons "Nepartak","Megi" and "Noul" are improved by-3.05%,34.38% and 5.11% respectively,but the positive effect can't be maintained all the time.The results of assimilation experiment on the three factors of time frequency,system noise and resolution of simulated GEO-MW data are as follows.As the data time frequency increases,the 4DVar system can use more observation data to improve the assimilation effect,which leads to a reduction in the average path deviation of the typhoon.As the system noise of the data increases,the quality of the geostationary orbit microwave data decreases,which leads to an increase in the average path deviation of the typhoon.As the resolution increases,the resolution of the geostationary orbit microwave data is only improved at the eye of the typhoon,and there is no obvious difference in the typhoon path prediction of the assimilation experiments with different resolutions.