Improved Study Of AMSU-A Cloud Detection Method In GRAPES Assimilation System

AMSU-A(Advanced Microwave Sounding Unit-A)is notable for its ability to retrieve the vertical temperature profile of the atmosphere and cloud and rain information,and the assimilation of AMSU-A data contributes significantly towards reducing global NWP forecast errors.As AMSU-A detects the radiation of the earth-air system above the atmosphere,it is inevitably affected by clouds,accurate identification of clear-sky or cloudy-sky observations is a prerequisite.This work improves the cloud detection method of AMSU-A data in the Chinese operational numerical forecast system GRAPES?GFS(Global/Regional Assimilation and Pr Ediction System?Global Forcast System)by merging the data from AMSU-A and MHS(Microwave Humidity Sounder)on the same satellite platform.Due to the strong surface emissivity and high spatial and temporal variability that makes it difficult to distinguish between the radiative contributions of clouds and the atmosphere,the AMSU-A terrestrial surface cloud detection has always been a challenge.Based on the differences in the response characteristics of different channels to clouds,develop a new index for cloud detection over land,by further matching the MHS cloud detection index,we can effectively distinguish between cloudy and clear-sky observations.By referring to the MODIS(Moderate Resolution Imaging Spectroradiometer)cloud product,the POD(probability of detection)of the cloud fields of view with the new method was nearly 85%,22% higher than empirical cloud detection methods.By using the new cloud detection method to remove the cloudy data,the bias and standard deviation of the observation-minus-simulated brightness temperature(O-B)were significantly reduced.Finally,the new cloud detection method was added to the data assimilation system of GRAPES?GFS.The results show that the forecast model provides improved results for both the central and outer temperature and geopotential height fields of the typhoon and that the forecast tracks of Likima and Losa were closer to the observations.Using FNL(final analysis)data to quantitatively assess the effectiveness of new cloud detection methods for improving GRAPES?GFS forecasts of large-scale weather systems.The results show that the mean and standard deviation of the temperature and geopotential height fields are reduced,the spatial anomaly correlation coefficients between the 500 h Pa geopotential height fields and the FNL data are significantly improved.In a single assimilation experiment,the spatial anomaly correlation coefficient between the 120-hour forecast and FNL increased by 0.05,exceeding the 95% significance test,and the improvement is more significant as the forecast time progresses.