Research On The Scene-adaptive Observation Error Of Fengyun-3 D-satellite Microwave Sounder Based On ARMS Observation Operator

The assimilation of massive satellite data has played a key role for Numerical weather prediction(NWP),especially the assimilation of microwave sounding data.The fast radiative transfer model is the basis of satellite data assimilation.In order to meet the growing demand for the application of Chinese satellite data,CMA has developed its own all-sky satellite observation operator ARMS.In order to apply the domestic ARMS model to the numerical forecast system,it is of great importance to construct the corresponding observation error matrix.In this study,based on the new observation operator ARMS,the key technologies for the construction of observation error covariance matrix such as scene identification scheme and allsky observation error analysis are conducted for applications of Micro Wave Temperature Sounder(MWTS)as well as Micro Wave Humidity Sounder(MWHS)onboard Chinese polarorbiting Fengyun-3D satellite.The main research covers the following four topical areas:(1)The clear-sky radiative transfer simulation is the basis for constructing the observation error covariance matrix.In order to improve the accuracy of clear-sky radiative transfer simulation,the microwave fast transmittance model considering the curvature effect of the Earth is constructed in this paper.The newly constructed ODPAS?WV algorithm,which considers the ODPS algorithm based on the original ODAS algorithm and adds the water vapor related predictors,is used in the scheme.In the comparison results with the accuracy of the existing ODAS algorithm,it is proved that the curvature-corrected ODPAS?WV algorithm is more suitable for calculating the absorption of water vapor in the microwave spectrum.In water vapor absorption channels,using the LBLRTM as the standard,a maximum reduction of 42.8%in the root-mean-square error of the ODPAS?WV algorithm results compared to the ODAS algorithm.By comparing with the observations,the standard deviation of the ODPAS?WV algorithm results is reduced by up to 0.7K compared to the ODAS algorithm.(2)Cloud detection is important for the construction of the scene-dependent observation error matrix.To enhance the applicability of the cloud emission and scattering index(CESI)detection scheme,the corrections of the scan-angle-dependent bias under different surface conditions are added based on the original scheme.In addition,different combinations of CESI thresholds for clouds of different heights and surface types,are selected for the identification of clear-sky area.The results show that the new threshold combination can be more effective for cloud identification compared with the original fixed thresholds and the probability density distribution of the screened clear sky data for most channels approximates a Gaussian distribution.(3)In this paper,the simulation accuracy of ARMS model is compared with CRTM under all-sky conditions,and the all-sky simulation error of ARMS is quantified.The results show that the accuracy of ARMS is comparable to that of CRTM under all-sky conditions.In addition,the simulation accuracy of 4 stream HRTS is comparable to that of 16 stream ADA for most microwave sounding channels when being coupled into ARMS,meeting the accuracy requirement of all-sky assimilation.(4)The error of the observation data is an important part of observation error covariance matrix.In this paper,a common calibration algorithm for microwave sounders is established to quantitatively evaluate the instrument observation errors of FY-3D microwave sounders.In the calibration algorithm,the Allan deviation method is used as a standardized method to compute the equivalent noise temperature(NEd T).The stability of the Allan deviation method is verified by comparing it with the standard deviation method for the calculation of unstable time-mean data.Meanwhile,based on the calibration results,a more comprehensive analysis of the observation error of FY-3D MWHS is conducted through the proposed quantitative multisource evaluation scheme,and it is found that the overall performance for MWHS temperature sounding channels is better than for water vapor channels.