Experiments Of Assimilating Amsu-A Observations At The Surface Sensitive Channel

Up to now, satellite data assimilation research is concentrated on the variational assimilation method and the Kalman filter is not widely used. Moreover, in the most published research studies only the satellite data at the channels which are less sensitive to the surface are adopted while the observation data at the so-called surface sensitive channels have been discarded. Doing so may cause a waste of resources on the one hand, on the other hand may also lose some important observation information, reduing the satellite data assimilation effect. Based on above consideration, in this study the ensemble square root filter (EnSRF) data assimilation scheme is adopted to investigate the possibility of improving the initial model states by assimilating the AMSU-A observations at the surface sensitive channels into a mesosacle model.Firstly, by using RTTOV v10(Radiative Transfer for TOVS) model with including the new surface microwave emissivity model, the simulated brightness observations for the AMSU-A first channel are assimilated into the weather research and forecasting (WRF) model with the EnSRF in order to investigate the possibility of improving the model initial fields in Jianghuai area of China. The results showed that: on the sunny day, the four elements of analysis fields for the potential temperature, water vapor mixing ratio and horizontal wind speed u and v have improved but the degree of improvement among the different variables is not the same at the different model levels. Specifically speaking, the improvement of the horizontal wind speed is the largest while that of the temperature is the slightest. For the rainy weather condition, the performance in the four analysis fields is generally similar to that on the sunny day. However, the spatial distribution of improvement on the rainy day is different with that on the sunny day.Finally, the actual AMSU-A first channel satellite observations are assimilated to investigate the possible impact on the heavy rain forecasts which occurred in June2003in Jianghuai valley. The results showed that:the assimilation of actual satellite observations can effectively narrow the gap between the simulated results and observed results, and improve four model states with different degrees respectively, showing that assimilating AMSU-A first channel can improve the model states. Precipitation intensity forecasts have also been improved after using the analysis field as an initial field, but there exists some spurious precipitation forecast and the location of rain belt has been moved to the south so further more tests are need.