Bias Correction And Assimilation Study On FY-3A Microwave Humidity Sensor Data

The use of observation data to revise forecast model to get closer to the live state of the atmosphere, and then provide a more accurate prediction of the initial field for the next time is the goal of data assimilation. Affected by the terrain factors, conventional data lack of the the region of oceans and plateau observing data. Satellite data have the advantages of more consistent information, wider coverage and high spatial resolution to make up for the shortage of conventional data. At the beginning of the 21st century, china began to focus on the satellite data assimilation research, and a new generation of satellite data has been initially applied in the assimilation system, but the percentage of the assimilation data is lower compared with the total observation data. Therefore, in order to further improve the numerical prediction effect and compensate for the lack of conventional data on the impact of the initial value of the model, it is a meaningful topic that using China’s satellite data effectively in mesoscale numerical model.FY-3A satellite is the first star of China’s second generation of polar orbit meteorological satellite, which carries Microwave Humidity Sounding (MWHS) sensor can provide the global humidity information for medium-range weather forecasts. According to the emissivity data bias correction methods of Harris and Kelly, and combined with the feature of FY-3A MWHS data, this research focus on the bias correction method on the basis of MWHS data in 2010 January 1-15. On the basis of WRF-3DVar system, a bias correction software system has been developed, which will be joined to the system that can be directly assimilating MWHS data. Based on the successful foundation of the assimilation system, one assimilation experiment of MWHS data at 06:00 on January 18,2010 has been designed with a 3 hours time window, then analysis the assimilation results and evaluate the influence of MWHS data assimilation in numerical weather prediction.The main research results are as follows:(1)Based on the Harris and Kelly emissivity data bias correction methods, and combined with the feature of FY-3A MWHS data, a bias correction method for FY-3A MWHS data has been established, and a bias correction software system has been successfully developed. By using the developed bias correction software system to test showed that:The distribution of observation residuals after bias correction was very close to Gaussian distribution. MWHS radiance data shows the high quality and potential for data assimilation in this research. This bias correction scheme can be further joined to assimilation system, and can provide the condition for direct assimilation of MWHS radiance data.(2) On the basis of WRF-3DVar system, integrating the research results of the quality control, error correction, error of observation of FY-3A MWHS data, a satellite data direct assimilation system for MWHS data has been created.(3) The experiment of direct assimilation of FY-3A MWHS data shows that: MWHS data assimilation has played a certain improvement for various elements of different heights layer reanalysis, and an active positive effect for predict field. This research shows that the MWHS data can be further applied to the numerical weather prediction.