Study On Retrieving The Atmospheric Temperature And Humidity Profiles From Measurements Of Microwave Humidity And Temperature Sounder On FY-3C Satellite

Atmospheric temperature and humidity are the important parameters which are used to describe the atmospheric state.They play important roles in the field of atmospheric science,such as climate-change research,real-time weather analysis,numerical weather prediction,to name a few.Satellite-based Microwave remote sensing which has several merits such as overlay the region wide,continous observation and lower cost,is an important way to obtain the information of atmospheric temperature and humidity.Microwave Humidity and Temperature Sounder is one of the most important sensors on FY-3C satellite which is the new generation of polar orbiting meteorological satellite.MWHTS is a microwave radiometer including a microwave temperature sounder and a humidity sounder,can detect atmospheric temperature and humidity simultaneously.In this dissertation,the measurements of MWHTS are studied,and the MWHTS retrieval system of atmospheric temperature and humidity is built.The goal is to provide the high-precision profiles of atmospheric temperature and humidity which can meet the requirements of applications in the field of atmospheric science.Firstly,the principle of atmospheric microwave radiative transfer and the principle of microwave remote sensing are analyzed.The atmospheric microwave absorption models which are widely used in the world,are compared.Given the timeliness of the operational retrieval of atmospheric temperature and humidity,fast radiative transfer model RTTOV is introduced.The simulation accuracy of RTTOV is compared with that of line-by-line model.The results of comparison show the feasibility of RTTOV.The principles,advantages and shortcomings of retrieval methods of atmospheric temperature and humidity using microwave measurements is analyzed and compared.This provide the theoretical support for the building of MWHTS retrieval system of atmospheric temperature and humidity.Secondly,the one-dimensional variational algorithm based on physical retrieval method is used to retrieve atmospheric temperature and humidity using MWHTS observations.The factors of the algorithm affecting the retrieval accuracy is analyzed,then the MWHTS ocean temperature and humidity retrieval system in clear sky is built.From a physical retrieval perspective,the detecting ability of temperature and humidity of MWHTS and the quality of MWHTS observed brightness temperatures are evaluated.The retrieval results are compared with ECMWF reanalysis data,NCEP forecast data and RAOB data.The results of comparison show that MWHTS has strong detection ability of humidity and has strong detection ability of temperature in upper atmosphere.The high-precision profiles of atmospheric temperature and humidity can be obtained using MWHTS observed brightness temperatures.Thirdly,for the observations of MWHTS over land,in order to further improve the retrieval accuracy of atmospheric temperature and humidity and avoid the classification computation of land surface emissivity,a hybrid retrieval approach based on united matrix and individual matrix of background covariance is proposed,and MWHTS land temperature and humidity retrieval system in clear sky is built.The retrieval results show that the hybrid retrieval approach can improve the retrieval accuracy effectively.In the meantime,the retrieval results are compared with ECMWF reanalysis data,NCEP forecast data,NCEP analysis data,RAOB data,AMSU retrieval results and MIRS retrieval results.The results of comparisons show that the retrieval results of MWHTS are sound,and this further evaluated the detection ability of atmospheric temperature and humidity of MWHTS and the quality of MWHTS observed brightness temperatures.Fourthly,in order to increase the temperature retrieval accuracy and retrieve atmospheric temperature and humidity in all weather conditions,the retrieval system based on MWHTS retrieval system in clear sky can be improved by two ways.one is that the statistical retrieval results are taken as the first guess in the physical retrieval system.This method can increase the humidity retrieval accuracy effectively,improve the MWHTS data utilization rate and avoid dependence on the third-party data source.The other one is improving the correction method of bias between simulation and observation of MWHTS.Through the analysis of a correlation between brightness temperature and atmospheric temperature and humidity parameters,two bias correction method are proposed,which are based on neural network algorithm and linearity regression algorithm,respectively.The experiments of bias correction and the experiments of atmospheric temperature and humidity are conducted.The experimental results show that the neural network correction has better performance.The algorithm stability test result show that the neural network correction can be used in physical retrieval system.Based on the above improvements,MWHTS temperature and humidity retrieval system is built.Finally,retrieving the atmospheric temperature profiles using the measurements of FY-3C/MWTS-II is studied,the retrieval results are compared with MWHTS retrieval results,and the detection ability of atmospheric temperature in 50~60 GHz and 118 GHz is evaluated.According to the comparison inversion results,the first guess of temperature in MWHTS temperature and humidity retrieval system is replaced by the statistical retrieval.This further improve the atmospheric temperature retrieval accuracy.