Retrieval Of Parameters Of Various Frozen Hydrometeor Of Ice Cloud Based On Terahertz Brightness Temperature

The particle density in the ice cloud is closely related to the formation of precipitation material,which is of great significance for the analysis and study of the characteristics of the ice cloud,the prediction of the weather state and estimating the impact of the climate.Because the wavelength of the terahertz wave is similar to the size of the ice cloud particles,the ice cloud particles have a strong scattering effect on the terahertz wave,and it is the best wave band for theoretically detecting the microphysical properties of the ice cloud.After years of development,a series of researches on ice cloud detection theories,detector schemes and a series of experiments have led to certain progress in terahertz ice cloud detection technology.However,the research of terahertz detecting ice cloud on the experimental system and load of started relatively late in china,and it is necessary to strengthen the research on the relevant principles and technology of terahertz ice cloud detection.The current technology of terahertz ice cloud detection has the following three problems.One of the problems is that the the current terahertzice cloud detection research does not take the contribution of trace gases in the upper atmosphere to the terahertz radiation brightness temperature of the top layer of the atmosphere into consideration;the second problem is that the ice phase particles are seen as a kind of particles in the current research on the using terahertz detect ice clouds.While from the perspective of atmospheric physics,the ice crystal particles in the ice cloud actually consist of three parts such as ice particles,graupel particles and snow particles.third problem is that at present,the field of meteorology divides ice clouds into two types: convection and cirrus.This article focuses on the above three problems of using terahertz detect ice cloud technology.First,the middle and upper trace gas components are considered in the retrieval and the influence of the trace gas on the retrieval accuracy.Second,the three phases of ice,snow,and graupel are respectively retrieved from the brightness temperature data of the terahertz multi-band ice cloud detection.Finally,the classification of convection and cirrus in the terahertz ice cloud detection is realized through the ice cloud classification algorithm of retrieval.The experiment uses neural network and one-dimensional variational algorithms.First,the88+118GHz frequency band is used to retrieval the graupel particle parameters,then the obtained graupel profile is used as the prior value,and the 150+183+243GHz frequency band is used to retrieval the snow particle parameters,and finally the graupel profile and snow profile obtained are used as a priori,and the 325+448+664+874GHz frequency band to retrieve the ice particle parameters.The results show that trace gas has a certain effect on the detection of ice clouds in terahertz.The upper trace gas was considered in the retrieval which improved the retrieval accuracy;the experiment verified the effectiveness of the using 88?118GHz brightness temperature retrieval of graupel parameters,150?243GHz brightness temperature retrieval of snow parameters and 325?874GHz brightness temperature retrieval of ice parameters.The relative errors of GWP,G?Dme and G?Zme are within 33.7%,10.8%,and 8.0%,respectively.The relative errors of SWP,S?Dme and S?Zme are within 30.6%,32.4%,and6.8%,respectively.The relative errors of IWP,I?Dme and I?Zme are within 31.6%,20.3%,and 6.5%,respectively.and it is also beneficial to improve the retrieval accuracy when classifying scenes into convective clouds and cirrus cloud retrieval;using one-dimensional variational algorithm can optimize the neural network retrieval results.