Research On Inversion Methods For Passive Remote Sensing Of Ice Cloud By Terahertz Wave

Ice clouds play an important role in the balance of the earth-atmosphere system radiation budget.At present,a variety of remote sensing methods have been developed to measure ice cloud parameters.Recent studies have shown that terahertz waves have unique advantages in remote sensing ice cloud characteristics because the terahertz wavelengths are similar to typical ice cloud particle sizes.In this paper,several representative channels in terahertz band are selected to conduct simulation based on ARTS(the Atmospheric Radiative Transfer Simulator),and various inversion methods of ice cloud parameters are studied.Firstly,a new idea of multi-channel regression method for ice water path inversion by terahertz wave is proposed.The simulation results show that the fitting degree of the stepwise regression equation can reach 0.951.This method can effectively separate the effect of ice water path under comprehensive influence of complex multi-factors,and can quickly and effectively perform terahertz ice water path inversion.Secondly,an intelligent remote sensing scheme of ice cloud parameters by terahertz wave based on genetic algorithm and simulated annealing algorithm is established.The fast forward operator of bright temperature difference is constructed by combining threedimensional interpolation,and the integrated intelligent inversion of multiple parameters such as ice cloud height,ice water path and effective particle diameter is realized.It is further clarified that setting multiple initial values can improve the effect of simulated annealing inversion algorithm.Finally,from the perspective of probability and statistics,a priori retrieval database of China region is established based on observed cloud profile data,and the BMCI(Bayesian Monte Carlo Integral)method is adopted to realize simultaneous inversion of ice water path and liquid water path by combining terahertz wave and microwave.The experiments show that the Bayesian inversion effect of the cloud water path depends on the amount of information in the priori retrieval database.