Research On Minute Precipitation Estimation Method In Southwest Mountain Region By Integrating Weather Radar And Satellite Data

Precipitation is the primary component of the water cycle system and has important climatological,meteorological,and hydrological significance.Therefore,the application of various remote sensing data sources for accurate precipitation inversion is a significant scientific research work.Using remote sensing technology to retrieve precipitation can make up for the shortcomings of traditional surface rain gauge statistics,improve the scope of precipitation estimation,operational efficiency and reduce costs.In this thesis,on the basis of comprehensive domestic and foreign remote sensing precipitation retrieval theoretical methods and practical applications,the use of weather radar and stationary meteorological satellite data to carry out the study of precipitation estimation methods in the southwest mountainous region.The main research contents of this thesis are as follows:This thesis selects Himawari-8 satellite image data,weather radar data,and ground precipitation data,using random forest(Ramdom Forest)method,least square method(Least Square Method)fitting,simulated annealing algorithm(Simulated Annealing Algorithm)optimization,ZR Relations to establish a precipitation inversion model.Taking Yunnan Province as an example,carry out precipitation estimation methods in the mountainous regions of Southwest China.Taking the three precipitation processes on May 27,2020,June 13,2020,and June 30,2020 as typical cases,establish the brightness temperature-rain intensity function relationship of satellite data-precipitation data,and the Z-R of radar data-precipitation data Relations and fusion of the two inversion results to obtain the final fusion weather radar and satellite data precipitation inversion model.The research results show that the relationship between Brightness Temperature and Rain Rate can be fitted to an exponential function relationship,and it decreases monotonously.The application of random forest algorithm can greatly improve the inversion accuracy,and invert moderate to heavy rain precipitation.The effect is better.The specific cloud top brightness temperature-rain intensity relationship function is=1.866×10⁴0)^-0.0268×13-20.96.The relationship between radar reflectivity and rain intensity can be fitted into an exponential function relationship of the form Z=a*B⁽(7),and the retrieval effect of moderate to heavy rain precipitation is better.The specific ZR function relationship is Z=160.7308*R^-0.1718,this ZR relationship function is suitable for the rainy season in mountainous areas of Yunnan Province and used for weather radar to retrieve precipitation.Considering comprehensively,the precipitation inversion model based on the fusion of weather radar and satellite data established in this paper has a high accuracy rate of96.08%,which has reduced the rate of precipitation false alarms,and solved the problems of false alarms and errors to a certain extent.It can effectively solve the problem of precipitation estimation bias in the southwest mountain area relying solely on weather radar or satellite data,and optimize the overall running time of the fusion inversion model algorithm to about 151 seconds to achieve minute-level precipitation estimation,which can meet the needs of the southwest mountain area.Demand for real-time precipitation estimates.