Font Size: a A A

Research On Application Problems Of Rainfall Measurement Technology By Using Line-of-sight Microwave Link

Posted on:2020-08-26Degree:MasterType:Thesis
Country:ChinaCandidate:B S HeFull Text:PDF
GTID:2480306548495544Subject:Journal of Atmospheric Sciences
Abstract/Summary:PDF Full Text Request
Real-time,accurate and detailed monitoring of precipitation is of paramount importance in meteorological and hydrological research as well as in meteorological operations.Since 2006,the theory of using the attenuation information in the microwave communication signal to measure precipitation has been proposed.This new method of precipitation monitoring with fine temporal and spatial distribution has great potential for improving precipitation information.However,there are still a series of problems in the actual applications of measuring precipitation using microwave links,such as difficulty in extracting rain attenuation in low frequency links,significant differences in data quality of different communication devices,and etc.,and these factors limit the available equipment for rainfall monitoring.Based on the method of monitoring precipitation by using microwave link,this paper focus on the working characteristics of microwave equipment in communication network and some key technical aspects of precipitation monitoring,the theoretical analysis and corresponding experiments are conducted,and corresponding solutions to the problems are proposed.Firstly,the theoretical model is used to calculate the rain attenuation characteristics of the microwave link.Combined with various attenuation factors of microwave transmission in the atmosphere,the objective conditions of the microwave link inversion rainfall are analyzed,and the applicable range of the microwave link that can be used for measuring rainfall are improved.Secondly,the distribution law of multi-time scale features of microwave link signals is studied and the variables that re sensitive to precipitation are fully exploited.And the strategy of the identification of wet/dry period by using the decision tree and LSTM neural network algorithm is proposed.The average accuracy of the final rainy season is as high as 80%..Finally,aiming at the sparse regional distribution of the microwave links,a construction method by jointly using microwave link,weather radar and rain gauges are proposed,and a two-dimensional rainfall field strategy based on the Kriging interpolation algorithm of the correction factor is constructed.The error between the accumulated rainfall and each reference rain gauge is generally less than 20%.The work of this paper has significance for making full use of the existing communication microwave link to inverse precipitation and meet the needs of the public and scientific research for precipitation information.
Keywords/Search Tags:Microwave links, Data mining, Wet-dry classification, Classification and Regression Tree, Long Short-Term Memory
PDF Full Text Request
Related items