Research On Quantitative Precipitation Estimation Method Of Dual Linear Polarization Radar Combined With Raindrop Spectrometer

Quantitative precipitation estimation(QPE)of dual polarization radar is an important application of weather radar,and improving the accuracy of radarq uantitative estimation of precipitation is not only a long-term research hotspot in the application of weather radar data,but also an important requirement to promote the development and upgrade of weather radar technology.The dual-polarized variables enrich microphysics information on the precipitation process,and has a positive effect on improving the radar-based quantitative precipitation estimate,the hydrometeors phase classification and the data quality control.In the past ten years,It is widely ongoing for many countries around the world to upgrade Doppler weather radar with the dual polarization detection function..Compared with the S and C bands,the X-band differential phase is more sensitive to strong convection and its accompanying hail processes,and has the advantages of less equipment,light weight,and low cost.Therefore,the X-band dual-polarization weather radar is used to artificially affect the weather and coordinated network observation.X-band dual-polarization weather radar quantitative estimation of precipitati on accuracy is affected by many factors such as rain attenuation along the elec tromagnetic wave propagation path,partial beam blockage caused by complex terrain,system bias caused by insufficient calibration accuracy,and precipitation relationship differences caused by regional climate diversity.Raindrop size distribution(DSD)observation data over a period of time can reveal the rainfall type,empirical constraints between radar parameters,and empirical constraints between radar variables and rainfall rate during the period.It has a positive effect on improving precipitation relationships,mitigating attenuation effects,and mitigating system bias.Therefore,in view of the limiting factors that affect the accuracy of QPE estimation,this paper combines the observation data of raindrop distrometer and X-band dual-polarization weather radar to conduct related research and experiments.Firstly,the radar parameters and attenuation factor Ah?Adp and rainfall rate R were simulated using DSD data combined with T matrix method.Then regression is used to fit the attenuation relationship Ah(Kdp)Adp(Ah),consistency relationship Kdp(Zh,Zd.),and rainfall rate estimator R(Kdp),R(Zh,Zdr)R(Zh).Furthermore,quality control is performed for radar data:1.To estimate Kdp from raw ?dp using median filtering,Kalman filtering,and polynomial fitting.2.To Correct attenuation using the fitted attenuation relationship Zh,Zdr.3.To correct the energy lost due to partial beam blockage with Digital Elevation Model(DEM)data.4.To perform contrast expeirment between "low angle light rain" and "dry snow" to correct the Zdr systematic bias.5.Use consistency relationship to correct Zh system deviation.Finally,the QPE relationship was used to retrieve the rainfall using the radar parameters after quality control.This article conducted an experimental data analysis and evaluation of an X-band polarization radar(XFXLD)deployed in Xinfeng,Guangdong,during the flood season before 2016.The focus is on parameter quality control and rainfall product inversion for XFXLD.After inversion of the drop-spectrum relationship in this area,the quality control for radar variables was performed:1.The?dp was filtered to calculate the more accurate Kdp.2.Corrected attenuation using a fitted Zh,Zdr attenuation relationship.3.After comparing the accuracy of the SRTM task and the DEM data of Google,the latter is selected to calculate and compensate the Zh energy lost due to occlusion.4.According to the observation results of "low-angle light rain" and "dry snow",it is found that the standard deviation of the Zdr system deviation is less than 0.2dB.5.The Zh system deviation corrected by the parameter consistency and cross-comparison with two standard S-band radars show that the mean deviation is less than 1.5 dBz and the standard deviation is less than 0.5 dBz.Then,15 precipitation events were analyzed to prove the performance of XFXLD's rainfall estimation.Use its corrected echo volume to bring into the accumulated rainfall of 1,2,and 3 hours calculated by R(Kdp),R(Zh,Zdr)which is compared with the rainfall estimates of national rain gauge data and two standard S-band radars.The results show that the R(Kdp),R(Zh,Zdr)-based rainfall estimates are in good agreement with the rainfall gauge observations and the rainfall estimates from both radars.The performance of R(Kdp)is slightly better than R(Zh,Zr).Finally,according to the test method of this paper,developing a software system for correcting and display dual-polarized variables with QT programming.