Study On Radar Rainfall Estimation And Hydrology Application Based On Z-R Relation Optimization

Rainfall is one of the most important factors affecting the occurrence and development of flood disaster.Due to climate change and the rapid development of urbanization,the frequent occurrence of extreme rainstorm events has caused significant economic losses.Therefore,in the development of rainfall monitoring technology,higher requirements are put forward for the resolution and spatial and temporal distribution characteristics of rainfall data,so as to better simulate,predict the flood process and reduce the impact of extreme rainstorm events.Weather radar,as an important part of "sky and earth integrated monitoring",has the characteristics of wide rainfall monitoring range and high spatial and temporal resolution of data,which is expected to solve this problem.However,the low accuracy of weather radar rainfall data has seriously restricted its further development.It is necessary to optimize the relationship between radar reflectivity Z and rainfall intensity R to improve the accuracy of radar rainfall estimation by combining the data of ground rain stations.Therefore,how to optimize Z-R relation to calibrate radar rainfall data has become one of the research hotspots in recent years.This paper took the valley above Biliu River Reservoir.10 rainfall processes from 2015 to2019 were selected as the research object,and radar data were calibrated by ground measured rainfall data and weather radar base data of Dalian Station.Based on the Z-R relationship of radar inversion,the factors affecting the Z-R relationship were analyzed.The methods of "overall optimization" and "segment optimization" were adopted to improve the accuracy of radar estimation of rainfall,and the results of point rainfall and area rainfall of inversion data were analyzed and evaluated.The HEC-HMS hydrological model was combined with the calibrated radar rainfall data to simulate the flood in a watershed.The differences in the simulation results between the two rainfall data-driven hydrological models of rain stations and radar were compared and analyzed.Through the research and analysis of this paper,the following results are obtained:(1)The radar data of Dalian station has the problem of beam blocking in the valley above Biliu River Reservoir,which leads to the calculation result of its reflectivity is seriously low.(2)After the radar beam blocking and other errors are suppressed,the "overall optimization"method is adopted to improve the accuracy of the radar estimation of rainfall to some extent.However,the estimation results of the stations far from the radar center are poor,and there are still serious phenomena such as the underestimation of the maximum daily rainfall.(3)By considering the reflectivity segmentation and site according to the distance of grouping optimization idea,namely "section optimization method,can effectively solve the high reflectivity value underestimate,the phenomenon of low overestimate,surface precipitation and radar result analysis shows that after the calibration of radar rainfall data can better reflect the time and space distribution characteristics of rainfall.(4)The HEC-HMS hydrological model was driven by rainfall station data and radar rainfall data,respectively.The results show that the radar data with high spatial and temporal resolution can better simulate the flood process,and has better applicability in distributed hydrological model application.