Evaluation Of Multiple Precipitation Products In The Qinling-Daba Mountains

As one vital hydrometeorological variable,precipitation has important practical and theoretical significance to the research of meteorology,climate,hydrology,and geology and other fields.The accurate acquisition of the spatial and temporal distribution of precipitation is the basis and precondition for the understanding of the hydrological process,the monitoring of drought and the utilization and development of water resource.The Qinling-Daba Mountainous areas in the central part of China are the geographical dividing lines of southern and northern China,and the dividing lines of subtropical zone and warm temperature zone.At the same time,they are an important source of water for the Middle Route Project of South-to-North Water Diversion Project.They also are the treasure house of natural animal and plant resources,whose ecological and geographical significance is self-evident.Accurate and continuous precipitation data is an important basis for scientific management of water resource in this region.With the improvement and development of satellite observation technology and numerical simulation technology,a large number of precipitation products with good space and time continuity have emerged,which provide useful supplements to the traditional dot precipitation observation methods.The accuracy of precipitation products in different regions is uncertain.Applicability of different precipitation products must be fully evaluated before application.This study conducted the precision test of precipitation products in the Qinling-Daba Mountains to provide the basis to comprehensively understand the performance of precipitation products at various time scales and choose more suitable precipitation products in the study area.In view of this,27 meteorological stations are regarded as “true value” in this study to analyze the temporal and spatial distribution of eight precipitation products including CMORPH,GPCP-2,TRMM 3B43,GPCC,CMFD,ERA-Interim,JRA-55,and NCEP/NCAR-1 through the precipitation centroid method,correlation analysis method,and so on,furthermore,the overall accuracy on a monthly,seasonal,and annual scale and the performance at individual stations is also elaborated.The main conclusions are as follows:(1)From the angle of time change,most precipitation products present consistent temporal distribution of precipitation with the observed precipitation to some extent,especially CMFD,GPCC,and TRMM 3B43,but the numerical simulation capabilities of ERA-Interim and JRA-55 still need to be improved.From the perspective of spatial distribution,CMFD,GPCC,TRMM 3B43,GPCP-2,and JRA-55 have better performance,all of which can show the basic spatial distribution characteristics of more precipitation in the southwestern part and less precipitation in the northwestern part of the study area,and CMFD is the best one among them.Judging from the spatial location and moving direction of precipitation center of gravity,most precipitation products can reflect the basic change trend of precipitation in the Qinling-Daba Mountains.CMFD show the best consistence with the actual precipitation,and TRMM 3B43 and GPCC performed well,but the difference between NCEP/NCAR-1 and JRA-55 with the observed precipitation is the greatest.Meanwhile,no obvious precipitation gradient was found in the Qinling-Daba Mountains from the available data.(2)On a monthly scale,CMFD is highly and obviously correlated with the observed precipitation in 12 months with a pretty small amount of error,followed by GPCC,TRMM 3B43,GPCP-2,and ERA-Interim.JRA-55 and CMORPH perform poorly among them,and NCEP/NCAR-1 is too poor to show the ability to simulate the precipitation at a monthly scale.On a seasonal scale,CMFD outperforms the other precipitation products and NCEP/NCAR-1 performs worst among them,and performance ranking of the other six precipitation products is GPCC,TRMM 3B43,GPCP-2,CMORPH,ERA-Interim,JRA-55.In addition,eight products have better consistence with the observed precipitation in summer and autumn than winter and spring,which showed the precise of precipitation products may be related to the abundance of rainfall.On an annual scale,the performance of eight products is ranked as: CMFD > GPCC > TRMM 3B43 > GPCP-2 > CMORPH > JRA-55 > ERAInterim > NCEP/NCAR-1.(3)At the individual site scale,the performance of eight precipitation products varies in the representative months at four seasons.Overall,GPCC,CMFD and TRMM 3B43 show pretty good precision,and CMORPH and NCEP/NCAR-1 are the worst among them.However,the performance of NCEP/NCAR-1 is improved compared the overall accuracy on the time scales.CMFD,GPCC,TRMM 3B43,and GPCP-2 show pretty stable ability in the relatively arid and humid years,whose PPDs(Precipitation Percentage Difference)keep in the range of 10%.Meanwhile,the performance of ERA-Interim and JRA-55 in the relatively humid year are much better than that in the relatively arid year.(4)All in all,CMFD,GPCC,and TRMM 3B43 show strong applicability in the QinlingDaba Mountains,while CMORPH and NCEP/NCAR-1 are the worst of the eight precipitation products.