Research On The Applicability Of Precipitation Reanalysis Data In Cra-interim Of Climate And Weather Characters In China

Atmospheric reanalysis data is a powerful tool for research in atmospheric science.The precipitation products of atmospheric reanalysis data,compared with ground precipitation observation and other data assimilation products which is comprised of satellite and radar observation,is characterized by comprehensive spatial coverage,long time scale,dynamic and physical significance.Thus,it is widely-used and playing an irreplaceable role in climate detection and seasonal forecasting,climate variability and Change research,regional climate research,global water cycle research,and stratospheric research.Therefore,it is necessary to conduct a systematic evaluation of the applicability of reanalysis of precipitation data at different regional scales,time scales,and precipitation levels.This study uses the"China Ground Meteorological Station hourly observation data"in the China Meteorological Administration National Integrated Meteorological Information Sharing Platform?CIMISS?and three Atmospheric Reanalysis Datasets including CRA-Interim,ERA5and JRA-55 from 2007 to 2016.The 3-hourly precipitation data in global atmospheric reanalysis data were compared and analyzed on spatial distribution characteristics,seasonal distribution,variation characteristics and daily variation characteristics of precipitation.And also,precipitation prediction indicators for seasons,different precipitation quantitative different regions were analyzed and compared.At last,taking the heavy storm happened on 21^st July 2012?“7.21”?in Beijing and Autumn rain of West China in 2014 as examples,analyzed the simulated error characteristics of CRA-Interim on summer meso-scale extreme precipitation and frontal precipitation events.Preliminary conclusions of all above are as follows:?1?Overall,the magnitude of the error of CRA precipitation data and the other two sets of reanalysis data are simlar.Among them,ERA5 and JRA-55 show positive deviation overall,CRA shows negative deviation overall,and CRA error level is greater than The other two sets of reanalysis data,the overall error level of ERA5 and JRA-55 is basically the same,JRA-55 is slightly smaller than ERA5.?2?On the atmospheric reanalysis data error of precipitation spatial distribution,it is found that the applicability of reanalysis data in plain areas is better than that in mountainous areas.The Free Arm Rate?FAR?and Missing Rate?PO?in plain areas is significantly lower than that in complex terrain areas;In sparse mountainous areas,reanalysis of data may reflect details that some observations cannot reflect.The reanalysis data is superior to the precipitation in the arid regions,but it is also necessary to consider the discontinuous characteristics of the precipitation data,which will lead to the high precipitation area,which is the high value area of the precipitation variance.The CRA-Interim score is likely to affected by the topography.?3?On the atmospheric reanalysis data error of precipitation seasonal distribution,it is found that the simulation analysis of the seasonal movement of the rainband is slightly advanced in time,and the simulation of the spring precipitation is dominated by positive deviation;CRA-Interim gives a better simulation of Meiyu in the middle and lower reaches of the Yangtze River while has a negative deviation for the pre-flood precipitation in the southern China and autumn rain in West China.?4?On the error of precipitation daily variation,the day-time precipitation reflected by the reanalysis data is closer to observation value than the night precipitation;CRA-Interim performs better in daytime precipitation in most parts of China,and JRA-55 performs better in the night rain in southwest China.?5?On the error of different magnitude precipitation,the three sets of reanalysis data have consistent scores for the quantitative indicators of weak precipitation simulation,which can better reflect the seasonal distribution characteristics of climatic precipitation;however,the error fluctuation of strong precipitation It is relatively large,reflecting that the reanalysis data has a large simulation error for weather scale precipitation;CRA-Interim performs better in heavy precipitation than in weak precipitation,and its error in moderate rainfall is relatively higher than the other two sets of reanalysis data.?6?From the results of error synoptic analysis,taking the"7.21"heavy rainfall in Beijing in2012 as an example,the simulation error of the reanalysis data for summer extreme precipitation events is relatively large,especially for the mesoscale precipitation,the simulation ability is limited,the precipitation level is smaller than observation,the precipitation area is eastward,and the precipitation area moves faster.Taking the autumn rain event in West China in 2014 as an example,the reanalysis data can basically reproduce the falling area and periodic characteristics of persistent frontal precipitation,but the overall precipitation level is small,and the moderate intensity precipitation simulation ability is obviously better than the heavy precipitation,and the occurrence time of precipitation in some areas has a certain lag.