新疆地区多源降水融合试验

Multisource precipitation data merging experiment in Xinjiang

利用新疆地区2016年夏季(6—8月)近2000个区域自动站小时降水数据,以最新一代GPM(Global Precipitation Measurement Mission,全球降水测量计划)IMERG(Integrated Multi-satellite Retrievals for GPM)卫星降水产品为初始场,运用概率密度匹配与最优插值两步融合校正方法(probability density function-optimal interpolation,PDFOI),开展干旱区多源降水融合试验。首先,通过将实况观测与卫星反演降水进行概率密度函数(PDF)匹配,从而实现对IMERG降水的偏差订正;然后以偏差订正后的IMERG降水为初估场,实况降水为观测场,通过最优插值(OI)将偏差订正后的IMERG降水量估计值与雨量计实况相结合,最终得到经过两步校正后的新疆地区小时雨量数据。交叉验证结果表明,与雨量计实况和原始卫星降水数据相比,PDF-OI两步校正方法构建的融合降水不仅大大消除了系统误差,而且显著提高了数据精度。

Precipitation is a fundamental component of the Earth’s water cycle.Recognizing the temporal and spatial distributions of precipitation in a specific area is helpful to better understand the ecological and environmental changes in this area,and the most important of these is to accurately obtain the precipitation data for the entire area.Therefore,developing high spatiotemporal resolution and high-quality precipitation data is an urgent need as the development of modern meteorological services.In this paper,the study on a fusion method based on two-source precipitation,that is,satellite precipitation data and ground observation,is carried out in the Xinjiang area with sparse stations in China,which can be of great significance to accurately reflect the precipitation distribution information and the characteristics of precipitation evolution process in this area.Using hourly precipitation data in summer(June-August)2016 of about 2000 automatic weather stations(AWS)over Xinjiang and using the latest generation of GPM IMERG satellite precipitation data as the first guess,a twostep strategy based on the probability density function(PDF)matching and optimal interpolation(OI)technique(PDF-OI)was conducted to a study on fusion methods of precipitation in arid regions.First,the bias of the IMERG precipitation is corrected by matching the PDF of the observation with the satellite data.Then,the bias-corrected IMERG precipitation(pdf_CRT)is used as the initial estimation field and the observation field(OBS)as the truthvalue;the combination of the pdf_CRT and OBS through the OI method is carried out to finally obtain hourly areal precipitation data(pdf-oi_CRT)in Xinjiang.The cross-validation results show that,compared with the OBS and the satellite precipitation,the PDF-OI fusion method not only greatly eliminates the bias but also significantly improves the accuracy of the original satellite precipitation data.In this study,a useful exploration is made in developing a grid precipitation data set in Xinjiang,which is located in a middle-to high-latitude arid area.The results show that the study on the multisource precipitation fusion method helps develop grid precipitation data set.The high temporal and spatial resolution grid precipitation data set will help us understand the characteristics and occurrence and development regularity of hourly heavy precipitation in the Xinjiang area.Furthermore,it will also provide data support for improving the accuracy of forecast and preventing secondary meteorological disasters and the technical foundation for developing high spatial and temporal resolution precipitation products in the Central Asian region.