干旱典型山区CFSR降水数据的偏差校正方法研究——以新疆开孔河流域为例

摘要

CFSR降水数据在西北干旱半干旱区存在较大偏差,需要寻求合适的偏差校正方法.基于新疆开孔河流域内4个不同海拔高度的雨量站1981-2010年日观测数据,与同期CFSR降水数据对比分析后,分别采用LOCI法、Gamma分布映射法以及这2种方法的耦合(gamLOCI法)对CFSR降水数据进行偏差校正和检验.结果表明:新疆开孔河流域内原始CFSR降水数据表现出低估强降水、高估弱降水的特征, 且春秋两季与观测值的相关性优于夏冬两季,高海拔山区的相关性优于平原地区,夏秋两季的偏差小于春冬两季,山区和低海拔平原区的偏差大于较低海拔平原区;gamLOCI法综合了LOCI法和Gamma分布映射法的优点,既能校正CFSR降水数据的湿日频率和湿日降水强度,又能保留降水序列中的极值;在平原区,3种校正方法的效果均较好,无显著差别,而在高海拔山区,gamLOCI法校正效果最优.%Since there is obvious bias in reanalysis precipitation datasets in Central Asia, the appropriate bias correction has to be considered before applying these datasets in mountainous regions.Based on the daily precipitation from four weather stations with different altitude and CFSR precipitation reanalysis data during 1981-2010 in Kaikong river basin, the precipitation of CFSR data was compared with the observation firstly.Then, the precipitation of CFSR data was corrected and tested by applying three bias-correction methods, including local intensity scaling (LOCI), Gamma distribution mapping (Gam) and their coupling method (gamLOCI).The results are as follows: (1) Compared with the observation, the overestimation for light rainfall and underestimation for heavy rainfall of CFSR datasets were significant in Kaikong river basin.(2) The precipitation of raw CFSR in mountainous regions was better related with the observed data than that in plain area, and the correlations in spring and autumn were more remarkable than those in summer and winter.The precipitation biases in summer and autumn were less than those in spring and winter, and the biases in mountainous area and low altitude plain area were bigger than that in mid-altitude plain region.(3) The gamLOCI method synthesizes the advantages of LOCI and Gam methods, not only the wet-day frequency and precipitation intensity of CFSR data could be corrected, but the extreme was preserved.Three correction methods could equally and effectively improve the reliability of CFSR reanalysis data in plain area, while the correction effect with gamLOCI method was the best in mountainous region.