【摘要】 降水资料中的分析和再分析降水数据可以弥补台站观测数据的不足,对于气象学和水文学具有重要的应用价值。利用目前国际上7种具有代表性的分析和再分析格点降水数据与全国621站的降水数据进行了对比分析。结果表明,7种逐月降水资料在不同地区的适用性不同,用格点资料研究西部地区的逐月降水情况会导致较大误差。中国西部部分地区多年平均偏差在50%以上,累计误差数超过总样本数的30%。中国东部地区平均偏差在10%左右,累计误差数不超过总样本数的30%。利用卫星反演的降水比模式计算的降水适用性更好,局地主要降水系统类型和台站数据的不足是影响各降水数据适用性的主要原因;利用卫星反演和数值模式计算的降水都会导致降水数据中的系统偏差和正态分布,与降水资料的非正态性不符。未来的降水产品应该是综合台站观测数据、卫星反演产品和模式计算结果各自的优点,互补不足,以提高降水产品的适用性和应用潜力。更多还原

【Abstract】 The analysis and reanalysis precipitation data could compensate the shortage of surface observational data.This is especially important for the relevant meteorology and hydrology researches.In this paper,7 kinds of analysis and reanalysis grid-precipitation data were compared and analyzed to the real precipitation data collected from 621 surface observation stations in the China’s mainland.The results showed that the applicability was distinctive in different regions.In western part of China’s mainland,the bias between grid-precipitation data and surface observation was above 50% and accumulated sampling error ratio was beyond 30%.In eastern part of China’s mainland,the bias between grid-precipitation data and observations was about 10% and accumulated sampling error ratio was below 30%.In general,the products retrieved from satellites remote sensing data performed better than that calculated by using numerical models because of the types of local raining system and the shortage of observation data.It was proved that both numerical models and inversion methods would cause system errors and deviated the normal distribution in reanalysis grid-precipitation data.In the future,the surface observation data,satellite remote sensing technology and numerical models should be synthesized to improve the applicability and potential of the grid-precipitation data.更多还原