遥感降水产品在中国不同气候区的适用性研究

遥感降水产品相对于气象站观测数据能够更好地反映降水的空间分布特征,对其进行不同气候区上的差异性评价对数据产品选择和遥感降水反演算法改进均有重大意义。本文选择中国典型气候区（干旱区、过渡区、湿润区和青藏高原地区）,以649个经偏差矫正后的气象站降水数据为标准,评估了5种国际常用的遥感降水产品 (CHIRPS v2.0, CMORPH v1.0, MSWEP v2.0, PERSIANN-CDR, TRMM 3B42v7) 在中国典型气候区的适用性。研究发现,各产品的性能存在空间差异性。MSWEP在各气候区的相关系数(CC),Kling-Gupta efficiency（KGE）,均方根误差（RMSE）等基本统计性能指标均优于其他4种产品。相对偏差（BIAS）方面,在干旱区、湿润区、青藏高原、过渡区表现较优越的产品分别为MSWEP、CHIRPS、PERSIANN、TRMM。在评估遥感降水产品对降水事件发生概率的估算能力方面,选择了误报率（FAR,降水事件预报错误的比例）、命中率（POD,降水事件预报正确的比例）、关键成功指数（CSI,降水事件正确预报综合性能）、精度指数（ACC,等级预报综合性能）和降水等级概率分布（PDF）5个指标作为评估依据,结果表明,就POD,CSI和ACC而言,在各气候区MSWEP表现明显优于其他产品;对于FAR,TRMM和CMORPH产品在湿润区表现优越,其余气候区仍以MSWEP表现较为优越;就PDF而言,PERSIANN和MSWEP产品对1~20 mm的日降水量的估算偏高,特别是MSWEP在青藏高原和湿润区对小雨的估算频率明显偏高, MSWEP有待在该区提高频率预报的能力。综合而言,多源数据融合的MSWEP在各气候区的基本统计性能和降水等级性能较好,可作为可靠的降水数据源用于中国水文气象研究,同时也表明多源数据融合产品具有良好的应用前景。

Research on the Applicability of Remote Sensing Precipitation Products in Different Climatic Regions of China

Compared with the observation data from meteorological stations, remote sensing precipitation products can be used to well present the spatial distribution of precipitation. So it is of great significance to make a comparative study on the differences of remote sensing precipitation products in different climatic regions. This paper selects the typical climatic regions in China for comparison analysis. Based on the corrected observations of 649 meteorological stations, the performance of five typical remote sensing precipitation products (CHIRPS v2.0, CMORPH v1.0, MSWEP v2.0, PERSIANN-CDR, and TRMM 3B42v7) are evaluated in different climatic regions. Results show that: (1) The performance of each product varies in different climatic regions; (2) The MSWEP product shows higher CC, KGE, and RMSE values in each climatic region. Based on BIAS, the MSWEP, CHIRPS, PERSIANN, and TRMM have better performance in arid, humid regions, the Tibetan Plateau, and transition regions, respectively; (3) In terms of POD, CSI, and ACC, the MSWEP has better performance in all climatic regions. In humid region, TRMM and CMORPH have advantages with lower FAR, while MSWEP is better in other climatic regions. Above all, MSWEP, the multi-source precipitation product, has better basic statistical performance and category performance in all climatic regions and can be used as a reliable precipitation data source for various hydrometeorological studies in China, which indicates that multi-source data fusion has a good application prospect in future.