摘要
利用江西省2015年4月至2016年3月水文站观测降水数据,在小时尺度上,对中国国家气象信息中心研制的5和10 km融合降水产品进行质量评估,同时与美国国家海洋大气局(NOAA)气候预测中心卫星反演降水产品(CMORPH)、中国国家气象信息中心研制的东亚区域多卫星集成降水产品(EMSIP)两套卫星降水产品进行对比评估。研究分析各类降水产品的数据误差及其时空变化规律,验证融合降水产品在特征区域的适用性。研究结果表明:融合降水和卫星降水均能较好地反映年内小时降水的变化趋势,与水文站观测降水相比,四套降水资料均存在一定程度低估,其中卫星降水产品低估较大。融合降水产品的数据质量较高,其中5 km融合降水产品的数据精度(R=0.81,RMSE=2.12 mm·h^(-1),RE=-5.4%)基本优于10 km融合降水产品(R=0.78,RMSE=2.3 mm·h^(-1),RE=-5.1%),卫星降水产品与水文站观测降水存在较大的偏差,CMORPH和EMSIP的相关系数分别仅为0.19和0.24。各降水产品误差具有相同的月变化趋势,融合降水产品的误差变化幅度明显要小于卫星降水产品。四套降水产品的相关性随着降水量级增大而增加,融合降水产品能够准确反映降水的空间结构和中心位置,5 km融合降水产品对强降水的监测能力更具有优势。
In this paper, the quality of 5 km and 10 km merged precipitation developed by China National Meteorological Information Centre is assessed based on the hydrological observed precipitation. Besides the U.S. National Oceanic and Atmospheric Administration （NOAA） climate prediction center satellite pre- cipitation products （CMORPH） and the East Asian multi-satellite integrated precipitation （EMSIP） developed by National Meteorological Information Centre are also assessed by comparing the two merged precipitation products. The data error of each precipitation product and its spatio-temporal variation characteristic are analyzed to verify the applicability of merged precipitation product in this selected region. The results show that the merged precipitation and satellite precipitation can well reflect the variation tendency of annual hourly precipitation. Four precipitation products all have a certain underestimate compared with the hydrological observed precipitation, and the underestimate of the satellite precipitation is larger. Merged precipitation has better quality, and the accuracy of the 5 km merged precipitation （R=0.81, RMSE= 2. 12 mm - h-1, RE=-5.4%） is better than the 10 km merged precipitation （R=0.78, RMSE=2.3 mm · h -, RE=- 5.1%）. There is a great difference between the satellite precipitation and the hydrological observed precipitation. The correlation coefficient of CMORPH and EMSIP precipitation products is only 0.19 and 0.24, respectively. The error of each precipitation product has the same monthly variation tend ency, and the error variation range of the merged precipitation product is obviously smaller than that of the satellite precipitation product. Besides, the quality variation of the four precipitation products in different rainfall level is also discussed, which shows that the correlation of the four precipitation products increases with the increase of precipitation. Merged precipitation products can accurately show the spatial structure and the central location of precipitation, 5 km merged precipitation takes great advantage in monitoring the severe precipitation events.
出处
《气象》
CSCD
北大核心
2017年第12期1534-1546,共13页
Meteorological Monthly
基金
国家自然科学基金项目(91437220)
公益性行业(气象)科研专项(GYHY201506002和GYHY201306045)
中国气象局"气象资料质量控制及多源数据融合与再分析"创新工程项目共同资助
关键词
融合降水产品
卫星降水产品
水文站观测降水
质量评估
merged precipitation product, satellite precipitation product, hydrological observed precipitati-on, quality assessment
