中国北方冬季降水的多源资料产品评估和融合优化

Assessment and merged optimization of multi-source winter precipitation products over northern China

为了考察不同来源降水产品在中国北方冬季(特别是固态降水)的精度和可用性,优化融合降水产品质量,利用2019年12月—2020年2月美国CMORPH和IMERGE卫星反演降水、日本GSMaP、中国气象局雷达定量估测降水(MOC-QPE)、CMAMESO模式预报以及地面观测插值等不同来源分析的降水产品,以地面站观测逐小时降水量数据为基准,从KGE评分、相关系数、平均误差和均方根误差等精度统计指标以及命中率(FOD)、虚警率(FAR)和TS评分等降水事件发生角度开展评估,结果表明:中国区域单源降水产品中地面插值分析产品对冬季降水描述精度最高也最稳定,但存在明显的系统偏低;其次是MOCQPE和IMERG卫星产品,对中国北方偏南部地区的降水有一定的描述能力,但对北方高纬度地区固态降水的反映能力较差;卫星产品中IMERG精度最高,CMORPH则基本没有反演能力;CMA-MSEO模式产品虽然误差较大但与地面站观测的降水特别是固态降水存在较高相关,明显优于雷达和IMERG、GSMaP等卫星产品。采用BMA技术融合雷达、模式、卫星降水形成优化背景场,评估逐步引入不同的数据源对融合降水在冬季的精度影响,引入IMERG卫星和CMA-MESO模式产品均能提升高分辨率融合产品的质量,其中模式产品的改进效果最显著。

Based on the data between December 2019 and February 2020, multi-source precipitation products such as CMORPH,IMERG, GSMaP satellite retrieved precipitation, quanity precipitation estimate(MOC-QPE), CMA-MESO model forecast, and grid analysis product based on gauges etc. are assessed. The statistical error indicates that the grid analysis dataset based on gauges is the best and robust among all the precipitation products. The MOC-QPE and IMERG are better in the southern area of northern China but worse in high-latitude area, especially for solid precipitation. IMERG is the best one of satellite products, while CMORPH is the worst in winter over northern China. CMA-MESO product has a high correlation with gauged observations, higher than satellite and radar products, and is valuable for merge analysis. Based on BMA method, the MOC-QPE, IMERG, and CMA-MESO are gradually added to and analyzed in merged products. Results indicate that both IMERG and CMA-MESO can improve the winter precipitation,but CMA-MESO has the largest contribution to the enhancement of the merge analysis accuracy.