GPM卫星降水数据在黑河流域的适用性评价

论文分别从时间和空间两个角度对GPM卫星降水数据在黑河流域适用性进行评价。以TRMM卫星的降水产品和气象观测站点实测的降水数据作为参考,选用相关系数（R）、均方根误差（RMSE）、相对误差（BIAS）、平均误差（ME）以及标准偏差（SD）多种统计分析指标进行精度评价。结果表明：1）GPM卫星降水数据具有与观测数据较好的一致性以及良好的时间模式;2）GPM卫星降水数据与观测数据之间有着较高的相关系数（R>0.72）以及较低的误差范围 （-0.59%~1.62%）;3）相比TRMM,GPM具有较低的均方根误差（RMSE<28.76）,尤其是对夏季降水,GPM具有相对更高的相关系数（R>0.76）和更小的相对误差（BIAS<1.14）;4）卫星估测降水量的精度与高程具有明显的相关性。总的来说,GPM卫星降水数据在地形复杂的干旱半干旱区域也能够具有较好的精度,能够较为准确地反映干旱半干旱地区的降水分布。

Evaluating the Applicability of GPM Satellite Precipitation Datain Heihe River Basin

The precipitation varies across space and time. The estimation of precipitation is thus difficult to achieve high precision, leading to the significant difference and enormous uncertainties when applying precipitation products in different areas. The accuracy could be influenced by terrain or climatic factors. Therefore, it is necessary to verify its suitability and accuracy before using GPM data for precipitation monitoring. In this paper, the applicability of GPM satellite precipitation data in Heihe River Basin (HRB) is evaluated at both spatial and temporal scales. Based on the precipitation data of the TRMM satellite and the precipitation data measured at the meteorological observation sites, the correlation coefficient (R), root mean square error (RMSE), relative bias (BIAS), mean error (ME) and standard deviation (SD) were selected to evaluate the accuracy of satellite data. The results demonstrate that the monthly GPM data can capture the precipitation well. It shows a good consistency with the measured data and reveals a good temporal pattern. There is a high correlation between the GPM data and the observed data (R> 0.72) and a low error range (-0.59% - 1.62%). Comparing with TRMM, GPM produces lower RMS error (RMSE <28.76). In summer, GPM performs better, with higher correlation coefficient (R> 0.76) and lower error (BIAS <1.14). Both estimated precipitation and measured precipitation decline from southeast to northwest. The elevation was found to have a significant impact on the precipitation monitoring accuracy. In general, the GPM satellite precipitation data are applicable in the arid and semi-arid areas with complex terrain and can accurately reflect the precipitation distribution, bearing in mind that further improvement is still needed.