高分辨率卫星反演降水数据在四川省的适用性分析

Analysis of the Suitability of High Resolution Satellite Inversion Precipitation Data in Sichuan Province

摘要: 本文利用2013~2015年四川省814个自动站的逐小时降水观测数据对同期CMORPH、TRMM和PERSIANN卫星反演降水估计产品进行了检验和评估。经过对比分析得出：CMORPH在各个时间尺度上对降水的探测都是最准确的，并且CMORPH所记录的逐日降水量与地面雨量计的值变化趋势一致，特别是在逐月尺度上，CMORPH与雨量计数据相关性很高，说明该数据具有良好的探测四川省降水情况的能力。TRMM对三种时间尺度降水的记录不如CMORPH准确，但仍具有一定的探测能力，例如TRMM的逐日降水变化趋势与雨量计所记录的情况类似。PERSIANN数据是三种降水产品中探测效果最差的，在三种时间尺度上均高估降水量，而且在逐日降水变化趋势上也和真实情况相差甚远。

Abstract: The hourly precipitation observation data from 814 rain gauge records are used to verify and evaluate the satellite precipitation estimation products of Climate Prediction Center Morphing Technique (CMORPH), Tropical Rainfall Measuring Mission (TRMM) 3B42 and Precipitation estimation from remotely sensed information using artificial neural networks (PERSIANN) during 2013-2015 in Sichuan Province. The results show that the CMORPH has the most accurate result for detecting precipitation on all time scales, and the daily precipitation recorded by CMORPH is consistent with the value of ground rainfall gauges, especially on the monthly scale. The correlation between CMORPH and rain gauge data is very high, indicating that the CMORPH data has a good ability to detect precipitation in Sichuan Province. The TRMM records of three time-scale precipitations are not as accurate as CMORPH, but they still have certain detection capabilities. For example, the daily precipitation trend of TRMM is similar to that recorded by rain gauges. The PERSIANN data is the worst among the three types of precipitation products. It overestimates the precipitation on all three time scales, and it is far from the real situation in the daily precipitation variation trend.