CMPA降水资料在中国地区不同地形下的精度评价研究

Precision evaluation of CMPA remote sensing precipitation data in different terrains of China

以2001—2010年中国地面自动站降水资料为基准,对中国大陆范围内CMPA(CMPA_Daily)降水资料进行精度评价研究,并与CMORPH1.0(CPC MORPHing technique gauge-satellite)、TRMM3B43V7(Tropical Rainfall Measuring Mission 3B43)降水资料精度进行对比,并进一步结合高程、坡度、坡向、坡向修正因子分析地形因子对数据质量的影响并探讨不同地貌类型下数据的精度。结果显示:CMPA在年、月尺度上均能较好地反映降水的多寡,与站点实测数据具有较高的相关性,误差波动较为平稳,数据质量及稳定性优于CMORPH与TRMM;从时间序列曲线显示CMPA的精度呈现较为明显的季节性差异,均方根误差夏季高于冬季,相关系数、百分比偏差、平均相对误差冬季高于夏季,总体而言CMPA夏季的误差高于冬季,由于夏季降水的基数大而导致了百分比偏差以及平均相对误差较低;分析地形的影响表明,高程、坡度对数据质量的影响大于坡向与坡向修正因子;在复杂地形下,高海拔与高坡度地区CMPA精度均有所降低,但降水资料的精度仍然优于CMORPH与TRMM。

The accuracy of CMPA remote sensing precif)itation data in China from 2001 to 2010 was evaluated using the ground rainfall observation data as the reference data,and compared witli CMORPH 1.0(CPC MORPHing Technique Gauge-satellite)and TRMM3B43V7"Tropical Rainfall Measuring Mission 3B43).In addition,combined with elevation,slope,aspect,correction factor to analyze the impact of topographic factors on data quality,the accuracy of the data in different terrains was also explored.The result shows that CMPA remote sensing precipitation data can reflect the distribution of precipitation in China on the year and month scale,with a high correlation with ground rainfall observation data and a stable fluctuation,which performs better than CMORPH and TRMM.Time series curve shows the accuracy of CMPA fluctuates in the period of year;RMSE in summer is higher than that in winter;correlation factor,percent eeor and MRE in winter is higher than that in summer.In general,the eeor of CMPA in summer is higher than that in winter,and the low percent error and MRE are caused by the large base of summer precipitation.Terrain performance analysis shows that the influence of elevation and slope on data quality is greater tlian that of asj^ect and correction factor.In complex terrain,the accuracy of CMPA slightly decreases is high altitude and high slope areas,but the performance of reflecting the distribution of precipitation is still stronger than that of CMORPH and TRMM.