中文摘要：

利用青藏高原77个地面台站的2003年~2009年夏季(6月~9月)的降水资料,对月尺度和年尺度上CMORPH(Climate Prediction Center morphing)多卫星降水数据的精度进行研究,并引入Sokol模型对年尺度上的CMORPH数据进行修正,旨在为基于卫星降水数据的青藏高原地区气候、水文等方面的研究提供科学依据。研究结果表明:1CMORPH数据对青藏高原降水的时间变化趋势和空间变化趋势的模拟精度较低,且存在明显的时空不稳定性。2不同时间尺度的CMORPH数据在青藏高原东南部的模拟精度要高于其他地区,而喜马拉雅山脉北麓以及青藏高原东北部的模拟精度最低。3CMORPH年数据存在明显的高值高估、低值低估的现象,其模拟值与误差之间的相关系数均在0.53以上。4经过Sokol模型修正后,CMORPH年数据均方根误差明显降低,而相关系数均有不同程度的提高,表明该模型能够提高CMORPH数据对青藏高原地区降水的模拟精度。

英文摘要：

The performance of CMORPH (Climate Prediction Center morphing)satellite precipitation estimates on Tibetan Plateau in month scale and year scale was assessed by compared it with ground observations from 77 meteorological stations in the summer (June,July and August)between 2003 and 2009,and a model called Sokol was introduced to modify the year-estimates of CMORPH.The result showed that:(1 )The CMORPH could not precisely simulate the spatial pattern and the temporal variability,and the precision of CMOPRH presented obvious spatial and temporal instability.(2)The performance of CMOPRH in southeast part of Tibetan Plateau is much better than other parts,while along the north of Himalayas and the northeast part of this area the performance of CMORPH is much poorer.(3)The CMORPH yearly estimates presented over-estimate with high value and under-estimate low value,the correlation coefficients between the CMORPH yearly estimates and the errors of it are all above 0.53 between 2003 and 2009.(4)After corrected by Sokol model,the RMSE (Root Mean Square Error)of CMORPH yearly estimates show obviously decrease,while the correlation coefficients increased in different extent.It means that Sokol model is suitable to correct the CMORPH yearly estimates.This research aims to provide scientific basis to the correction and use of CMORPH to study the precipitation,climate and other subjects over Tibetan Plateau.