Comparison of different precipitation datasets in Lancang-Mekong River basin
Comparison of different precipitation datasets in Lancang-Mekong River basin
复制标题
澜沧江-湄公河流域不同降水数据集比较
DOI：
--
复制
发表时间：
2011
期刊：
影响因子：
--
通讯作者：
Xiao Zi
中科院分区：
文献类型：
--
作者：
Xiao Zi
澜沧江-湄公河，在中国被称为澜沧江，发源于青藏高原海拔约4500米的地方。在穿越6个国家后，它通过越南宽阔而泥泞的三角洲流入南中国海。根据径流量，澜沧江-湄公河是世界第十大水系，也是该地区最重要的国际河流，使近80万平方公里、约6000万人的流域干涸，随着全球气候变暖和澜沧江经济合作等区域经济合作行动的快速发展，湄公河次区域近几十年来，无论是气候变化还是经济变化对该流域水资源系统的影响，已成为气象学、经济学、为了探讨气候变化对水文水生态影响的基本事实，要研究澜沧江-湄公河流域的水环境变化，首先要了解流域水循环中最重要的组成部分--降水的变化规律。分析比较了澜沧江流域及澜沧江全境几种不同网格化降水资料和观测资料的年、季平均降水量变化的时空特征，结果表明，不同来源的降水资料的均值和变率存在很大差异，但在1979年以后均呈上升趋势，在1979年以后呈上升趋势。通过对比分析发现，在澜沧江流域，用Xie-Arkin和CRU降水资料分析年降水量序列比用其他降水资料更具有适应性，而对于季节降水量变化，春、秋两季的降水分析宜选用Xie-Arkin降水资料，而夏、冬两季的降水分析，CRU资料的结果更接近观测结果。各数据集的降水量均呈现出由上游向下游逐渐增加的趋势，CRU降水数据由于其高分辨率，能够捕捉到更多的空间分布细节，空间分辨率。经过相关性分析，年径流量与年降水量的相关性很好，相关系数大于0.7，远远超过了99%的可靠性检验水平（46个样本）在此基础上，建立了黄河上游九洲站和下游云景洪站降水量与径流量的一元线性回归方程，在澜沧江流域，仅用降水量资料，利用上述一元线性回归模型就能提供较好的径流预测结果。
The Lancang-Mekong River,known in China as the Lancangjiang,begins its course at an altitude of approximately 4 500 meters in the Qinghai-Tibet Plateau.After traveling some 4 500 kilometers across six countries it empties into the South China Sea through its wide and muddy delta in Vietnam.Measured by runoff,the Lancang-Mekong river is the tenth largest river system in the world and the most important international river in this area,draining a watershed of nearly 800 000 square kilometers and some 60 million people,many of whom depend directly on the river for their livelihood.With the global warming and the fast development of the regional cooperation actions in economy such as the economic cooperation in Lancang-Mekong river subregion in last several decades,the changes,either from climate or from economy influencing the water resource system in this drainage basin have become the focus of the attention and hotspot of research from meteorology,economy,hydrology and other fields internationally.In order to investigate the basic facts which the climate change influence on the hydrological and water ecological environment variations of the Lancang-Mekong river basin,we should understand the regularity of precipitation in this region firstly,which is the most important component in the water cycle.In this paper,we analyzed and compared temporal and spatial characteristics of changes of annual and seasonal mean precipitation derived from several different gridded precipitation data and observation data in the Lancang River basin and the whole Lancang-Mekong River basin.The gridded precipitation datasets include GPCP data,CMAP data,CRU data,Xie-Arkin data,which are all datasets recongnized by most researchers from several fields.The results show that there is a lot of differences in the mean value and variable ratios of precipitation data from different sources,but they all showed an increasing trend after 1979,which is consistent in both the Lancang-Mekong River basin and the Lancang River basin.By the comparing analysis,we found that in the Lancang River basin,using the Xie-Arkin and CRU precipitation data to analyze the annual precipitation series was more adaptive than using the other precipitation data.While for seasonal precipitation changes,we should choose Xie-Arkin precipitation data to analysis the spring and autumn precipitation,and for the summer and winter precipitation,the results from CRU data were more closer to the observation results.For the spatial distribution patterns,the precipitation from various datasets could all show the increasing trend from the upstream to the downstream.CRU precipitation data could catch more detail of spatial distribution than other precipitation data because of his high spatial resolution.After correlation analysis,we found there was a very good correlation between annual runoff and annual precipitation which far exceeded the 99% level for reliability test with the correlation coefficient greater than 0.7(46 samples).Basing on the relationship we established the unary linear regression equation between the observation precipitation data and runoff of the Jiuzhou station upstream and Yunjinghong station downstream in Lancang River,we can provide a good runoff prediction by using the above unary linear regression model only with precipitation data.