中文摘要：

基于1998—2008年Spot—Vegetation归一化植被指数（NDVI）数据集，采用遥感与地理信息技术以及最大值合成法获取了石羊河流域下游植被覆盖的时空变化特征，利用同期气象数据和地下水埋深数据，构建Matlab多元回归模型，综合分析了NDVI与地下水埋深、年降水量和年均气温的响应关系，平均相对误差为5．83％．研究发现11年来石羊河流域下游植被覆盖整体上经历了1998-2002年持续增加期、2003-2004年短暂下降期及2005-2008年稳步回升期，植被改善的区域面积大于植被退化的区域面积，植被覆盖整体呈现增加趋势．研究区NDVI与年降水量和年均温呈现显著相关，且年降水量对NDVI的影响大于年均气温，而NDVI与地下水埋深的相关性不显著．研究区植被覆盖受年降水量的影响相对于年均气温和地下水埋深较大，且与当地退耕还林政策和关井压田政策的实施关系密切．

英文摘要：

Based on the vegetation index data sets of Spot-Vegetation NDVI, RS/GIS techniques and maximum value composite were adopted in the paper to get the spatiotemporal variations of the vegetation cover in the lower reaches of Shiyang River basin. The contemporaneous meteorological data and groundwater depth data were used in to build a multiple regression model and to comprehensively analyze the response relationship between NDVI, groundwater depth and climatic factors. An average relative error of 5.83% was got eventually. The study found that the vegetation cover in the lower reaches of Shiyang River basin in eleven years experienced three phases on the whole： a continuous increase from 1998 to 2002, transient decrease from 2003 to 2004 and unfaltering recovery from 2005 to 2008. The area of vegetation improvement exceeded that of degradation in the time period, indicating that the vegetation cover degree had presented an increasing tendency on the whole. Furthermore, the NDVI in the study area exhibited no remarkable relevance with groundwater depth, while correlating significantly with the annual precipitation and annual mean temperature respectively. It turned out that the influence of annual precipitation on the vegetation in study area was greater than that of the annual mean temperature and groundwater depth. Moreover, the implementation of the policy of returning farmland to forest and of closing driven wells and reducing arable land had a close relationship with the variations of vegetation cover in the study area.