中文摘要：

在全球气候变暖的背景下湖泊变化已是国内外地学界关注的环境问题，并日益受到各国政府和研究机构的重视。本文以长江中下游的6个典型湖泊为研究对象，基于2003年-2009年ICESat—GLAS测高数据，采用高斯滤波法对ICESat—GLAS测高数据所对应的回波进行去噪后，用于识别水体与陆地；采用融合算法从LandsatTM影像数据提取湖泊边界；提取湖泊边界内的ICESat—GLAS水位观测值构建了湖泊的年际和年内水位变化，进一步分析了6个湖泊的水位变化时空特征与原因。结果表明：从LandsatTM中提取湖泊边界的精度达到10m，ICESat—GLAS测高数据推算出洪泽湖蒋坝站、洞庭湖城陵矾站、鄱阳湖星子站和都昌站、太湖西山观测站、巢湖闸站和高邮湖高邮站的ICESat测高数据验证的平均精度分别为-3．9cm、-8．3cm、-9．0cm、-2．5cm、-3．1cm和-4．1cm；在整个观测期间，长江中下游流域的湖泊水位呈周期性波动，但较历史观测资料，湖泊平均水位有所降低，且自2006年10月以后三峡工程对洞庭湖的水位调控明显。

英文摘要：

Changes in lakes because of global climate change has become a national and international issue. We selected six typical lakes in China and analysed ICESat-GLAS data （ice, cloud, and land elevation satellite-geoscience laser altimeter system） to extract lake water elevations for 2003-2009. ICESat-GLAS altimeter waveform data were denoised by a Gaussian filter and used to distinguish between water and land. Lake boundaries were extracted from pre-processed Landsat TM imageries using a Fusion algorithm. Water level values within lake boundaries were used to construct inter-annual and annual water level time-variation curves. We found that the spatial accuracy of lake boundary data extracted from Landsat TM imageries is 10 m. The mean accuracy of measured ICESat altimeter data in the vicinity of Jiangba station at Hongze Lake, Chenglingji station at Dongting Lake, Xingzi and Changdu stations at Poyang Lake, Xishan station at Tai Lake, Zha station at Chao Lake and Gaoyou station at Gaoyou Lake are validated and assessed by observation data, and mean accuracy was -3.9 cm, -8.3 cm, -9.0 cm, -2.5 cm,-3.1 cm and-4.1 cm, respectively. Lake water levels in the middle and lower reaches of the Yangtze River basin show periodic fluctuation, and compared with historical observation data, the mean water level has decreased slightly. Lake water levels have been affected by regulation of the Three Gorges Project since October 2006. The relationship between lake water levels in the middle and lower Yangtze River basin and annual precipitation was correlated. Average water level changes for Hongze and Tai Lakes in 2006 fluctuated significantly. The average water level of Dongting, Poyang, Chaohu and Gaoyou Lakes peaked from June to October 2006.