湖泊提供重要的生产生活用水，在维系生态安全方面发挥显著的价值。中国地理环境复杂多样，湖泊类型各异，不同湖区间的湖泊时空变化特征及驱动机制复杂。两次全国湖泊调查是国家尺度的湖泊变化监测的开创性工作，近年来伴随遥感数据源的海量积累和湖泊遥感提取方法的快速发展，大尺度湖泊水体变化遥感监测研究取得了长足的发展。然而，由于人工判读、湖泊类型判别依据和遥感数据源时相的差异，湖泊水体遥感提取结果及面积估算存在不同程度的出入。首先，本文总结了现有全国尺度湖泊监测研究的遥感数据源、方法和结果，分析了不同研究分析结果差异的可能原因。其次，本文基于1980s—2010s时段Landsat影像制图的湖泊水淹频率数据集的最大水域范围信息，提取并生成了全国湖泊新编目，以其作为该时 段内湖泊判定和水体变化监测空间约束的统一参照，避免了因遥感数据时相与年际、年内湖泊水域范围波动对遥感提取结果及湖泊面积计算带来的差异。最后，本文从湖泊水淹范围频度的统计角度，初次提出了可体现湖泊面积年内与年际变化情况的统计指标——概率等效面积，用于监测和指示湖泊水域动态。结果表明，近30年来，中国有3741个最大水域范围大于1 km²的湖泊，总面积约93723 km²，其中青藏高原湖区的湖泊数量约占全国三分之一，总面积超过了全国的一半。全国湖泊总面积呈显著上升趋势，不同湖区的变化趋势差异明显：青藏高原湖区和新疆湖区湖泊面积显著上升，东部平原湖区、内蒙古高原湖区和云贵高原湖区显著下降，东北平原与山地湖区波动式变化、趋势不显著。研究结果总体表明中国湖泊水资源的时空不平衡状况在加剧。

Lakes play an important role in supplying water resources and sustaining human living. The broad geographical extent and complex environment of China result in the different changing patterns and complicated driving factors of lakes. The two-epoch national lake surveys provide the groundbreaking knowledge of lake distribution and change characteristics in China. Thus, an increasing number of researchers have conducted the national-scale remote sensing monitoring of lake changes with the rapid development of satellite techniques in recent years. However, evident discrepancies exist in the results and conclusions among these prior studies due to differences in level of manual quality control, lake mapping method, and the acquisition time (year and month) of used remote sensing images. Thus, comprehensively investigating their differences and causes and producing an updated lake inventory are required.We reviewed and compared the data sources, methods, and results of the existing works on national-scale remotely sensed lake changes and analyzed the causes of such differences. We also produced a new national lake inventory by interpreting the maximum water inundation area (1980—2010), which can minimize the differences of manual interpretation and lake mapping in different time periods. With the spatial constrain of maximum water extent for each lake from the new inventory, this study developed a novel approach, namely, the Probability Equivalent Area method, to detect the long-term change trajectory of China’s lakes during 1980—2010.A total of 3741 lakes in China had a maximum area exceeding 1 km² in the past 30 years. The Qinghai-Tibet Plateau lake zone accounted for approximately one-third in count and half in area of the inventoried lakes in China. The total area of China’s lakes showed an overall upward trend, but it had strong spatial heterogeneity. The Qinghai-Tibet Plateau and Xinjiang lake zones significantly increased lake areas, while the Eastern Plain, Inner Mongolia Plateau, and Yunnan-Guizhou Plateau lake zones significantly decreased lake areas. The area changes in the Northeast Plain and Mountain lake zone were statistically insignificant.This study suggested that the main causes of different results in previous remote sensing research of China’s lakes include inconsistent timing of satellite imagery, the manual interpretation standards for lakes and reservoirs, and quality assurance procedures. To produce a more accurate estimation of lake area, a new inventory dataset, which minimized interpretation differences and provided a unified spatiotemporal constraint, was proposed on the basis of our findings. A statistics-based method named the Probability Equivalent Area, which eliminated differences in satellite imagery selection, was also proposed. The lake area changing patterns and distribution of lake water resources in different lake zones are obviously imbalanced in China. China’s sparsely populated areas are generally experiencing a dramatic increase in lake area. However, the densely populated areas are experiencing a significant decline in lake area. This situation may further increase the imbalance of water resources per capita in China. Effective measures should be taken to slow down this trajectory.