Monitoring the water level and volume changes of lakes and reservoirs is essential for deepening our understanding of the temporal and spatial dynamics of water resources in the Yellow River Basin, with a view to better utilizing and managing water resources. In recent years, there have been many studies on monitoring water level and volume changes in inland waters, but they were mainly focused on radar altimetry and the full waveform LiDAR ICESat, which was retired in 2010. Few studies based on the latest photon-counting LiDAR ICESat-2 have been reported. Compared with previous sensors, ICESat-2 has great advantages in footprint size, transmitting frequency, pulse number, etc, but its performance in monitoring water level and volume changes in inland waters has not been fully explored. Here we investigated the spatial distribution of water level and volume changes of 11 lakes and 8 reservoirs in the Yellow River Basin based on ICESat-2 and Google Earth Engine, and analyzed the factors affecting the measurement uncertainties. In-situ validation of lake level in Lake Qinghai indicates that the Root Mean Square Error (RMSE) of our result is only 7 cm after the reference coordinate system conversion. We found that the water level trend of the natural lake shows significant seasonal variations, while the water level trend of the reservoir shows a sharp rise and fall. In addition, precipitation plays a decisive role in the changes in natural lake levels and indirectly affects the artificial control of reservoirs’ water discharges. The uncertainty of water volume change monitoring is mainly affected by water level measurement uncertainty for lakes, while for reservoirs, that is affected by the combination of water level and area measurement uncertainties. The stability of lake level measurement increases with the increase in photon counts. The introduction of ICESat-2 ATL13 Significant Wave Height might lead larger standard deviation in water level measurement. According to the law of propagation of uncertainty, the uncertainty of the water volume change estimation by the combination of ICESat-2 and GEE is less than 9 %.

监测湖库水位和水量变化，对于加深对黄河流域水资源时空动态的认识，更好地利用和管理水资源具有重要意义。近年来，关于监测内陆水域水位和水量变化的研究很多，但主要集中在雷达测高和2010年退役的全波形LiDAR ICESat。基于最新光子计数的研究很少LiDAR ICESat-2 已被报道。与以往的传感器相比，ICESat-2在足迹尺寸、发射频率、脉冲数等方面具有很大优势，但其在监测内陆水域水位和体积变化方面的性能尚未得到充分挖掘。本文基于ICESat-2和Google Earth Engine对黄河流域11个湖泊、8个水库水位和水量变化的空间分布进行了调查，分析了影响测量不确定性的因素。青海湖湖水位现场验证表明，我们的结果在参考坐标系转换后的均方根误差（RMSE）仅为7 cm。我们发现，天然湖泊水位趋势呈现明显的季节变化，而水库水位趋势呈现急剧上升和下降。此外，降水对天然湖泊水位变化起决定性作用，间接影响水库放水的人工调控。水量变化监测的不确定性主要受湖泊水位测量不确定性的影响，而水库则受水位和面积测量不确定性的综合影响。湖水位测量的稳定性随着光子数的增加而增加。ICESat-2 ATL13 有效波高的引入可能会导致水位测量的标准偏差更大。根据不确定性传播规律，ICESat-2与GEE联合估计水量变化的不确定性小于9%。ICESat-2 ATL13 有效波高的引入可能会导致水位测量的标准偏差更大。根据不确定性传播规律，ICESat-2与GEE联合估计水量变化的不确定性小于9%。ICESat-2 ATL13 有效波高的引入可能会导致水位测量的标准偏差更大。根据不确定性传播规律，ICESat-2与GEE联合估计水量变化的不确定性小于9%。