基于同化数据的GRACE卫星反演中国陆地水储量变化降尺度分析

利用GRACE重力卫星反演得到陆地水储量变化受数据低空间分辨率的限制,目前仅应用于大尺度的空间范围内,为此提出一种简易的降尺度方法将GRACE陆地水储量变化精度提高了16倍。借助全球水文模型模拟的陆地水储量变化数据,在中国九大流域上进行分区对比分析,发现高分辨率数据基本保持了原始数据的时空变化趋势,同时又增加了水文模型模拟的水储量变化信息。此外,利用全国455个地下水位监测站的水位数据,验证了降尺度后的陆地水储量变化数据,可知高分辨率陆地水储量变化数据与实测地下水变化值的相关系数最高达0.847,普遍高于低分辨率GRACE陆地水储量变化值与实测地下水变化值的相关系数,证明降尺度后的数据更加准确有效。进而基于降尺度数据,利用EOF方法分析了2005年1月~2012年12月中国各地陆地水储量变化的时空变化特征,即在总体趋势上,长江流域及以南地区春夏两季的陆地水储量有明显增加,华北地区的地下水超采及南方流域的极端气候事件为影响该区域陆地水储量异常的重要因素。

Downscaling Analysis of GRACE Terrestrial Water Storage Changes Based on GLADS in China

GRACE-based terrestrial water storage change (TWC) has been widely applied in many studies. Limited by the low spatial resolution of its data, it can only be applied in the large-scale spatial area. Therefore, a simple downscaling method is proposed to downscale the TWC derived from GRACE to the high spatial resolution. Subsequently, the downscaled TWC is firstly compared with the TWC calculated from GLADS in terms of the change trend and the spatial distribution in nine major river basins of China. In addition, the correlation coefficient between the GRACE TWC and the measured groundwater change is also calculated to validate the accuracy of the downscaled TWC. The results show that the correlation coefficient between the downscaled TWC and the measured groundwater change is up to 0.847, which is generally higher than the value calculated by the original GRACE TWC. Finally, EOF method is used to analyze the temporal and spatial variation characteristics of TWC from January 2005 to December 2012 in China based on the downscaled TWC data. The spring and summer terrestrial water storage changes in the Yangtze River Basin increase obviously. The underground water overexploitation in north China and extreme climate events in south China are important factors for affecting the abnormal terrestrial water storage in this area.