基于GRACE卫星和GLDAS系统的地下水水位估算模型——以和田地区克里雅河流域为例

地球重力场的变化是导致陆地水储量变化的重要因素之一,利用GRACE(Gravity Recovery and Climate Experiment)重力场恢复与气候实验重力卫星数据,结合GLDAS(Global Land Data Assimilation Systems)全球陆面数据同化系统和实测地下水位数据,反演和田地区克里雅河流域11年间四季和田地区的陆地水储量动态变化,模拟计算地下水等效水高变化趋势,构建了地下水水位估算模型。研究结果表明：和田地区春、夏两季的陆地水储量呈现出增加趋势,而秋、冬两季出现亏损状态;GRACE地球重力卫星所反演的陆地水储量比GLDAS同化系统所模拟的水资源变化更为剧烈,但2类数据的动态变化拟合度很高;GLDAS水资源等效水高二阶微分、GLDAS水资源变化倒数一阶微分、GRACE陆地水储量变化倒数变化、地下水储量变化一阶微分的敏感程度最高,构建的多元逐步回归模型明显优于线性函数,且水位深度越浅,该估算模型的适用性越高。

Mathematical fitting of influencing factors and measured groundwater level: Take Keriya River Basin in Hetian area as an example

The change of Earth's gravity field is an important factor that leads to the change of terrestrial water storage. Using GRACE (gravity recovery and climate experiment) gravity field recovery and climate experiment gravity satellite data, combined with GLDAS (global land data assimilation systems) global land surface data assimilation system and measured groundwater level data, inversion of the dynamic changes of terrestrial water storage in the Hetian area during four seasons in 11 years was simulated and the trend of change in equivalent water height was calculated, and the estimation model of groundwater level was constructed. The results indicate that the terrestrial water storage showed an increasing trend in the Hetian area in the spring and summer, in contrast to the autumn and winter; inversion of the terrestrial water storage by GRACE is more intense than simulation of the water resources change by GLDAS, but correlation of the dynamic changes of the two kinds of data is very high; equivalent water height by GLDAS second order differential, change of water resources by GLDAS reciprocal first order differential, change of terrestrial water reserves by GRACE reciprocal transformation, and change of groundwater reserves first order differential are most sensitive, and the stepwise regression model is superior to the linear function. The lower the water level, the higher the applicability of the stepwise regression estimation model.