摘要
基于GRACE和GRACE-FO卫星陆地水储量遥感数据,采用长短期记忆(LSTM)神经网络模型,结合水量平衡方程和全球陆地数据同化系统(GLDAS)重建GRACE与GRACE-FO间的陆地水储量变化量,分析黄河流域2002年4月至2020年3月陆地水储量变化特征,探究影响陆地水储量变化的环境因子。结果表明:LSTM模型可以有效填补GRACE与GRACE-FO间的陆地水储量变化量;黄河流域陆地水储量呈明显下降趋势,上、中、下游下降趋势依次增大,陆地水储量与地下水储量的变化特征高度相关;黄河流域上、中、下游年陆地水储量变化量与年降水量和年干燥度指数呈极显著相关关系,表明黄河流域陆地水储量变化受到降水和蒸散发的影响。
Based on GRACE and GRACE-FO satellite remote sensing data,the terrestrial water storage change between GRACE and GRACE-FO was reconstructed using the long short-term memory(LSTM)neural network model,combined with the water balance equation and global land data assimilation system(GLDAS).Then,the change characteristics of terrestrial water storage of the Yellow River Basin from April,2002 to March,2020 were analyzed,and its environmental factors were explored.The results show that the LSTM model can obtain the missing data of terrestrial water storage change between GRACE and GRACE-FO effectively,and the terrestrial water storage of the Yellow River Basin shows an obvious decreasing trend,with the decreasing trend getting more significant along the upper,middle,and lower reaches.The change characteristics of terrestrial water storage and groundwater storage are highly correlated,and there are also significant correlations between the annual terrestrial water storage change,annual precipitation,and annual dryness index in the upper,middle,and lower reaches of the Yellow River Basin,indicating that the terrestrial water storage change in the Yellow River Basin is affected by precipitation and evapotranspiration.
作者
任立良
王宇
江善虎
卫林勇
王孟浩
张怡雅
REN Liliang;WANG Yu;JIANG Shanhu;WEI Linyong;WANG Menghao;ZHANG Yiya(State Key Laboratory of Hydrology-Water Resources and Hydraulic Engineering,Hohai University,Nanjing 210098,China;College of Hydrology and Water Resources,Hohai University,Nanjing 210098,China)
出处
《水资源保护》
EI
CAS
CSCD
北大核心
2022年第4期26-32,共7页
Water Resources Protection
基金
国家自然科学基金(51979069,U2243203)
中央高校基本科研业务费专项(B200204029)。
关键词
陆地水储量
GRACE
长短期记忆神经网络模型
GLDAS
水量平衡方程
黄河流域
terrestrial water storage
gravity recovery and climate experiment(GRACE)
long short-term memory(LSTM)neural network model
global land data assimilation system(GLDAS)
water balance equation
the Yellow River Basin