利用Forward-Modeling方法反演青藏高原水储量变化

Water Storage Variation Inversion in the Tibetan Plateau by Using Forward-Modeling Method

青藏高原水资源对我国经济社会发展及气候变化影响深远。借助重力反演与气候实验(gravity recovery and climate experiment,GRACE)卫星重力数据,针对其滤波泄漏误差影响,采用Forward-Modeling方法进行定量估算。在顾及冰川均衡调整(glacial isostatic adjustment,GIA)效应情况下,反演青藏高原水储量变化,将结果与全球水评估与预测模型(water-global assessment and prognosis hydrology model,WGHM)进行比较分析。利用短时傅里叶变换提取时序信号的时频谱特征,并结合全球降水气候中心(global precipitation climatology centre,GPCC)降水数据探讨水储量与降水的关系。研究发现:(1)恢复泄漏信号后水储量变化呈现明显的空间差异性,大部分时段青藏高原正北和东南部水储量大幅盈余,正南和西南部水储量快速亏损,其空间特征与WGHM模型结果基本一致。(2)水储量时频谱分布以低频信号为主,其动态变化具有明显的季节性和阶段性变化,周年振幅达到7.5 cm。2003—2004年水储量增加速率为(3.9±4.9)cm/a;2005—2010年亏损速率为(-0.3±0.8)cm/a;2011—2014年上升速率为(0.2±1.6)cm/a。(3)降水是影响陆地水储量变化的主要因素,在降水明显偏少的2006年和2009年,水储量均存在明显负异常;在降水偏多的2010年,水储量显著回升。实验证明,该泄漏误差改正方法和研究结果对于区域水储量变化的定量研究具有重要参考价值。

The water resources in the Tibetan Plateau have a profound impact on China’s economic,social development and climate change.Based on gravity data from gravity recovery and climate experiment(GRACE)satellite,considering its filtering leakage error,this paper proposes the Forward-Modeling method for quantitative estimation.In view of the glocial isostatic adjustment(GIA)effect,we first infer terrestrial water storage(TWS)changes in the Tibetan Plateau,and compare the results with the water-global assessment and prognosis hydrology model(WGHM).Then we extract the time-frequency spectrum of timeseries signals using short-time Fourier transformation algorithm.In addition,we discuss the relationship between TWS and precipitation by combining with precipitation data from Global Presipitation Climatology Centre.The study finds that:(1)TWS shows obvious spatial differences after recovering leakage signal.In most of the period,TWS in the north and southeast is in large surplus,and TWS in the south and southwest is rapidly depleted.The spatial characteristics are basically consistent with the WGHM results.(2)The time-frequency spectrum distribution of TWS is dominated by low-frequency signals,and time-series dynamic changes have obvious seasonal and periodic changes,with an annual amplitude of 7.5 cm.The increase rate of water storage in 2003―2004 was 3.89 cm/a.The loss rate in 2005―2010 was-0.31 cm/a.The rate of increase in 2011―2014 was 0.22 cm/a.(3)Precipitation is the main factor affecting the variation of TWS.In 2006 and 2009,precipitation was significantly lower,the water storage all decreased significantly.In 2010,there was heavy rainfall,the water storage increased significantly.The leakage error correction method and research results in this paper have important reference value for quantitative research on regional TWS varition.