利用GRACE数据联合新型尺度因子校正法提高陆地水储量变化准确性

Improve the accuracy of GRACE terrestrial water storage changes using GRACE data combined with a new scale factor correction method

本文围绕GRACE数据在信号处理过程中存在泄露误差开展了探索性研究.第一,在传统尺度因子法的基础上,根据模型与CSR-SHc数据的均方根误差和相关性赋予权重,构建了新型尺度因子校正法.第二,以长江流域为例,评估该方法的校正效果,研究结果表明:新型尺度因子校正法校正结果综合GLDAS(Global Land Data Assimilation System)水文模型计算的尺度因子校正结果空间分布趋势的优点,避免了CSR(Center for Space Research)官方提供的尺度因子、WGHM(Water GAP Global Hydrology Model)尺度因子和迭代恢复法校正结果空间分布趋势不均匀的现象.在长期趋势上,该方法校正结果优于GLDAS水文模型计算的尺度因子校正结果;在周年振幅上,新型尺度因子校正法校正结果明显优于迭代恢复法和CSR Mascon数据的结果.第三,基于该方法校正结果显示,长江流域、上游和中下游在2002年4月—2017年1月水储量呈现上升趋势,分别为0.29 cm·a^(-1)、0.14 cm·a^(-1)和0.49 cm·a^(-1),相比于校正前CSR-SHc数据在长江流域、上游和中下游上升趋势0.21 cm·a^(-1)、0.07 cm·a^(-1)和0.40 cm·a^(-1),分别提高了38%、100%和23%.长江流域水储量上升趋势主要集中在中下游.

The purpose of this study is to reduce leakage errors caused by GRACE signal processing.Firstly,a New Scaling Factor Correction Method(NSFCM)is developed by assigning weight to the mean square error and correlation calculated by the model and CSR-SHc.Secondly,the Yangtze River Basin is selected as the example to evaluate the performance of the NSFCM.Results indicate that the correction results using NSFCM,on the one hand,take the advantages of the spatial distribution trend of the scaling factor correction results calculated by the GLDAS(Global Land Data Assimilation System)model and WGHM(Water GAP Global Hydrology Model).On the other hand,it can avoid the uneven spatial distribution trend of the scale factor and iterative restoration method officially provided by CSR(Center for Space Research).In terms of the long-term trend,the NSFCM outperforms the scaling factor correction results calculated by the GLDAS and WGHM.With regard to the annual amplitude,the NSFCM results are significantly better than that of the iterative recovery method and CSR Mascon data.Thirdly,the NSFCM results show that the water storage of the Yangtze River Basin,upstream,and middle and lower reaches reveals an increasing trend at 0.29 cm·a^(-1),0.14 cm·a^(-1) and 0.49 cm·a^(-1) from April 2002 to January 2017,respectively.Compared to the rising trends before correction,the uptrend increases by 38%,100%,and 23%,respectively.Moreover,the rising trend of water storage is mainly concentrated in the middle and lower reaches of the Yangtze River Basin.