利用三维加速度点质量模型法解算华北地区陆地水储量变化

三维加速度点质量模型法为反演陆地水储量变化提供了新的途径，采用三维加速度点质量模型法计算了中国华北地区2003—2014年的水储量变化。为了检验反演结果，采用球谐系数法以及德克萨斯大学空间研究中心(Center for Space Research，University of Texas at Austin，CSR)发布的RL06 Mascon模型进行对比分析。研究结果表明，两种方法反演结果均反映出华北地区陆地水储量长期处于亏损趋势，但不同方法计算的亏损速度有一定的差别，三维加速度点质量模型法采用CSR提供的RL06数据反演的华北地区陆地水储量亏损速度为-3.09$\mathrm{c}\mathrm{m}/\mathrm{a}$，而球谐系数法反演结果为-2.60$\mathrm{c}\mathrm{m}/\mathrm{a}$；三维加速度点质量模型法特征点的反演结果与Mascon法相关系数更高，而球谐系数法与三维加速度点质量模型法结果之间的差异主要是由条带噪声约束平滑策略不一致导致的。

Monitoring the Variation of Terrestrial Water Storage in North China by Three-Dimensional Acceleration Point-Mass Modeling Approach

  Objectives  The lack of water resources in North China has severely affected the further development of industry and agriculture in the region. Previous studies have shown that terrestrial water storage (TWS) in North China are in a long-term loss trend, but the inversion methods are different, and three-dimensional acceleration point-mass modeling approach (3D-PMA) provides a new way for inversion of changes in TWS.  Methods  We invert TWS in North China from 2003 to 2014 based on 3D-PMA. 3D-PMA decomposes the changes of the satellite's perturbation force into the three directions of the coordinate axis, and establishes the least square equation through theoretical calculations and observations.We use the average spherical harmonic coefficients from 2003 to 2014 as the background model, and replace C₂₀ with satellite laser ranging (SLR) data. The monthly difference relative to the background model is used as the change in the spherical harmonic coefficient caused by the change in surface mass. For the ill-posed problems of least squares, the 600 km linear-type spatial constraint is adopted. In order to verify the results, spherical harmonic approach (SH) and the RL06 Mascon model released by Center for Space Research, University of Texas at Austin (CSR) are used for comparative analysis.  Results and Conclusions  The inversion results of the 3D-PMA and SH approaches indicate that TWS in North China have been in a long-term loss trend, but the loss rates calculated by different methods have certain differences. The loss rate of 3D-PMA is -3.09$\mathrm{c}\mathrm{m}/\mathrm{a}$, and SH is -2.60$\mathrm{c}\mathrm{m}/\mathrm{a}$ while using RL06 data provided by CSR to inverse TWS in North China. According to the calculation results of the characteristic points, 3D-PMA and Mascon have a higher correlation, and the difference between SH and 3D-PMA is mainly due to the inconsistency of the striping noise constraint smoothing strategy.