Estimation of small surface displacements in the Upper Rhine Graben area from a combined analysis of PS-InSAR, levelling and GNSS data
Estimation of small surface displacements in the Upper Rhine Graben area from a combined analysis of PS-InSAR, levelling and GNSS data
复制标题
DOI：
10.1093/gji/ggv328
复制
发表时间：
2015-10-01
影响因子：
2.8
通讯作者：
Heck, B.
中科院分区：
文献类型：
--
作者：
Fuhrmann, T.;Cuenca, M. Caro;Heck, B.
利用大地测量技术研究了位于中欧的上莱茵河地堑（URG）的板内变形。本文提出了一种新的方法来计算由干涉合成孔径雷达、水准测量和GNSS测量得到的组合速度场。由于URG地区预期的构造运动较小（小于1 mm a（-1）），因此第一步估计了单一技术分析的线速度率的最佳可能解。其次，我们结合联合收割机的速度率从干涉合成孔径雷达（视线速度率在上升和下降的图像几何形状），水准测量（垂直速度率）和GNSS（水平速度率）使用最小二乘调整（LSA）。针对北方URG地区，我们分析了SAR数据上的四个不同的图像堆栈（ERS上升，ERS下降，Envisat上升，Envisat下降）使用持久性散射（PS）的方法。在上升和下降的图像的几何形状，分别的线速度率估计在LSA从联合的时间序列分析的ERS和环境卫星数据。水准测量的垂直速度率是通过使用运动位移模型对40 000多个测得的高度差进行一致调整而获得的。通过对URG区域76个永久运行的全球导航卫星系统站点每日坐标估计值进行时间序列分析，计算出了东向和北向的水平速度。由于PS-InSAR、水准测量和GNSS的测量数据所处的位置不一致，在严格处理的几个步骤中需要进行空间插值。我们使用普通克里格法从给定的一组数据点插值到感兴趣的位置，特别关注误差的建模和传播。最终的三维速度场是在200 m网格上计算的，其值仅接近PS点的位置，导致平均水平和垂直精度分别为0.30和0.13 mm a（-1）。合成速度场的垂直分量显示，地堑北方部分有约0.5 mm a（-1）的显著沉降，与一个著名的第四纪盆地结构相吻合。在地堑外观察到东南方向的水平位移速率高达0.8 mm a（-1），与最大水平应力的平均方向合理对齐。在地堑内部，非沉降段的速度方向向东旋转，而沉降段的速度方向则相反。所观察到的速度场的复杂性是兼容的地质力学情况，在我们的调查区域，其特征在于从一个过渡到释放弯曲设置。冰川均衡调整是影响观测速度场的另一个潜在来源，也是一些地方已确定的采矿、石油勘探和地下水使用所产生的人为信号。
The intra-plate deformation of the Upper Rhine Graben (URG) located in Central Europe is investigated using geodetic measurement techniques. We present a new approach to calculate a combined velocity field from InSAR, levelling and GNSS measurements. As the expected tectonic movements in the URG area are small (less than 1 mm a(-1)), the best possible solutions for linear velocity rates from single-technique analyses are estimated in a first step. Second, we combine the velocity rates obtained from InSAR (line of sight velocity rates in ascending and descending image geometries), levelling (vertical velocity rates) and GNSS (horizontal velocity rates) using least-squares adjustment (LSA). Focusing on the Northern URG area, we analyse SAR data on four different image stacks (ERS ascending, ERS descending, Envisat ascending, Envisat descending) using the Persistent Scatterer (PS) approach. The linear velocity rates in ascending and descending image geometries, respectively, are estimated in an LSA from joint time-series analysis of ERS and Envisat data. Vertical velocity rates from levelling are obtained from a consistent adjustment of more than 40 000 measured height differences using a kinematic displacement model. Horizontal velocity rates in east and north direction are calculated from a time-series analysis of daily coordinate estimates at 76 permanently operating GNSS sites in the URG region. As the locations, at which the measurement data of PS-InSAR, levelling and GNSS reside, do not coincide, spatial interpolation is needed during several steps of the rigorous processing. We use Ordinary Kriging to interpolate from a given set of data points to the locations of interest with a special focus on the modeling and propagation of errors. The final 3-D velocity field is calculated at a 200 m grid, which carries values only close to the location of PS points, resulting in a mean horizontal and vertical precision of 0.30 and 0.13 mm a(-1), respectively. The vertical component of the combined velocity field shows a significant subsidence of about 0.5 mm a(-1) in the northern part of the graben coinciding with a well-known quaternary basin structure. Horizontal displacement rates of up to 0.8 mm a(-1) in southeast direction are observed outside the graben, in reasonable alignment with the average direction of maximum horizontal stress. Within the graben, the velocity directions rotate toward east in the non-subsiding part, while an opposite trend is observed in the subsiding part of the graben. The complexities of the observed velocity field are compatible to the geomechanical situation in our investigation area which is characterized by a transition from a restraining to a releasing bend setting. Glacial isostatic adjustment is another potential source influencing the observed velocity field, as well as anthropogenic signals due to mining, oil exploration and groundwater usage that have been identified in some places.