基于Cokriging的三维重力张量随机反演

3D stochastic inversion of gravity tensor based on Cokriging

重力张量是重力位的二阶导数,与传统的重力测量相比,重力张量具有更高的分辨率。协同克里格法是地质统计学中常见的一种方法,使用协同克里格法对三维重力张量各分量数据进行随机反演。由于位场数据在深度方向上的分辨率较低,因此需通过使用深度加权函数和钻孔数据来解决这一问题,对比了带深度加权和带钻孔数据的协同克里格反演结果。反演结果表明,协同克里格法能够较好地反演重力张量各分量数据,两种方式都能够较好地反演出地下异常位置,带深度加权的协同克里格法反演出来的密度值与实际值有一定的偏差,而带钻孔数据的协同克里格法能够较好地反演出剩余密度值,并且对于不同埋深的异常体都能很好地反映。

Gravity tensor is the second derivative of gravitational field. Compared with traditional gravitational exploration, gravity gradient tensor has a better resolution. Cokriging is a common method in geostatistics. In this paper, we applied cokriging to 3D stochastic inversion of each gravity tensor componentdata. Because potential field data have little resolution a- long the depth direction, we use depth weighting function and borehole data to solve this problem, and we compare the inver- sion results between both above methods. The inversion results show that the eokriging method can invert each gravity tensor component data. And both methods are capable of inverting the position of causative in the subsurface. In addition inversion re- sults with depth weighting function cannot precisely recovered the anomalous bodies. But incorporating borehole date into in- version can improve the results with better resolution along the depth.