The gravity and magnetic data sets over a geothermal source are investigated to explore the geological features and geothermal system manifestations. The raw gravity data sets are firstly corrected for drift, latitude and free-air effects to obtain free-air anomalies. A new approach is utilized for Bouguer and terrain corrections. The model space, which is the mass from the reference surface to the ground surface, is parameterized by a very accurate and advanced meshing algorithm (octree and quadtree). The gravity effect of the model space is computed via numerical forward modelling and is considered as the Bouguer and terrain corrections. These corrections are subtracted from the free-air anomalies, which yields the complete Bouguer anomaly. The computed Bouguer anomalies are de-trended and transferred to residual anomalies by methods including polynomial fitting and two-stage methods. The residual gravity anomalies are inverted to obtain the subsurface density distribution. The density contrasts are estimated by minimizing the data and model objective functions through an efficient algorithm. The probable source of the geothermal system and other new gravity anomalies are successfully detected and inverted contrary to the results in previously published papers. The residual magnetic anomaly is also inverted and the obtained geometrical and physical parameters of the source body including minimum depth, shape and susceptibility are very close to the previous results. The new detected gravity anomaly and the main magnetic anomaly in the overlapped area of gravity and magnetic grids are located at almost the same place with almost the same depths, which confirms their probable common source. We developed the open source package (http://docs.simpeg.xyz/content/tutorials/03-gravity), which is accessible through SimPEG (Simulation and Parameter Estimation in Geophysics), for inverting the gravity and magnetic data sets.

研究了地热源上的重力和磁性数据集，以探索地质特征和地热系统表现形式。原始重力数据集首先针对漂移、纬度和自由空气效应进行校正，以获得自由空气异常。一种新方法用于布格和地形校正。模型空间是从参考表面到地面的质量，通过非常精确和先进的网格划分算法（八叉树和四叉树）进行参数化。模型空间的重力效应通过数值正向建模计算，并被视为布格和地形校正。这些修正从自由空气异常中减去，产生完整的布格异常。计算出的布格异常通过多项式拟合和两阶段方法等方法去趋势化并转化为残差异常。对残余重力异常进行反演以获得地下密度分布。通过有效的算法最小化数据和模型目标函数来估计密度对比度。与先前发表的论文中的结果相反，地热系统和其他新的重力异常的可能来源被成功检测和反演。剩磁异常也被反演，获得的源体几何和物理参数包括最小深度、形状和磁化率与先前的结果非常接近。新探测到的重力异常与重磁网格重叠区域的主要磁异常几乎位于同一地点，深度几乎相同，证实了它们可能的共同来源。我们开发了开源包 (http://docs.simpeg.xyz/content/tutorials/03-gravity)，可通过 SimPEG（地球物理学中的模拟和参数估计）访问，用于反转重力和磁数据集。