大地电磁测深和重力数据同步正则化联合反演

在以往正则化联合反演中,不同的地球物理方法均采用相同的正则化因子。当不同的地球物理方法的数据拟合泛函和模型稳定泛函的数量级不一致时,单正则化因子条件下的正则化联合反演不能充分发挥不同地球物理方法各自模型约束泛函的作用。针对这点不足,以大地电磁测深(MT)和重力联合反演为例,提出了双正则化因子的确定方法,新方法根据MT和重力方法各自的数据拟合效率实现了正则化因子的自适应调整。设计了非完全共界面的断块模型。模型试验表明:与传统的MT和重力正则化联合反演相比,双正则化因子条件下的MT和重力联合反演可以更好地发挥各自的模型约束泛函的模型约束作用,联合反演的解更符合真实模型。

Regularized synchronous joint inversion of magnetotelluric and gravity data

The same regularization parameter is used in the previous study of regularized joint inversion. The stabilizing functional of different geophysical method cannot play full role with single regularization parameter, when there is the disagreement on the magnitude order of the misfit functional and the stabilizing functional between different geophysical methods . Aiming at this disadvantage, we take the joint inversion of magnetotelluric(MT) and gravity for example, and present the approach to determinate bi-regularization parameters, in which regularization parameters can be modified adaptively according to the efficiency of respective data fitting for MT and gravity .The fault block model under non-perfect common interface condition is designed. Testing on synthetic model shows that, compared to traditional regularized joint inversion of MT and gravity, the new joint inversion with bi-regularization parameters can promote every stabilizing functional to exert better function in model constrain , and help us to obtain a solution which is close to the fact.