The galvanic distortion caused by small-scale electrical inhomogeneities near the shallow surface perturbs the overall frequency response curve of magnetotelluric (MT) records and adversely affects the imaging of deep electrical structures. Various efforts have been made to suppress the galvanic distortion, but they are still difficult to use in complex regional structures. To effectively recover the deeper electrical structures, we develop an inversion technique in which the shallow electrical structures are constrained by other types of electromagnetic data and used as part of the prior model to remove the galvanic distortion. When other types of electromagnetic data are unavailable, we can also simulate the galvanic distortion by using a refined forward mesh and putting small-scale electrical resistivity anomalies beneath the surface. We validate our inversion approach using synthetic examples and apply it to field MT data collected from the northern Trans North China Orogen. A three-dimensional electrical resistivity model is obtained based on the prior model designed according to the locations of distorted MT stations. In our resulting model, artifacts caused by galvanic distortion are removed as compared to previous work. We demonstrate that our method is effective in improving the reliability of MT inversions in complex regions, leading to better interpretations of deep electrical structures.

由浅表面附近的小规模电气不均匀性引起的电流畸变扰动了大地电磁（MT）记录的整体频率响应曲线，并不利地影响了深层电气结构的成像。已经进行了各种努力来抑制电流畸变，但是它们仍然难以在复杂的区域结构中使用。为了有效地恢复较深的电气结构，我们开发了一种反演技术，其中较浅的电气结构受到其他类型的电磁数据的约束，并被用作现有模型的一部分以消除电畸变。当其他类型的电磁数据不可用时，我们也可以通过使用精细的前向网格并将小范围的电阻率异常置于地表以下来模拟电变形。我们使用综合实例验证了我们的反演方法，并将其应用于从华北北部华北造山带收集的野外MT数据。根据变形的MT站的位置设计的现有模型，获得了三维电阻率模型。在我们得到的模型中，与以前的工作相比，消除了由电流失真引起的伪影。我们证明了我们的方法可有效提高复杂区域MT反演的可靠性，从而更好地解释深层电子结构。在我们得到的模型中，与以前的工作相比，消除了由电流失真引起的伪影。我们证明了我们的方法可有效提高复杂区域MT反演的可靠性，从而更好地解释深层电子结构。在我们得到的模型中，与以前的工作相比，消除了由电流失真引起的伪影。我们证明了我们的方法可有效提高复杂区域MT反演的可靠性，从而更好地解释深层电子结构。