Depth Estimation for Magnetic/Gravity Anomaly Using Model Correction

The Tilt-depth method has been widely used to determinate the source depth of a magnetic anomaly. In the present study, we deduce similar Tilt-depth methods for both magnetic and gravity data based on the contact and sphere models and obtain the same equation for a gravity anomaly as that for a magnetic anomaly. The theoretical equations and the model tests show that the routine Tilt-depth method would result in unreliable depth estimation for deep sources. This is due to that the contact model is no longer valid for causative sources under the condition in which the depths of causative sources are significantly larger than their horizontal lengths. Accordingly, we suggest that the Tilt-depth derived from the contact model can be used to detect a shallow source, whereas the Tilt-depth derived from the sphere model can be used to detect a deep source. We propose a weighting method based on the estimated depths from both the contact model and the sphere model to estimate the depth for real data. The model tests suggest that the determined depths from the contact model and the sphere model are shallower and deeper, respectively, than the real depth, while the estimated depth from the proposed method is more close to the actual depth. In the application to the Weigang iron ore located in Jiangsu province, China, the routine Tilt-depth method results in -76% relative error, whereas the proposed method obtains the reliable depth estimation compared with the drill holes. In addition, the proposed method works well in the application for the Shijiaquan iron ore located in Shandong province, China. These results indicate that the proposed weighting equation is a general improvement.