LAUSR

The Curie-point depth is an important magnetic parameter for studying the geological structure and thermal evolution of the lithosphere. In this study, we first apply wavelet transform to the centroid method, and compare it with the conventional Fourier spectral centroid method based on a 3-D fractal magnetization model. We find that the Fourier transform method is suitable for estimating smaller Curie-point depths (less than 20 km), whereas the wavelet transform method has better accuracy in estimating large Curie-point depths (more than 30 km). The application of both methods to the western Pacific magnetic anomalies from the Earth Magnetic Anomaly Grid of 2-arc-minute resolution (EMAG2) proves their reliability and comparability. Our results show small Curie-point depths in areas with active magmatism and high heat flow, like mid-ocean ridges and island arcs of subduction zones, but large Curie-point depths in sedimentary basins and forearcs of subduction zones with low surface heat flow. We wrote our Curie-point depth inversion software CPDINV using FORTRAN language, which includes completed modules of minimum-curvature gridding, projection, wavelet and Fourier transforms, least-square fitting, and depth estimation.