As analyzing geomagnetic data at low latitude areas for instance the Mekong Delta (latitudes  11,07^o), significant problem is that both of the magnetization and ambient field are not vertical totally, making magnetic anomalies antisymmetrical and often skewed to the location of the sources. In this paper, two-dimensional continuous wavelet transform (2-D CWT), using Farshad-Sailhac complex wavelet function is studied and applied for reducing the magnetic anomaly to a symmetrical one - this located on the source of the anomaly, and then determining the position of the center of the object causing anomalies by wavelet transform modulus maxima (WTMM) method. Next, magnetic data is extracted in two perpendicular directions passing through the center of the source to perform one-dimensional continuous wavelet transform (1-D CWT) to estimate the shape, depth and size of the source. Then, using the Marquardt algorithm to solve the inverse problem by least-squares method to further identify other characteristic parameters of the source such as: vertical size, remanent magnetization vector. The reliability of the proposed method is verified through theoretical models before application for analyzing the geomagnetic data in the Mekong Delta. The results are consistency with deep hole data, having small root mean square error, contribute to a better interpretation of the geological nature of the magnetic anomaly sources in the study area.

Keywords: low latitude, Marquardt algorithm, remanent magnetization vector, vertical size, 2-D CWT.