The local-wavenumber method estimates the depth to a magnetic source based on the spectral content of a single anomaly assuming that the base of the magnetic body is at infinite depth. However, the “infinite-depth” assumption can lead to significant underestimation of the depth to the top of magnetic bodies, especially in areas where the depth to the bottom of the magnetic layer is not large compared to the depth to the top, as would occur in high heat-flow regions and thinned continental margins. Such underestimation of depths has been demonstrated in model studies and using real data with seismic and well control. We evaluated a modification to the local-wavenumber approach to estimate the depth to the top of magnetic sources assuming that the depth to the bottom of the magnetic sources is controlled by the Curie temperature or crustal thickness. We applied this new method to a simple model of a continental margin and to magnetic survey data over the central Red Sea where the Curie isotherm is shallow. The effective structural index of this finite depth extent model is found to increase continuously from the continent to the ocean as the depth to the magnetic basement increases and the depth to the bottom of the magnetic layer decreases. We have also discovered in this study that the local-wavenumber maxima correlate well with major seafloor spreading magnetic reversal epochs in the central Red Sea segment.