This letter deals with the analysis of a novel data processing approach to estimate the local depth in a shallow coastal area starting from data collected by an in situ X-band radar system. The reconstruction approach is based on the maximization of the normalized scalar product (NSP) between the measured and the theoretical wave dispersion relation, which embeds the dependence on the searched for parameter (local depth). The use of NSP approach allows obtaining a high-resolution spatial map of the investigated area, and a thorough statistical analysis is carried out by comparing the local depth estimation results with the ground-truth data made available by a multi-beam echo-sounder survey. In addition, the improved accuracy of the NSP approach is demonstrated with respect to other methods previously used for the analysis of the same area.