Bathymetry and Water-Level Estimation Using X-Band Radar at a Tidal Inlet

Knowledge of bathymetry in coastal areas is a key factor for maintaining shipping channels, identifying hazards to navigation, and measuring sediment transport processes. Remote-sensing techniques are crucial for this purpose, because in situ methods are often costly, time-consuming, and less frequent in occurrence. This paper presents a derived bathymetry and water-level retrieval algorithm that operates on time-series X-band marine radar data. The algorithm is designed to operate on a long-term continuous data stream, generating updated water-level measurements and bathymetry maps as new data are collected. In step 1, individual data sets are first processed to retrieve dominant wave numbers over a pyramid of overlapping tiles. In step 2, depths are estimated at each pixel location in the domain using the shallow-water wave dispersion relation. Last, results from step 2 are used to estimate the current water level. Updated water levels are used to tide correct a series of individual depth estimates, which are averaged together to produce an updated bathymetry map. In this way, no a priori or external data sources from the survey area are required. A 2-month-long experiment was conducted at Beaufort, North Carolina, to test the algorithm and quantify the frequency at which adequate signal levels were present for valid retrievals. Derived bathymetry and water-level results are presented and compared against ground truth. Derived measurements agree well with ground truth, although bathymetry accuracy degrades in areas of the domain with large bottom gradients or strong tidal currents.