Nearshore bar crest location quantified from time-averaged X-band radar images

Abstract The remote sensing of the sea surface is commonly regarded as a logistically simple and cost-effective way of obtaining long-term (>years) time series of nearshore sandbar location. In this paper, we demonstrate that, similar to video imagery, time-exposure X-band (∼3 cm wave length) radar images contain alongshore, breaking-related high-intensity bands that can be used to compute the location of the underlying sandbar crests. Analysis of time-exposure X-band images and bathymetric surveys collected at Egmond aan Zee, Netherlands shows that the radar-based bar crest location, defined as the cross-shore location of maximum radar intensity, differs from the in situ surveyed crest location by a distance of O(10 m). This difference, Δxradar, depends linearly on the offshore water level, consistent with model predictions in which the radar intensity is assumed to behave as the energy of the surface roller. The model additionally suggests a dependence of Δxradar on the offshore wave height when the wave field across the bar is non-saturated. In general, such wave conditions did not result in radar images with sufficient contrast to determine the bar crest location and, accordingly, the wave-height dependence was not observed clearly in the field. Because of the dependence of Δxradar on the offshore water level, time series of radar-based bar crest location show artificial variability when the images are collected at different water levels. Various techniques for correcting this variability are discussed.