The Effect of Wind-Wave Growth on SAR-Based Waterline Maps

In this paper, we investigate an issue related to the use of synthetic aperture radar imagery to detect ocean and lake waterlines. In a previous publication, the waterlines indicated by L- and P-band AIRSAR imagery of a tidal flat in South Korea were compared and found to be offset from one another by approximately 80 to 170 m. The authors postulated that the difference was due to depth-dependent dissipation or dispersion of surface waves and the subsequent modulation of the radar backscatter as described by the Bragg model. In this paper, we present an alternative explanation based on the growth of wind waves as a function of distance (i.e., fetch). This new explanation is more consistent with the environmental conditions, radar look geometry, and surface wave theory, while also explaining several finer-scale features observed in the imagery that are not addressed in the original publication. Our results indicate that the detected waterline position should be a sensitive function of radar frequency only under a restricted set of conditions, namely, when the body of water in question is cutoff from incoming swell and the only surface waves present are locally generated by a land breeze. While such conditions may occur relatively infrequently in the coastal ocean, they appear to be common when imaging the windward shores of inland lakes, as illustrated by additional AIRSAR imagery. The analysis also serves as a remote-sensing-based validation of the existing theory for wind-wave growth in a wavenumber regime not previously studied in the field.