Portal de Periódicos da CAPES

Wave-Filtered Surf Zone Circulation under High-Energy Waves Derived from Video-Based Optical Systems

2021; Multidisciplinary Digital Publishing Institute; Volume: 13; Issue: 10 Linguagem: Inglês

10.3390/rs13101874

2072-4292

Isaac Rodríguez‐Padilla, Bruno Castelle, Vincent Marieu, Philippe Bonneton, Arthur Mouragues, Kévin Martins, Denis Morichon,

Tropical and Extratropical Cyclones Research

This paper examines the potential of an optical flow video-based technique to estimate wave-filtered surface currents in the nearshore where wave-breaking induced foam is present. This approach uses the drifting foam, left after the passage of breaking waves, as a quasi-passive tracer and tracks it to estimate the surface water flow. The optical signature associated with sea-swell waves is first removed from the image sequence to avoid capturing propagating waves instead of the desired foam motion. Waves are removed by applying a temporal Fourier low-pass filter to each pixel of the image. The low-pass filtered images are then fed into an optical flow algorithm to estimate the foam displacement and to produce mean velocity fields (i.e., wave-filtered surface currents). We use one week of consecutive 1-Hz sampled frames collected during daylight hours from a single fixed camera located at La Petite Chambre d’Amour beach (Anglet, SW France) under high-energy conditions with significant wave height ranging from 0.8 to 3.3 m. Optical flow-computed velocities are compared against time-averaged in situ measurements retrieved from one current profiler installed on a submerged reef. The computed circulation patterns are also compared against surf-zone drifter trajectories under different field conditions. Optical flow time-averaged velocities show a good agreement with current profiler measurements: coefficient of determination (r2)= 0.5–0.8; root mean square error (RMSE) = 0.12–0.24 m/s; mean error (bias) =−0.09 to −0.17 m/s; regression slope =1±0.15; coherence2 = 0.4–0.6. Despite an underestimation of offshore-directed velocities under persistent wave breaking across the reef, the optical flow was able to correctly reproduce the mean flow patterns depicted by drifter trajectories. Such patterns include rip-cell circulation, dominant onshore-directed surface flow and energetic longshore current. Our study suggests that open-source optical flow algorithms are a promising technique for coastal imaging applications, particularly under high-energy wave conditions when in situ instrument deployment can be challenging.