Observations of large amplitude internal waves at the Malin Shelf edge during SESAME 1995

Abstract

Synthetic Aperture Radar (SAR) images of the Malin Shelf break area on August 20th and 21st 1995 separated by 24 hours show nearly identical internal wave signatures, with small spatial displacements between the two images. The coincidence of these patterns, imaged at similar times in the tidal cycle, suggests that the waves are tidally generated. These SAR images are here related to in situ data obtained during the coincident NERC Shelf Edge Study (SES) and the SES Acoustic Measuring Experiment (SESAME). Internal waves were observed propagating either up-slope or on-shelf, with little evidence of off-shelf propagation. The waves seen over the continental slope have been tracked between the SAR images, and towed thermistor chain records, using the assumption that the internal wave is regenerated every tidal cycle. The resulting calculations of phase speed validate this assumption, and indicate that the waves propagate towards the shelf from a distant, deep source and separate into non-linear solutions. The waves are large amplitude, causing displacements of the seasonal thermocline of up to 50 m. Detailed analysis of the thermistor chain data showed that the leading soliton is well described by first-order Korteweg de Vries theory, despite the large amplitude of the wave. Further, the soliton includes contributions from the first three vertical modes, the first mode creating large displacements at depths around 100 m, and the second and third modes significantly displacing the near-surface layers. Another SAR image on 5th September shows similar features on the Continental slope, also shown in coincident current-meter data. Predictions of soliton amplitude and phase speed from their length scales on SAR using the KdV soliton relationships are compared with the in-situ data for the two periods. The results that the phase speeds are reasonably well predicted, to within 10 cm s −1 (10–20% of the value) while amplitudes are predicted to at least within a factor of two (when compared to the thermistor chain data). Errors and uncertainty may be incurred due to the limit of the SAR geolocation precision, the difficulty of measuring exact internal wave length scales from SAR, and the natural variability of internal wave amplitude along its wavefront.