Evidence for the influence of wave‐current interaction in a tidal boundary layer

Abstract

Near‐bed velocity profiles were measured in 24‐m water depth off the northeast coast of England. Measurements were made over two complete tidal cycles which spanned the waning stage of a storm. Superimposed on the tidal mean flow were progressively decaying wave‐orbital motions: the data thus span a wide range of relative wave and current energies and so offer an ideal opportunity to test wave‐current boundary layer theory. The magnitude of the tide‐modulated friction velocity, inferred from measured velocity profiles, appeared to decrease concurrently with the near‐bed wave energy, which is consistent with wave‐current theory. Also, the discrepancy between roughness inferred from the measured velocity profiles above the wave boundary layer and the “expected” roughness was greatest when the waves were most energetic. This, too, is consistent with wave‐current theory. The best evidence for a dynamic effect of the waves on the mean flow above the wave boundary layer was the correlation of the roughness discrepancy with the regular tidal variation in the strength of the wave‐orbital velocity relative to the mean flow. A model of the wave‐current boundary layer was used to predict the time‐averaged friction velocity, and the model predictions compared well with the observations with exceptions that, formed two groups. The first group, in which the model predictions were underestimates of the observations, comprised observations from the time of peak observed bed shear stress. A plausible explanation for the discrepancy is that an evolving ripple field and/or near‐bed saltating layer contributed an extra component of roughness that was unaccounted for in the predictions. The second group, in which the model predictions were overestimates of the observations, comprised observations from the times of minimum observed bed shear stress. These bursts also had the lowest observed roughness Reynolds numbers, all of which fell below the critical value for transition to fully rough turbulent flow. Since laboratory measurements imply wave‐current interaction does not occur in the smooth‐turbulent combined flow boundary layer, the predictions of time‐averaged friction velocity were repeated using a smooth‐turbulent pure‐current model. The predictions were significantly improved, thus supporting that contention.