Detiding ADCP data in a highly variable shelf sea

This paper presents a comparison of two conventional detiding techniques carried out for ship-mounted acoustic Doppler current profiler (ADCP) data collected in the European shelf area of the Celtic Sea during the summer of 1998. One technique consisted of extracting the vertically averaged tidal currents obtained from a barotropic three-dimensional numerical tidal model. The second technique consisted of fitting the spatiotemporal ADCP data using least squares and polynomial spatial functions. In the least squares technique, the incorporation of zero velocity normal to the coast appears to improve the estimation of the tidal currents near the coast. Quantitative comparisons of the results from both techniques with historical current meter observations are shown. However, both methods showed limitations in accurately representing the tidal currents in the study area. Consequently, an alternative detiding technique is proposed. This technique consists of blending the tidal currents derived from the numerical model with those fitted to the ADCP data from the least squares method. Improved results were obtained using the blending technique. ADCP-derived residual currents were comparable with contemporaneous flows measured using drifting buoys and also with estimates obtained by geostrophic calculations.