Validation of ADCP surface wave measurements in a shelf sea

The performance of two wave ADCPs was analyzed, which were deployed at the North Sea nearby the Dutch coast. A 1200 kHz ADCP was deployed in 13 m water depth, which was set-up with two different protocols: (1) at 2 Hz in mode 1, with depth cells of 0.5 m and (2) at 1.25 Hz in mode 12, with 1 m cells. A 600 kHz ADCP was deployed in 18 m water depth, sampling at 2 Hz in profiling mode 1, with depth cells of 1 m. At both locations, a wave directional buoy was co-located for reference data. Spectral estimates of wave height and wave period from surface tracking agreed well with corresponding wave buoy estimates, showing only a small bias. Regarding surface tracking, especially the configuration with the 600 kHz ADCP with 1 m depth cells performed well. Scatter plots of ADCP velocity versus wave rider buoy estimates of spectral wave period feature a substantial bias for all configurations: spectral wave periods are increasingly underestimated for longer waves. This is caused by a relative excess of high-frequency energy density in the surface elevation spectra derived from ADCP velocity spectra via linear wave theory. The agreement between wave buoy and ADCP velocity-derived estimates of the principal wave direction increases if the sea state becomes more energetic