Putting the AWAC to test

Comparison with directional wave buoys

In the winter 2002-2003, CEFAS conducted a comparison test where they deployed the AWAC with AST firmware and and a Datawell Directional Waverider. The two instruments had common a deployment period of November 27th, 2002 to January 4th, 2003.

The deployment site was "Gabbard" (see map) at 52 00’ N 002 20’ E. The AWAC deployment depth was 32 meters, the velocity depth cell size for the wave measurements was 3 m and the AST sampling rate was 2 Hz. One burst of 2048 point (about 17 minutes) was collected every hour and a current profiler was collected every 10 minutes. A total fo 1215 bursts were collected, 16 of which had an AST error rate higher than 5%.

The figure below presents a comparison of the two instruments for the most common wave parameters: Hs (significant wave height), Tp (peak period) and mean wave direction. s can see the significant wave height compare very favorably. This is similarly true with the peak period estimate. It is interesting to note that early in December, both instruments detect a peak period of over 12 seconds. Closer inspection of the time series suggests that there was a passing ship during a relatively calm sea state.

The wave direction estimates, which is the estimate of the direction of the dominant waves, provides good agreement for the long waves (Tp>5 s) but show more scatter for the shorter waves. This is attributed to the fact that the AWAC's projected array just below the surface is too large to resolve waves with small wavelengths. For the depth of 31 meters, this means that the AWAC could not unambiguously resolve waves smaller than 4.5 seconds wehn using the three slanted velocity beams only and 3.2 seconds when including the AST wave height in the directional alogrithm. This illustrates the only know weakness of the AWAC-AST system, which is the limited capability to measure the direction of short waves. As an aside, we have generally noted that the wave direction for waves with short period can be quite closely correlated with the wind direction. This is because waves of this period are commonly generated from local weather events.

The spectrogram show the logarithmc energy for each burst and the significant wave height. The purpose is to provide a sense of how the dominant energy distribution shifts over time, either as the sea developed over time (waves gets longer with time) or as long wave waves approach from a distant storm (waves gets shorter with time).