Video measurements of θm(f) are validated against collocated in situ ADV

probe measurements. ADV observations of horizontal velocity components were

sampled at a rate of 20 Hz for a duration of 204.8 sec, resulting in 4,096 samples.

Time series records were linearly detrended, tapered, and ensemble averaged

resulting in a frequency resolution of 0.078 Hz and 32 deg of freedom with a

maximum frequency cutoff of 5 Hz.

To facilitate comparison, ADV energy density and direction values were

interpolated using a piecewise cubic Hermite interpolation within the frequency

bands observed by the video system. Superimposed on the rectified (lower)

images shown in Figures 26, 27, 29, 31, and 32 are solid circles that represent

ADV and CIIS measurement collocations. Figure 36 is a representative graphical

comparison of normalized energy density and direction versus frequency at a

single measurement location for a random wave experiment (*f*peak= 1.25 Hz). The

peak spectral density for graphical comparison normalizes the variance spectrum.

For visual reference, Appendix P presents a graphical comparison of energy

density and mean direction versus frequency for each Structure 1 collocated

CIIS-ADV location.

Figure 36. Representative collocated video-ADV spectra

Several statistics were selected to compare mean wave directions obtained

near the peak of the spectra for the two measurement techniques. These statistics

include mean angular difference (scalar wave direction bias), standard deviation

of mean angular difference, root-mean-square error and correlation coefficient.

These statistics are defined as follows for *X*i (ADV observation) and *Y*i (CIIS

observation), and a wave direction vector magnitude of unity:

50

Chapter 5 Video-Based Wave Direction Measurement

Integrated Publishing, Inc. |