for each sensor). These were converted onboard to millimeters of seawater. After arrival 

 at WES. the data were stored in a database. 



A copy of the data was then automatically put through a routine analysis procedure. 

 The data were first decimated to 1 Hz, then truncated to the first 2,048 decimated data 

 points. An edit routine checked for spikes and flat spots. The data were then converted to 

 pressure (newtons per square meter), detrended, and demeaned. An auto- and cross- 

 correlation matrix was produced using a Welch method to produce 3 1 data segments 

 (50 percent overlapping segments of 128 points each). Thus, the frequency resolution for 

 this analysis was 1/128 sec or 0.00781 Hz. A 10-percent cosine taper (window) was 

 applied to each segment. The resulting 2-D spectrum was then surface corrected on a 

 frequency-band by frequency-band basis using linear wave theory. This spectrum was 

 truncated when the surface correction factor exceeded 100. Because the three gauge sites 

 were at different depths, each had a different minimum cut-off period; approximately 

 4.3 sec for the offshore gauge, 2.9 sec for the ebb shoal gauge, and 2.5 sec for the throat 

 gauge. 



The co-spectrum was then integrated to obtain the wave height. The frequency band of 

 the co-spectrum with the maximum energy was used to designate the peak period. The 

 quad-spectrum was used to determine the wave direction of the peak frequency. These 

 wave parameters are defined as follows (see Earle, McGehee, and Tubman (1995) for 

 additional information): 



a. Zero Moment Wave Height (H m0 ): Spectrally-derived wave height, in meters; 

 equivalent to time-domain-derived significant wave height in deep water. 



b. Peak Wave Period (T p ). Peak spectral period, in seconds; inverse of the frequency 

 of the peak (highest energy) of the one-dimensional (1-D) power spectrum. 



c. Peak Wave Direction (D p ): Peak spectral direction, in degrees clockwise from true 

 north; mean direction from which energy is coming at the peak of the 1-D power 

 spectrum. 



The time stamp associated with these products refers to the time at the beginning of the 

 1-hr data record. 



During very low-energy wave conditions, while H m0 can be calculated, the calculation 

 of other wave parameters is difficult, subject to greater relative error, and may be 

 misleading. Therefore, wave period and direction calculations were omitted from the data 

 set whenever the wave height dropped below a threshold of 0.2 m. 



Application of these wave parameters to engineering solutions requires judgement 

 and understanding of their limitations. Critical decisions should take all available 

 information, such as the directional spectra of the original measured time series, into 

 account. However, heuristic rules of thumb have evolved among the wave measurement 

 community that may be of assistance in utilizing these data. Reasonable assumptions for 

 typical engineering applications with these data are that the uncertainty in significant wave 

 height is on the order of 10 percent, and on the order of 10 deg for peak direction. 



Chapter 2 Instrumentation and Procedures 



