similar value, being equivalent to 0.007 m of water for an 11 -bit binary 

 discretization of the full-scale signal. 



91. Spacing between the gages along the linear array appears irregular 

 in Figure 3 but is, in fact, distinctly specified by analysis requirements. 

 Minimum gage spacing is 5 m, and maximum spacing (the length of the array) is 

 255 m . These distances are associated with directional resolving ability for 

 waves with the shortest wavelength (highest frequency) and longest wavelength 

 (lowest frequency) of interest. The remaining gages are spaced more or less 

 so that unique multiples of the minimum spacing are distributed evenly up to 

 the length of the array (which is about 51 multiples of the minimum spacing) . 

 With nine gages, there are 36 possible unique spacings. In this array, there 

 are some redundant spacings left intentionally for checking spatial homogenei- 

 ty of the wave field, but well over half of the 51 spacing increments are 

 present. These are distributed for optimum resolution of wind wave directions 

 in the frequency range 0.05 to 0.32 Hz. 



92. It is not critical that the spaces are exact multiples of the 

 minimum spacing, but it is critical that gage locations are well measured. 

 Measuring was done by surveying the gage mounts (using a range pole which 

 protruded through the water surface and was held by an operator on the FRF's 

 Coastal Research Amphibious Buggy (CRAB)) with a Zeiss Elta-2 total station 

 referenced to the FRF coordinate system (see Birkemeier and Mason 1984). 

 Accuracy of the total station and CRAB combination is ±0.03 m. Accuracy of 

 pole positioning is estimated to be ±0.15 m. The maximum directional error of 

 this array is ±3 deg, which occurs only if two minimum- spaced gages are used. 

 In normal analysis, a minimum of five gages is used. Spacing is such that 

 positioning errors result in a wave direction error of ±0.3 deg. 



Data Collection 



93. Data from the 10 pressure gages were sampled in groups called 

 collections . Each collection consisted of sequentially sampling the digitized 

 analog voltage from each gage every 0.5 sec (2 Hz). The A/D sampling sweep 

 rate was 10 kHz, so it took 0.001 sec to sample 10 channels. This introduced 

 a negligible phase shift between channel samples; i.e., the sampling was 

 almost instantaneous for the 10 channels. Time series length for each channel 

 was 16,384 points. At 2 Hz, this is a duration of 2 hr 16 min 32 sec. Raw 



36 



