many seasons, wherein nearshore reflectivity would be expected to vary con- 

 siderably. 



Another possibility is that linear wave theory is not adequate during high- 

 energy conditions when the wave field may become saturated with highly 

 nonlinear, actively breaking waves. While this may be true for the extreme 

 storm of Event T, it is unlikely to be true for the very marginally energetic 

 storm of Event I, which was inadvertently included in this analysis, and had 

 an H^ that exceeded 2 m for less than 9 hr. The common spectral pattern 

 described above also occurs in Event I (see pages A33 and A34), so it is 

 improbable that a saturated wave field accounts for these observations. 



Two other possible causes for these results are a common structure in the 

 overall driving wind fields or the net effect of the broad, shallow bathymetry 

 of the continental shelf adjacent to the FRF. The anemometer used in this 

 report is only a crude indicator of the larger wind systems that generate waves 

 observed at the FRF. A possible line of future research to examine the larger 

 systems would be to apply a wind field model like that used 

 in the CERC Wave Information Studies (Resio, Vincent, and Corson 1982) to 

 the storm events examined in this report. 



In conjunction with this possible study would be the inclusion of a reason- 

 ably accurate bathymetry for the United States mid-Atlantic continental shelf. 

 The shelf break, at depths of 60 to 100 m, is 80 to 100 km distant from the 

 FRF (Figure 1). Waves with periods of 8 to 11 sec (common periods for 

 waves observed at the FRF) begin to be influenced by the bottom at these 

 depths. Consequently, considerable wave refraction is expected for intermedi- 

 ate and long waves generated over the shelf and in the deeper Atlantic by 

 nonuniform wind fields associated with atmospheric fronts and low pressure 

 systems. This condition may account for the directionally broad, low-frequen- 

 cy peaks and side lobes in observed storm spectra. However, a quantitative 

 assessment of these effects should be made, either with a numerical model or 

 with more extensive observations. 



It is possible that the directionally bimodal, high-frequency spectra may be 

 due to a hummock or linear bar just offshore of the FRF array that causes 

 strong refraction for high-frequency waves. This hypothesis can be tested 

 with a detailed bathymetric survey that extends beyond the area routinely mea- 

 sured by the FRF staff (generally, the area shown in Figure 2). Such a sur- 

 vey has not yet been done, but the results presented herein should provide 

 strong motivation for such work. There is weak evidence against a local 

 bathymetric effect in the present data set. In the later stages of Events A, Y, 

 and 2, winds shifted to the southeast with just enough force to generate waves 

 from that quadrant. Though rather faintly shown in the figures (see pages A6, 

 A87, and A96), bimodal spectral structures appear in the southeast quadrant in 

 forms roughly symmetric to those observed in the northeast quadrant. It is 

 somewhat unlikely that a local bathymetric feature would induce this kind of 

 symmetry in S(f,6) . 



Chapter 4 Discussion 



19 



