Virginia Beach Phase II model tests to quantitatively evaluate overtopping 

 potential. He recommended using a different correction factor for each swl. 

 These factors are presented in Section VI of this report. The correction fac- 

 tors represented a percent increase in wind-induced overtopping potential for 

 each swl. Dr. Resio observed from the video tapes that at low swl's more 

 potential for wind-induced overtopping occurred than at the higher swl's. As 

 the surge increased, the seawall became inundated causing less wave breaking 

 and thus less wind-induced overtopping. 



41. The SPM (1984) recommends another method for estimating wind- 

 induced overtopping. Appendix C provides a sample calculation which is based 

 on the following equation: 



k' = 1.0 + Wf (h - ds/R + 0.1) sin Q 



where 



k' = the wind correction factor 



Wf = a coefficient depending on wind speed 



h = height of the structure crest above the bottom 



ds = depth at the toe of the structure 



R = runup on the structure that would occur if the structure were high 

 enough to prevent overtopping 



9 = structure slope (90 being vertical) 



42. The increase in overtopping depends on wind velocity and direction 

 with respect to the axis of the structure and structure slope and height. The 

 equation requires use of empirical relations based on monochromatic or regular 

 waves. A comparison of the wind- induced overtopping rates using Resio' s 

 method and those calculated using the SPM (1984) suggests the latter method 

 probably results in conservative predictions. This comparison is provided in 

 Part VI. 



43. The effects of wave-induced setup in the model were measured. How- 

 ever, there are no prototype measurements of wind- or wave-induced setup for 

 storms along Virginia Beach which would have allowed verification of the rela- 

 tive scaling of these effects. The SPM recommends a procedure for computing 

 setup based on monochromatic waves which would have limited relevance to the 

 spectral conditions used in model testing. Since the model setup would be 

 more consistent with the other model-generated data used in the study, these 



28 



