41. In principle, the value of the variability factor could be 

 dependent on surge height, wave height, and the median change. No such 

 dependency was identified in the data analyzed. 



42. The need for a variability factor has significant implications 

 relative to the required accuracy of the method. Obviously, with such a high 

 degree of variability, a precise estimate of erosion along an individual 

 profile is not required; however, a reasonable estimate of the maximum (here 

 chosen as the 75th percentile) change which may occur on any profile is 

 required. An important added consideration is the large degree of uncer- 

 tainty concerning the long-term stability of a particular beach prior to the 

 arrival of the 100-year storm. 



43. Although this report does not address additional requirements 

 introduced by long-term erosion (over periods of many years), the process 

 cannot be ignored. Birkemeier (1979) reported on significant changes 

 resulting from only a 2-year return period storm. The profile shape at any 

 particular time is related to the integrated effect of all previous storm 

 (and nonstorm) waves. Since predictions are invariably based on a limited 

 amount of beach survey data, the accuracy and usefulness of the prediction 

 will decrease with time. As an extreme example, a dune sufficiently sub- 

 stantial to survive a major storm may disappear under the combined attack of 

 a number of small storms over a number of years. The Grand Isle case 

 discussed in Part I is just one example of a scenario repeated along almost 

 every shoreline which is undergoing long-term erosion. 



24 



