multiplier to be used with a dune erosion estimate. Assuming that the 

 estimate was close to the median volume change on a beach and that a 

 desirable prediction should cover 75 percent of the changes (or any 

 acceptable percentage), then an average multiplier of the median value could 

 be computed to predict the 75 percent change. Of the 45 cases reported by 

 Birkemeier, Savage, and Leffler (in preparation) which had median above-MSL 

 volume losses in excess of -1 m-'/m of erosion, a median variability factor of 

 1.6 was computed and is plotted along with the distribution of factors in 

 Figure 11. Also shown in Figure 11 is a similar distribution of factors 

 based on the computed volume changes above the surge level. Only 35 of the 

 46 cases given in Table 1 had median above-surge level volume losses in 



MSL 



9.8 



ABOVE 

 MAXIMUM 



SURGE 



KEY 



EXTREME 



75% 



MEDIAN 



25% 



EXTREME 



1.9 

 1.6 



1.4 



I- 



1.3 



MEAN- 2.0 



MEAN.! .8 



Figure 11. Distribution of computed variability factors 



excess of -1 m^/m. It is interesting to note that the median above-surge 

 variability factor of 1.5 is nearly identical to the above-MSL factor. Use 

 of the 75 percent factor of 2.1 is recommended. This value can be inter- 

 preted as the multiplier of the median required to include volumetric erosion 

 expected on 75 percent of the shoreline. A higher value, for example 2.5, 

 would increase the percentage included and would conservatively account for 

 the uncertainty in the dune erosion estimate. This specific value of the 

 variability factor should be viewed as preliminary, to be revised as more 

 data become available. 



23 



