PART V: EVALUATION OF THE VELLINGA AND KRIEBEL MODELS 



93. Both the Vellinga and Kriebel models are highly empirical in their 

 conception and formulation. However, empirical modeling techniques have been 

 proven to be capable of predicting certain extremely complex natural phenom- 

 ena, such as dune erosion. These predictions can be reliable if 



a. The modeling parameters have been carefully selected for a 

 particular site. 



b. The model results have been properly interpreted in a manner 

 consistent with the assumptions and simplifications inherent 

 in the model formulation. 



c_. The modeled area and storm event do not significantly deviate 

 from the conditions for which the model was formulated and 

 tested. 



94. Since a certain amount of interpretative skill and engineering 

 judgment are required for use of both the Vellinga model and the Kriebel 

 model, a quantitative comparison of the two models was made to demonstrate 

 their capabilities. Sargent and Birkemeier (1985) used a number of indi- 

 vidual beach and offshore profiles for each of four storms to demonstrate the 

 accuracy of the Vellinga model. In this section, the Kriebel model is 

 applied to these same data and compared both to the Vellinga model results 

 and to measured changes. The comparison is based on only 14 points. A more 

 rigorous evaluation with a wider variety of data, though desirable, was 

 beyond the scope of this feasibility study. 



Field Data 



95. The storm data used by Sargent and Birkemeier (1985) are summar- 

 ized in Table 2. Deepwater wave data were either obtained from 3-hr wave 

 hindcasts (Jensen 1983) in 9 m of water (Westhampton) , from gage measurements 

 in comparable water depths (Duck), or estimated from visual breaking wave 

 observations (Long Beach Island). Surge heights (without wave setup or 

 runup) were computed from water level measurements collected by the nearest 

 available sensors. Sediment fall velocities were determined from repre- 

 sentative sediment grain size data from each beach. 



96. The severity of each of the four storm events used in the compar- 

 ison was quantified by both pre- and poststorm profile surveys, including 



47 



