551. The process of multiple bar formation was simulated and compared 

 with one case from the CE data set where two bars developed. Multiple bars 

 could be generated in the numerical model by allowing wave reformation and the 

 appearance of multiple break points. Data from the LWT experiments gave 

 little guidance about the cross-shore transport properties in zones of wave 

 reformation. The transport rate in this zone was determined through simple 

 functional relationships based on qualitative observations from the 

 experiments . 



552. The model was also used to simulate onshore transport and berm 

 formation by using one CE case. The size of the berm was well reproduced; 

 however, the model failed to adequately describe the seaward berm face slope 

 and inshore profile shape. The seaward berm face slope is only limited by the 

 angle of initial yield in the model because of the crude description of 

 transport in the swash zone . 



553. A comparison between the present model and the Kriebel (1982, 

 1986) model was conducted to evaluate how bar formation would affect beach 

 erosion. One hypothetical case involved a variation in water level which 

 prevented bar development; both models gave similar predictions of erosion. 

 The description of the profile at the dune toe was more realistically des- 

 cribed by the present model, based on experience with the LWT and other 

 experiments, than by the Kriebel model, which distributed the eroded material 

 more evenly over the surf zone. Another comparison case involved bar develop- 

 ment, giving a significant difference in dune retreat, for which the Kriebel 

 model produced a larger amount of erosion than the present model, as expected. 



554. The numerical model was also used to simulate bar movement in the 

 field at CERC's FRF in Duck, North Carolina. Four different storm events 

 showing erosive profile response and offshore bar movement were used in the 

 calibration, and another independent event was used for verification of the 

 model. Some empirical model parameters determined with the LWT data had to be 

 modified somewhat to achieve agreement with measured field profiles. In 

 particular, the transport rate coefficient took a smaller value for the field 

 simulation than for the LWT cases, with an overall best value in a least- 

 square sense of 0.7 10'^ m''/N. 



236 



