little loss of accuracy in determining an optimal calibration, since the 

 minimum of the sum of squares in most cases was located in a rather flat 

 region. 



433. Based on preliminary calibration runs, the coefficient expressing 

 the slope dependence of the transport rate (e in Equation 33) was set to 

 0.001 m^/sec. A smaller value of e will allow the trough to be locally 

 somewhat more pronounced, whereas a higher value will flatten the trough. The 

 angle of initial yield was set to 28 deg according to slope behavior inferred 

 from the LWT experiments, and the residual angle after shearing was set to 



18 deg. A larger angle of initial yield will allow the profile slope to 

 become steeper before avalanching occurs. During simulation of an erosional 

 event in the LWT data, avalanching typically takes place on the foreshore step 

 or on the shoreward side of the bar. 



434. At the initial stage of model calibration, both K and D^^ in 

 the transport equation (Equation 33) were used in the calibration procedure. 

 The transport rate coefficient K was varied together with D^^ for 10 

 erosional cases. Although it was considered desirable to avoid using D^^ as 

 a calibration parameter and instead determine its value from the design curve 

 given by Moore (1982), it was found that in order to achieve best agreement 

 between numerical model simulations and tank measurements , the value of D^^ 

 had to be reduced. The equilibrium energy dissipation controls the amount of 

 sand that is eroded before the equilibrium profile is attained. Moore's 

 relationship was derived by a least-squares fit of a power curve (Equation 1) 

 to beach profiles in general, making this method not entirely compatible with 

 the concept of regions with different transport rate relationships used in the 

 present numerical model. In most cases, the parameter combination which gave 

 the minimum sum of squares was located in the vicinity of an equilibrium 

 energy dissipation value of about 75 percent of that obtained by Moore's 

 relationship. This fixed reduction (0.75) of the equilibrium energy dissipa- 

 tion was applied in all cases, and the optimal value of the transport rate 

 coefficient K was determined by minimizing the sum of the squares of depths. 



435. Values of the transport rate coefficient for the 10 cases simu- 

 lated which gave the best agreement between measured and simulated profiles 

 varied in the range of 0.3 - 2.2 10"^ mVN, with an average of 1.4 10"^ mVN 



177 



