489 . Measured wave height and period were available every 6 hr and 

 water level every hour. Since a shorter time-step (typically, At = 20 min; 

 Ax = 5.0 m) was used in the numerical model, cubic spline interpolation was 

 used to provide input values for time-steps between measurements. The energy- 

 based wave height H^^ was determined from the wave spectrum as four times 

 the standard deviation, which corresponds to the significant wave height if a 

 Rayleigh distribution of the wave height is assumed. Wave period was given as 

 the peak period of the spectrum. Wave input data were obtained from gage 620, 

 located in 18 m of water directly seaward of the pier. The height was then 

 transformed using linear wave theory to the beginning of the model calculation 

 grid, located at a depth of about 8 m below mean sea level. 



490. The tide gage is located at the end of the pier and measures the 

 total water level variation with respect to mean water level. Thus, water 

 level measurements include both storm surge and tidal variation, the latter 

 being semidiurnal (two high and two low waters in a tidal day). 



491. Profile surveys were made at an average interval of 2 weeks, with 

 more frequent surveys when greater profile change took place. Although 

 profile data from the FRF represent one of the most detailed and accurate data 

 sets on profile change, horizontal spacing between measurement points along a 

 profile line is typically tens of meters. Small features along the profile 

 are not resolved, and the general shape has a more smoothed character than 

 actually exists. However, the data set is highly suited to the present 

 application. 



Calibration of numerical model with field data 



492. Parameter values given by calibration with the LWT data sets were 

 initially used in simulation of the field profile change data. However, it 

 became apparent that values of some empirical coefficients would have to be 

 modified to achieve agreement between measured and calculated profiles. Four 

 storm events (811022 - 811103; 811110 - 811116; 840210 - 840216; and 840403 - 

 840406) were chosen for calibration of the numerical model, and one event 

 (821207 - 821215) was used for verification. Calibration was performed by 

 minimizing the total sum of squares of the difference between calculated and 

 measured depths. The optimum transport rate coefficient K obtained for the 

 four events was smaller than the value obtained for the LWT data. As 



206 



