number of variables that must be adjusted. 



Offshore Model 



30. The tides of two time periods were simulated in the calibration of 

 the offshore model: 5-lU March 1962 (the March 1962 northeaster) and 11 and 

 12 September 1960 (Hurricane Donna). Predicted highs and lows for several 

 stations along the coast were taken from U. S. Coast and Geodetic Survey 

 (1962) tide tables for these two times to serve as prototype data. Four 

 stations (Capes Fear, Lookout, Hatteras, and Henry) yielded the initial esti- 

 mates for boundary conditions and served as the primary comparisons with WIFM 

 computations during the calibration process. Estimates of the M- amplitudes 

 and epochs for the grid boundaries were chosen so that Lhe computed one- 

 constituent tide would represent the entire predicted tide, and not just the 

 M- portion of the signal. In order to easily see the agreei.ent between WIFM 

 computations and the prototype data, computed marigrams were plotted against 

 spline-fit curves of the predictea highs and lows. 



31. Boundary conditions were first refined to match the predicted tides 

 of the March 1962 storm. The beginning of the WIFM computatioi. was chosen to 

 be midnight Greenwich Mean Time (GMT) on 5 March 1962. A 3-min time-step was 

 used in WIFM, and 111 hr of the tide were computed. Plate i shows computed 

 marigrams versus prototype curves for the four Cape stations. There are small 

 variances between the peaks of the computed and proi.otype tides due to the 

 lack of other constituents in the grid boundary conditions, but the two tides 

 match well for these gages. To check that the model correctly represents the 

 prototype tide at other points along the Outer Banks, other stations are 

 compark^.d with computations (Plate 2), Virginia Beach shows an excellent 

 match, while the calculated amplitudes at the Currituck Beach Lighthouse are 

 about 1/2 ft higher than predicted. This increase is caused by the shoreline 

 approximation made in the grid for the barrier island at Currituck, and by the 

 proximity of the WIFM gage to this shore. 



32. Plates 2 and 3 illustrate how the crude approximations of the three 

 inlets in the offshore grid (each inlet is only one cell wide) induce some 

 local amplification of the computed tide. However, the basic hydrodynamic 

 properties of the inlets are preserved in this model because the cross- 

 sectionil areas of the inlets match observations (Jarrett 1976). Tidal wava 



22 



