(19) 



where W is the 10-metre wind speed in metres per second. The surface 

 stress is calculated by Taylor's (1916) equation: 



T = p Cj^W^ (20) 



where p is the air density, assumed constant for this study. The total 



surface stress then is resolved into components for the F and F terras in 



X y 



VTFM. 



43. Due to the increase in flow velocities through Oregon Inlet during 

 the storms, the computational time-step in WIFM had to be adjusted for the 

 nearshore model. Both Hurricane Donna and the March 1962 northeaster were 

 simulated with a 3-min time-step in the offshore model (the same as for the 

 tide runs). The March 1962 computation begins at midnight GMT on 5 March and 

 goes to noon GMT on 9 March. The Hurricane Donna simulation begins at noon 

 GMT on 11 September 1960 and computes to 9 p.m. GMT on 12 September. For the 

 nearshore model, the time-step was reduced to 50 sec but the same starting and 

 ending times were kept. The March 1962 storm was sxraulated with a 60-sec 

 tine-step from hours 33 to 78 (9 a.m. GMT on 6 March to 6 a.m. GMT on 8 March). 

 The simulation of only 45 hr of this storm greatly reduced computational 

 costs. 



44. Two letter reports of SAW (Davidson 1961, Grygiel 1962) contain 

 the prototype storm surge data used in this study. Surge histories for the 

 peak periods (7 March 1962 and 11 and 12 September 1960) for most of the sites 

 in Figure 7 can be found in vhese reports. No prototype data were; available 

 at or near Oregon Inlet, but there are recorded marigrams available from 

 Rodanthe, Nags Head, Stumpy Point, and Point Harbor. These four sites all lie 

 within 25 miles of the ir.let, so that agreement between them and calculations 

 can indicate the accuracy of calculations at Oregon Inlet. Tabulated measure- 

 ments of the tide highs and lows for the months of March 1962 and September 

 I960 provided independent checks on the accuracy of the surge histories; these 

 two sources agreed completely. The tabulated data also gave an indication as 

 to where a "mean water level" may lie for each prototype gage. An average of 

 the month's highs and lows (excluding the storm period) seldom came close to 

 the datum listed with the tabulations. Differences are mainly attributable 



29 



