et al., 1973) shows the recorded water level with tidal signal 

 removed at Cedar Key during the passage of hurricane Agnes in 1972. 

 The 12 h period oscillation is obvious. 



The maximum surges on the order of 7-8 m on the Florida shelf 

 found in the PATH4 simulations are probably too large. Reid and 

 Whitaker (1981) showed that the response on the Florida shelf to 

 direct forcing by the M2 tide potential was very sensitive to the 

 local friction coefficient. Therefore, the simulation of HUR23 was 

 repeated with a friction coefficient of 7.5 x 10" m/s. This 

 coefficient is applied to the entire computing domain since only the 

 response on the Florida shelf is the primary concern in this 

 simulation. The maximum surge generated (6.23 m) is decreased by 

 about 25 % from that generated by the same storm simulation with the 

 smaller friction coefficient. The hydrographs from stations on the 

 Florida shelf (not shown) clearly demonstrated the effect of larger 

 friction. Examination of the remaining hydrographs reveals no 

 significcuat changes in water levels due to increasing friction, even 

 on the Texas-Louisiana shelf region. Presumably these stations are 

 located too far to the left of the storm track such that the 

 corresponding currents and bottom stress are small. 



The PATHS storm, HUR25, produced a 5.4 m maximum surge near 

 Galveston, Fig. 89. The sharp drop of water levels before the peak 

 surge is caused by a direct offshore wind stress. The maximum peak 

 of TiQ (Fig. 90) is less than 0.1 m .The estimated period of rjg for 

 this run is 32 h. 



149 



