and finally, the water level at this discrete point in space and time 

 is given by Equation 3-90 as 



At / _ _ \n+l 

 i+'A 

 (0.2) (0.0915 -0.0513) (5280) 



^"4 = C«-^(Q,-Q«)' 



= 3.32 + 



(10) (26.3) 



= 3.48 feet, say 3.5 feet . 



The significant digits indicated in the above computations do not 

 reflect the accuracy of the numerical procedure, but are retained to 

 reduce the accumulation of round-off errors. 



The wind setup hydrograph for the ends of the lake as determined by 

 computer is shown in Figure 3-63. The storm winds blew in the general 

 direction toward Buffalo at the northeastern end of the lake as indi- 

 cated by the setdown at Toledo and the setup at Buffalo. The spatial 

 steps Ax taken are quite large; smaller increments in Ax would give 

 a more accurate estimate. Calculations with the mathematical model 

 were initiated with the system in a calm state, i.e., Q = S = 0. 



The wind setup profile along the major axis of the lake determined 

 by the numerical scheme is shown in Figure 3-64 for three time periods 

 during the storm. The nodal point where the computed water surface 

 crosses the Stillwater level occurs near the center of the lake, but 

 this nodal point can vary with time. 



Although the comparison of the computed and observed wind setup is 

 not in complete agreement, particularly at the beginning and late stages 

 of the storm, the method gives reasonable results for the wind setup 

 amplitudes. To engineers, it is frequently the maximum departure of 

 the water level from its normal position that is of greatest concern. 

 Results of the simplified model should be interpreted with care, since 

 many of the physical processes which may be significant have been neg- 

 lected. Wind and bottom stress laws, in particular, are oversimplified 

 for the Lake Erie problem. Better agreement can be expected with two- 

 dimensional schemes such as the one developed by Platzman (1963) , since 

 they more accurately model for the physical processes involved. 



(3) Storm Surge in Semienclosed Basins . It is generally im- 

 possible to make reliable estimates of storm surge in semienclosed basins 

 (bays, and estuaries) with less exact procedures such as those described 

 for specific problems within enclosed basins or on the open coast. This 

 is because bays and estuaries are nearly always irregular in shape, and 

 basin geometry is often further complicated by the presence of islands, 

 navigational channels, and harbors. Moreover, many of these basins have 



3-140 



