where L^^^ is the length of the inlet and <irnf2x ^^ ^^® maximum water 

 depth in the inlet. A longer time step can be used for most tidal in- 

 lets, and as an upper limit, the time step should be one-hundredth of 

 the forcing wave period. 



Step 6 . Document all input data using the computer format shown in 

 the appendix. As a first estimate, set the flood and ebb entrance and 

 exit loss coefficients to equal one (CDF = 1.0 and CDE =1.0). As a 

 first approximation. Manning's n can be evaluated by the relation: 



n = CI - C2 D , (4) 



where D is the local inlet Stillwater depth. For depths greater than 

 4 feet and less than 30 feet, CI = 0.03777 and C2 = 0.000667; for depths 

 less than 4 feet, CI = 0.0550 and C2 = 0.005. The n for each grid may 

 be different if C2 ^ 0. 



Step 7 . For use with periodic forcing, run the program for several 

 sinusoidal cycles having the period and amplitude of the long wave of 

 interest to approximate the hydraulic characteristics of the inlet-bay 

 system. A sinusoidal tide is specified in the model by giving the forc- 

 ing period, T, in hours and the wave amplitude, A^ , in feet, on card 

 type 3 and by setting NPTS = on card type 8 of input to the program 

 INLET. Set ITABLE = 1 to obtain tables of instantaneoiis hydraulics at 

 points throughout the water level cycle and set IPLOT = 1 to obtain a 

 plot of predicted inlet velocities and discharge at sequential bay levels. 

 These outputs will indicate the importance of the terms in the equation 

 of motion describing water motion in the inlet. If temporal acceleration 

 is small during most of the water level cycle, then startup transients 

 will be small and the first or second cycle will contain little transient 

 effect (NCYCLES = 1 or 2 in input data) . However, if temporal accelera- 

 tion is significant during more than 25 percent of the cycle, approxi- 

 mately four cycles of model operation are required to eliminate startup 

 transient effects (NCYCLES = 4) . For aperiodic use such as with storm 

 surges or rapidly varying wave size (e.g., tsunamis), run the model for 

 the water level for approximately 10 hours before the time of interest 

 to build up initial conditions in the model similar to the prototype. 



Step 8 . Calibrate the computer model by varying Manning's n or 

 flood- and ebb- loss coefficients. The seawater level fluctuation can be 

 specified as a sinusoidal wave or in terras of an equal time series. For 

 an equal time series, start and stop the series when the seawater level 

 is at zero so that one or more complete cycles are described. Use at 

 least 20 points to describe each cycle. The sampling interval in minutes, 

 TDEL, and the number of points, NPTS, must be specified on card type 8 

 and the water level data on card type 9. 



The model is calibrated using short periods of field observations by 

 first comparing observed and predicted mean water velocities, if avail- 

 able, at the minimum cross-sectional area region of the inlet. If the 

 predicted velocities are higher or lower than observed, then the value 



10 



