in this area. No other changes in bottom friction values were necessary to 

 achieve reasonable agreement between model calculations and measurements. 



172. Two types of boundary conditions were used at the external bound- 

 aries. The Type-1 condition, values of water surface elevation specified as a 

 function of time, was used at Node 1, the sea boundary, and provided the 

 forcing for tidal flow. Therefore, at Node 1, the model expects a table of 

 values of water surface elevation as a function of time. This table is 

 supplied in the input data file EXTER.DAT (Appendix B) . 



173. The Type-2 boundary condition, discharge as a function of time, is 

 applied at Nodes 25 and 31, which represent the end nodes in Indian River and 

 Rehoboth Bays. Since neither the discharge nor water surface elevations were 

 recorded at these boundaries, a zero discharge was used as the boundary 

 condition. This is a good assumption because the magnitude of velocity at 

 these locations would be very small because of the large bay areas. 



174. At Cross Sections 4, 5, and 6 the transition loss coefficient was 

 initially set to 0.5 and then revised after the first few model runs to 0.4 to 

 simulate head losses at the pilings of the highway bridge and constriction of 

 the channel in general. In models not having the capability to simulate 

 transition head loss, a larger value of the bottom friction coefficient would 

 have to be used; however, increased bottom friction often results in an 

 underestimate of velocity. 



175. In the present study, tidal elevations available for hours 39 

 through 63 of a 63-hr-long measurement record were used in the calculations 

 and comparisons. Locations of cross sections and the single junction were 

 determined by inspection of the hydrographic survey charts, and needed depths 

 were read from the charts, visually interpolating and extrapolating readings, 

 as necessary. Total time required to set up DYNLETl for modeling Indian River 

 Inlet was approximately 50 person hours , extra time being required to check 

 the depth survey readings. Approximately 1 day was required to run the model, 

 graph and output results, make some slight changes for calibration, as 

 discussed in the following paragraphs, and conclude that the model was 

 accurately simulating tidal flow velocity, stage, and discharge. 



176. Execution time for the simulated 63-hr period with a time step of 

 1,800 sec (30 min) for Indian River Inlet (consisting of three channels, one 

 junction, and 31 nodes) took 187 sec (Pascal -language version of the program) 



69 



