values are reasonable estimates from previous studies and also calibration 

 parameters. As initial values, 0.02 was used everywhere for deeper channels 

 and the inlet throat with sand bottoms, and 0.04 was used everywhere in the 

 interior channel cross -sections with rough boundaries expected because of 

 vegetation. Sensitivity testing showed that these initial values provided 

 adequate results for the purpose of this example; therefore, they were not 

 optimized. 



139. At the sea boundary, Node 1, measured values of water surface 

 elevation were specified as a function of time (Type-1 boundary condition). 

 This boundary condition was also successfully applied at all external boundary 

 nodes in Masonboro Inlet employing the 1974 data (Amein 1975) . 



140. At the end nodes located in Masonboro Channel, Shinn Creek, and 

 Banks Channel (Nodes 13, 19, and 25), Type-2 boundary conditions of zero 

 velocity were applied under the assumption that the discharge would be 

 negligible far from the inlet (on the order of 20,000 to 40,000 ft). This is 

 a fictitious boundary condition, and for most accurate results, values for 

 these distances should be determined in model calibration. This boundary 

 condition produces correct values and is a very reasonable method given the 

 absence of a well-defined bay or well-defined boundary. Sensitivity of model 

 results can be investigated by changing the distance from the inlet, but, in 

 the present example, the initially specified distances were maintained because 

 the calculated results were sufficiently accurate. 



141. A third boundary type. Type 3, may also be used at the external 

 boundaries. This boundary was employed in the lumped -parameter models of 

 GITI. The boundary condition Type 3 requires values of bay area, the bank 

 slope of the bay, and an empirical coefficient called the flood repletion 

 coefficient (Seelig, Harris, and Herchenroder 1977). In the case of Masonboro 

 Inlet, the bay area would be a fictitious number presumably based on the tidal 

 prism. This type of boundary condition may be specified at Nodes 13, 19, and 

 25 inside Masonboro Channel, Shinn Creek, and Banks Channel. A number of 

 trials with DYNLETl using this method showed that the "bay area" is actually a 

 calibration parameter, and good results could be obtained with this method if 

 an appropriate fictitious bay area was specified. The Type -3 results are not 

 discussed further here. 



53 



