of waves and currents, as typified by this study, is one of the most compli- 

 cated and least understood processes. The sediment transport model employed 

 in this study uses fairly simple empirical formulas which are based on labora- 

 tory and field data. It reflects the inaccuracies inherent in the formulas as 

 well as the inaccuracies in the results of the three numerical hydrodynamic 

 models. 



127. In this study, a mean tide, and an average year's wave climate 

 based on 20-year averaging of wave statistics were used in running the sedi- 

 ment transport model to estimate the yearly shoaling rates in the navigation 

 channel. In reality, the tidal cycle is more complex, involving spring and 

 neap tides; and the wave climate varies from day to day, season to season, and 

 year to year. Severe storms such as hurricanes, which have a dramatic impact 

 on sediment transport and channel shoaling, have been excluded from this 

 study. As a result, sediment transport and channel shoaling rates in any 

 given year may deviate significantly from the values predicted in this study. 

 Moreover, short-term rates such as averages over a month or a season may dif- 

 fer markedly from average rates over a year. Even the nature of sedimentation 

 at a particular location may change from erosion to deposition and vice versa. 

 This change is exemplified by the field data on shoaling rates shown in Fig- 

 ure 20. Therefore, the results of this study will provide reasonable esti- 

 mates of the long-term yearly average values of sediment transport and channel 

 shoaling rates, provided severe storms are excluded and the uncertainty in the 

 results for the reach of channel between sta 325+00 and sta 399+74 is noted. 



128. In general, the uncertainty in the predictions of the sediment 

 transport model is reduced by verifying the model with field data from the 

 project site. For St. Marys Inlet, field data from navigation channel surveys 

 were available for about a year (November 1980 to December 1981) and are free 

 from the effects of extreme storms. The model verification, as shown in Fig- 

 ure 22, is good. In a sense, the verification shown is an indirect verifica- 

 tion of the modeling system approach as a whole. The yearly shoaling volumes 

 predicted by the model for existing conditions are comparable to yearly main- 

 tenance dredging volumes recorded by CESAJ. In general, with proper verifica- 

 tion numerical sediment transport models are better at predicting the effect 

 of a change from one condition to another, such as from base to plan, than at 

 predicting an absolute condition such as base or plan alone. In view of these 

 facts, it is estimated that Model B results on sediment transport are accruate 



75 



