V. SUMMARY AND CONCLUSIONS 



1. Evaluation of Method. 



The sediment mass deposited and the shoaling rate (estimated by the method 

 described in this report) resulted in values 12 percent less than those meas- 

 ured in Dillingham Harbor during summer (ice-free) conditions. The method was 

 also tested using shoaling rates and other data collected at and near a large 

 sedimentation tank at Anchorage, Alaska (Everts, 1976a). The predicted shoal- 

 ing rate at Anchorage was 10 percent less than measured in the tank. In both 

 cases the model predictions were less than that which was actually measured; 

 for a conservative application of the model it is recommended that the M^ 

 value (total mass deposited; eq. 9) be increased by 15 percent. 



2. Settling Velocity Distribution. 



This crucial factor may vary through time in a rising water cycle, and by 

 time of year. Importantly, it may be highly dependent on aggregation of dis- 

 crete particles in saline waters. Thus, when compiling the settling velocity 

 distribution diagram, the velocities used should be those which actually may 

 be expected to occur at the harbor site. This means particle settling behav- 

 ior should be measured in the fluid where it was obtained. Aggregation can 

 occur in glacial-source sediments with low clay percentages such as found in 

 Alaskan estuaries (Everts, 1976c), as well as in sediment suspensions com- 

 prised of a high percentage of clay minerals. 



3. Sediment Mass Carried Into Harbor. 



The part of the sediment which enters the harbor and is subsequently 

 deposited will probably be near constant from one tidal cycle to another, 

 even though the total quantity which enters may vary widely (Everts and Moore, 

 1976; Everts, 1976a, 1976b). Everts (1976b), for example, reported that the 

 quantity carried into the basin at Dillingham Harbor varied by a factor of 

 three between tidal cycles, but that the part deposited was near constant at 

 81 percent (range: 76 to 84.2 percent for five tidal cycles). Thus, to accu- 

 rately predict M>, a number of tidal cycles or floodflow stages during a 

 river rise must be sampled to obtain representative c and c- values. 



4. Effects of Navigation Channel . 



Part of the material that enters a harbor may be sediment that was 

 deposited in the navigation channel at a low tide stage and was resuspended 

 during the next rising tide (Everts, 1976b). The amount entering from this 

 source is in addition to that which enters directly from the estuary and which 

 would be obtained from samples collected before the harbor was constructed. 

 The suspended-sediment data presented in the example problem, however, were 

 collected in the channel at the sill at Dillingham Harbor. Compared with 

 other samples collected in the estuary, apparently 60 percent of the sediment 

 carried into the harbor originated in the estuary, and 40 percent came from 

 resuspension within the entrance channel. This channel source should 

 obviously be considered. Cost savings can be achieved when the entrance 

 channel is designed as short as possible (Everts, 1977a). 



17 



