21 



the initial tests. It was suspected from the beginning that the model 

 roughness was deficient^ but the very low ciorrent velocities throughout 

 the model for astronomical tide conditions, plus the close reproduction 

 of the gravitational buildup of the 1938 hurricane tide, had made addi- 

 tional roughness seem unnecessary. After the deficiency in roughness 

 became apparent the Manning "n" of the prototype channels was estimated 

 to be of the order of 0.026, and roughness elements were added to the 

 model as required to effect a scale reproduction of this estimated pro- 

 totype roughness. The model roughness elements consisted of three- 

 fourth- inch-wide metal strips set vertically into the concrete bed of 

 the model and extending to the water surface. An average of about one 

 strip per two square feet of model area was required to duplicate the 

 estimated prototype roughness. 



30. Use of the refined wind-setup computations for the 1938 hur- 

 ricane tide indicated that the gravitational buildup of this tide over 

 maximum elevation at Newport was slightly more than 2.0 ft at Somerset 

 and, as previously stated, was 3-1 to 3-3 ft at Providence. The 1938 

 hurricane tide was repeated in the model after completion of the rough- 

 ness adjustment described above and with the hurricane tide generator 

 adjusted to produce a maximum elevation of +11. 9* ft mlw at Newport. It 

 was found that maximum elevations at Providence and Somerset were 15-2 

 and li+.2 ft, respectively, above mlw, or a gravitational buildup of 3-3 

 ft at Providence and 2.3 ft at Somerset. Comparisons of prototype and 

 model gage records at Newport, Providence, and Somerset for conditions 

 of the 1938 hurricane tide are shown on fig. 13 . Astronomical tide 

 elevations and ranges shown on fig. 9^ s^J^cl current velocities at several 

 of the stations for which data are presented on figs. 11 and 12 were re- 

 checked to determine whether the change in model roughness had effected 

 changes in astronomical tides and tidal currents. It was found that no 

 measurable changes had occurred in astronomical tides and tidal currents 



* Model high tide was increased 0.2 ft over that recorded in the proto- 

 type to compensate for the effects of local wind setup which depress 

 water- svirf ace elevations near the bay entrance. 



