20 



procedure until the resultant tide at Newport (astronomical tide plus 

 hurricane tide component) was in agreement with the actual tide recorded 

 for 21 September 1938. As soon as a successful reproduction of the 

 Newport tide curve was attained, measurements of the resultant tides at 

 Providence and Somerset were made for comparison with prototype records. 



27. Maximum elevations reached by the September I938 hurricane 

 tide at Newport, Providence, and Somerset were 11.7 ft, I7.8 ft, and l^.f 

 ft, respectively, above mlx-r at Newport. The elevations reached at Prov- 

 idence and Somerset were therefore 6.1 ft and 3.1 ft, respectively, 

 higher than at Newport, including both gravitational buildup and wind 

 setup. Early computations of wind setup for the I938 hurricane tide in- 

 dicated this factor to be about 2.0 ft at Providence and less than 1.0 

 ft at Somerset, thus indicating the gravitational buildup to have been 

 of the order of k.O ft at Providence and between 2.0 ft and 3.0 ft at 

 Somerset. The initial test of the I938 hurricane tide in the model, 

 during which an elevation of +11. 7 ft mlw was reproduced at Newport, re- 

 sulted in maximum elevations of +15.5 ft at Providence and +1^.8 ft at 

 Somerset, or gravitational buildups of 3.8 ft and 3.1 ft at Providence 

 and Somerset, respectively. 



28. Since the results of the initial test were in close agreement 

 with the results of initial computations of wind setup for conditions of 

 the 1938 hurricane tide, a number of preliminary hurricane tide tests of 

 proposed barrier plans (through plan 26 of the model study) were made 

 for these conditions. However, the results of later and more refined 

 computations of the wind- setup component of the 1938 hurricane tide in- 

 dicated that the setup at Providence was of the order of 2.8 ft to 3.O 

 ft, instead of about 2.0 ft as indicated by the early computations. Use 

 of the refined wind-setup computation indicated that the gravitational 

 buildup of the 1938 hurricane tide between Newport and Providence was 

 about 3.1 ft to 3.3 ft, instead of the 4.0 ft shovm by the early computa- 

 tions and checked by the model diiring initial tests. 



29. The excessive gravitational buildup of the 1938 hurricane 

 tide in the model indicated a deficiency in model roughness, which con- 

 sisted only of a rough brushed finish of the concrete bed at the time of 



