maxiiaura probable runoff from the streams tributary to Narragansett Bay. 

 A hjc'rologic study of the August 1955 storm, transposed over the Narra- 

 gansett Bay area, was made to compare the runoff from a storm producing 

 excessive rainfall but not occurring coincident with tidal flooding. 

 The results of this study are svuranarized in Table C-Ij., 



The maximum precipitation for the September 1938 stoTTn vras con- 

 centrated over Portland (Buck), Connecticut, about one mile north of 

 Iliddletovm, where a total of 17 inches was recorded for the period 

 September 17-21, This storm was transposed over the Narragansett 

 Bay ?rea as studies by the U, S, Weather Bureau indicate it co\ald as 

 easily have been located there, 



G-7. DESIGN FLOOD FOR FOX POIOT BARRIER 



For the Fox Point Barrier, the September 1938 storm was centered 

 over the Woonas qua tucket and Iioshassuck Rivers to produce the maximum 

 amount of inmoff at thiis point. The rainfall distribution was based 

 on the Corps of Engineers' "Maximum Average Depth-Duration of Rain- 

 fall" for New England over a drainage area of 100 square miles (Plate 

 C-3), The total accumulative rainfall for a 2ij.-hour period was 9*5 

 inches. An infiltration rate of 0,2 inches per hour was used for 

 rural areas and 0,1 inches per hour for city drainage. Unit hj'dro- 

 graphs were developed for each river. The inflow hydrograph for 

 the transposed September 1938 storm at Fox Point, including the 

 contributions from the '.Voonas qua tucket and Iioshassuck Rivers and 

 the "local" area is shown on Plate C-5« 



a, Woonasquatucket River . The Woonasquatucket River, with 

 a drainage area of 52 #3 square miles, was divided into tliree areas; 

 (l) the 2li,5 square mile c.:-ea above the Woonasquatucket Reservoir 

 at Ljrmansville, Rhode Island, (2) the 20,5 square mile area between 

 the dam and the city limits of Providence, and (3) the 7 •3 square 

 mile city area. The peak rxinoff from the reservoir v:as assiimed to 

 be 100 cubic feet per second per square mile. To estimate the 

 nmoff from the 20,5 square mile area, a 3-hour unit graph was de- 

 veloped from the September 195U flood (the flood of record for this 

 stream) ba^ed on the runoff between the reservoir and the U, S, 

 Geological Survey gaging station at Centerdale, Rhode Island 

 (drainage area = 38,3 square miles). The unit graph for the city 

 area was computed synthetically, using Snyder's coefficients. The 

 rainfall excess for the September 1938 storm was arranged to give 

 the maximum peai: discharge when applied to the composite unit 

 hydrograph of the Woonasquatucket River at its mouth (see Table 



c-U) , 



C-6 



