Ch. 17] PRYORS ISLAND REACH, OHIO RIVER 303 



and for securing proper flow lines through the model. Operation 

 of the model was divided into two distinct phases: (1) adjustment 

 of the model and verification of its accuracy; (2) development and 

 testing of improvement plans for solution of the shoaling problem. 



The verification of a movable-bed model is a highly important 

 phase of such a study, inasmuch as the validity of model indications, 

 when transposed to the prototype, rests entirely on this procedure. 

 The time, slope, and other scale relationships necessary to insure 

 proper model bed-load movement are determined largely by the trans- 

 portability of the available and usable bed materials and cannot be 

 closely established by mathematical means. Hence the model verifi- 

 cation, which might be termed an empirical determination of scale 

 relationships, must be depended on to establish the reliability of 

 model data. Verification of the Pryors Island model consisted of a 

 series of tests during which the time scale, slope, bed load, and rough- 

 ness were progressively adjusted by trial and error until the model 

 would reproduce accurately the hydrographic changes that were 

 known to have occurred in the river between two surveys made 5 

 years apart. This was accomplished by molding the movable bed 

 to the configuration shown by the earlier survey and then reproducing 

 in the model the sequence of stages and discharges that occurred in 

 the river during the 5-year period, all artificial changes, such as dike 

 construction and dredging, being reproduced in chronological order. 

 The model verification was considered satisfactory when a survey of 

 the movable bed, made at the end of this operation, showed the bed 

 configuration to be an accurate reproduction of river conditions shown 

 by the later survey. It could then be assumed that any changes 

 brought about by training works tested in the model would be a true 

 indication of what might be expected in the prototype if similar works 

 were installed. The model verification fixed the procedure for all sub- 

 sequent tests of possible improvement plans, and this procedure was 

 followed exactly during all tests. Thus the results of all tests were 

 directly comparable with one another. 



The river survey of 1930 (Fig. 2) shows the condition of the river 

 before the adoption of effective remedial works and reveals intermit- 

 tent shoals from above Pryors Island to below Sisters Island, a shoal 

 above Stewarts Island, and a dangerous crossing above Dog Island. 

 Twenty different plans or modifications, generally involving channel 

 realignment and the construction of stone spur dikes, were tested in 

 the model to arrive at one that would alleviate the shoaling problem 

 in these areas. Figure 3 shows the recommended plan of improve- 

 ment which was developed in the model. According to this plan, the 



