are 5-cubic centimeter pipettes into which the samples are drawn by suc- 

 tion. Each cell is equipped with two platinum electrodes and a thermo- 

 couple. Separate conductivity cells are used for samples between 0.0 

 and 1.0, 1.0 and 10.0, and 10.0 and 40.0 parts per thousand; the cells 

 are accurate to within ±0.03, ±0,30, and ±0.50 parts per thousand, 

 respectively. The cells are frequently calibrated to ensure accuracy, 

 but the calibration has been determined quite stable. The accuracy to 

 which a model can be expected to duplicate salinities at any given point 

 from cycle to cycle for identical conditions is about ±3 to 5 percent. 



(20) Dye Concentration Measurement . Fluorescent dye tracers are 

 used to determine dispersion patterns and rates. Pontacyl Brilliant Pink, 

 Uranine and Rhodamine WT dyes are most commonly used in models. Dye con- 

 centrations are determined to within about ±3 percent by a Turner Model 

 III fluorometer. Since this instrument is sensitive to temperature 

 changes, all samples should be analyzed at a uniform temperature. In 

 most dye-dispersion tests, the dye is released at a point source and 

 then moves in high concentration clouds along the model for several tidal 

 cycles until the dye spreads throughout the estuary. These clouds move 

 away from the injection point with the ebb and flood currents and form 

 detached areas of comparatively high dye concentration which are dis- 

 cernable for several tidal cycles thereafter. Therefore, dye concentra- 

 tions measured at points 1 foot apart in the model differ greatly. The 

 same phenomenon occurs in nature and is probably responsible for the 

 reported difficulty in analysis of results of similar full-scale studies 

 in the field. 



(21) Temperature Measurement . During heat-dispersion tests, 

 water temperatures are measured either by an extensive array of thermo- 

 couples and multichannel recorders or by thermistors with individual 

 or multichannel recorders. Temperature differentials are accurately 

 measured to within ±0.5° Fahrenheit (±0.3° Celsius) with the thermo- 

 couples or ±0.2° Fahrenheit (±0.1° Celsius) with the thermistors. 



(22) Sediment at ioi 'm easurement . At the conclusion of a fixed-bed 

 model shoaling test, the moi- 1 sediment deposited within the limits of 

 marked areas (e.g., 1,000-fo. -long sections of a navigation channel) is 

 recovered with the apparatus discussed previously. After most of the 

 water has been separated from the model sediment, the material is then 

 measured volumetrically in graduated cylinders. The limit of accuracy 



of repeating identical shoaling tests is about ±10 percent. Assuming 

 that a satisfactory shoaling verification has been achieved, the model 

 and prototype shoaling rates can be related. Although the model shoaling 

 test results must be considered qualitative, the model predictions will 

 give a reasonable indication of prototype shoaling rates. 



4. Model Verification . 



The worth of any model study is completely dependent on verification 

 of the ability of the model to produce, with a reasonable degree of 

 accuracy, the results which can be expected to occur in the prototype 



96 



