Laserfiche WebLink
<br />-'\ 'Il'~ <br /> <br />where Aa, Ba, Ca, Da, Ea, and Fa are experimental coefficents. <br /> <br />Values of the correlation coefficient of 0.991 and 0.981 were obtained <br />for the diagonal and square spacing of elements, using a linear response <br />surface for the pi-terms. When the two patterns were combined, the value was <br />0.967. <br /> <br />Using a quadratic model, the values were 0.997, 0.987, and 0.979. An <br />exponential model yielded values of 0.991, 0.970, and 0.968. <br /> <br />Garton reached several conclusion~ from his experiments. He found that <br />an increase in size and density of roughness elements increased the resistance <br />of flow in the channel. Resistance to flow in the channel decreased slightly <br />with an increase in slope, and a diagonal-grid pattern of roughness elements <br />offered less resistance to flow then did a square grid pattern. Finally, he <br />found that a linear model and an exponential model gave comparable results. A <br />quadratic model gave an improved estimate, but it was more complex to <br />calculate. <br /> <br />SUMMARY OF THE METHODS <br /> <br />All of the methods previously preaented were analyzed for their <br />suitability in determining n values for flood plains. After examination of <br />the four methods it was determined that two, the vegetation-density method <br />(petryk and Bosmajian, 1975) and the roughness-concentration method (Tseng and <br />others, 1974), were very similar. Both methods were based on the balance of <br />the momentum equation, where the ~ . ~ M~~ees of a densely vegetated flood <br />plain was equated to the obott_rO"'<)8_'.~f.oJ!.1ll..".,ugh- on the flood <br />plain. Both were derived from the - .ba~e, where ~_~_l.~~_~ is equal <br />to .n._'..:f_...~due to form .......,.beundalty..aheaP'-~rce. The vegetation-density <br />method was chosen for comparison with field data, because it determined the <br />roughness characteristics in the form of Manning's n, and the determination of <br />n was easily applicable to field data. <br /> <br />The estil\1ation procedure of ~ (1956) was found to b_-","-.-&ul in <br />deter.mining n values, especially for channels. The.... estimation method was <br />used by Mdr1l!g'e ..mt.'.Garret.t. (1973), who attempted to systematize the <br />selection of roughness coefficients for Arizona streams. They expanded and <br />modified Cowan's estimation procedure. <br /> <br />The regression-analysis method of Garton (1970) is not applicable for <br />field deter.mination of n; therefore, it was not pursued any further. <br /> <br />11 <br />