Applicability of alluvial roughness predictors for Gumti, Kushiyara and Mohananda rivers

Abstract

This study has been performed to determine the applicability of alluvial roughness predictors for Gumti, Kushiyara and Mohananda rivers. Eleven roughness predictors are evaluated- six are based on division of roughness and five are based on overall roughness. The selected methods are evaluated on the basis of Manning's n, Chezy's C and Darcy-Weisbach friction factor f and, where possible, on the basis of depth and/or velocity. To predict the roughness characteristics of alluvial rivers, hydrological data, e.g. discharge, cross-sectional area, mean velocity, river width, water surface slope and grain size data of bed materials are needed. The discharge stations Jibanpur, Sherpur and Chapai Nawabganj are selected for the rivers Gumti, Kushiyara and Mohananda, respectively. The required hydrological and grain size data of the selected rivers have been obtained from BWDB and SWMC, respectively. The Manning, Chezy and Darcy-Weisbach formulae are applied to calculate the roughness coefficients n, C and f, respectively. The roughness coefficients are also predicted by using the selected roughness predictors and then these values were compared with the values obtained by Manning, Chezy and Darcy-Weisbach formulae. Accuracy of different predictors has been determined on the basis of MNE and the standard deviation ofthe MNE. Out of the eleven selected roughness predictors, the Einstein and Barbarossa (1952) is the only method, which can satisfactorily predict roughness coefficients for full range of discharges for all the rivers. Some of the roughness predictors seem to predict Manning's n satisfactorily for the full range of discharge of a particular river, e.g. the Shen (1962) method for Gumti river. Shen (1962) method for the Kushiyara river predicts Manning's n about 100% higher than the calculated values for all discharges. Simons' and Richardson (1966) method predicts Manning's n about 48% higher for the Gumti river and 87% higher for the Kushiyara river than the calculated values for all discharges. Some methods predict Manning's n satisfactorily only for moderate to high range of discharge of a particular river, e.g. Haque and Mahmood (1983) and Shen et al. (1990) methods for the Kushiyara river. Some of the methods, e.g. Engelund and Hansen (1967), Van Rijn (1993), Garde and Ranga Raju (1977), Brownlie (1983), White et al. (1987) and Karim (1995) methods predict roughness coefficients of poor quality for any discharges and all the rivers. The Einstein and Barbarossa (1952), Shen (1962) and Brownlie (1983) methods for the river Gumti and Brownlie (1983) and Karim (1995) methods for the river Kushiyara predict depth satisfactorily. During analyses of the hydrological data collected from BWDB for the river Mohananda, the data did not seem reliable, because the calculated Manning's n from observed data vary from 2.316 to 0.081 which are about 30 to 3 times higher than from its normal ranges. So no conclusion could be drawn for the predictors in case of the Mohananda river. But Einstein and Barbarossa (1952) and Shen (1962) methods seem to predict resonable roughness coefficients for the full range of discharges and Haque and Mahmood (1983) method seems to predict resonable results for moderate to high discharges of the Mohananda river. The Manning's n for the Gumti and the Kushiyara rivers is ma.ximum when the discharge is low and decreases with increasing discharge. For the river Gumti detail analyses indicate the presence of bed forms in all discharges, i.e. throughout the year. For the river Kushiyara; bed forms are only present in the range of low to medium discharges but at high discharges the river bed becomes plane. The response of grain roughness to the change in discharge is relatively less than that for form roughness. The variation of grain roughness is responsive to that of the total roughness but the variation of form roughness is similar to the total roughness. The Manning's n for the river Gumti is maximum during the month of March when discharge is low; and attains minimum value during the months of May to October when high flow takes place. The Manning's n for the river Kushiyara is maximum during the month of February when the discharge is low and minimum during the month of July to September when the discharge is high. The Mohananda river attains the maximum discharge in the month of September and the minimum discharges in the months of December to May.