AN INVESTIGATION OF REPRESENTATIVE DISCHARGE AND ITS IMPACT ON CHANNEL PLANFORMS

Abstract

The representative discharge of a channel having variable flows has been the subject of interest among researchers since several decades. Estimation of representative discharge in a channel had been made using various approaches such as bankfull discharge (based on concept when water overtops the river banks), the effective discharge (discharge which over a period of record transports the maximum sediment) and a recurrent interval discharge. In this study efforts have been made to collect and analyse published data as well as data collected for four river reaches of Jhelum and Kunhar Rivers in Pakistan as part of this study. The four approaches used in this study for estimation of effective discharge include; i) class based rating curve approach; ii) class based average approach; iii) half load approach and iv) event based approach. The estimated discharges using the above approaches for all four reaches were compared with bankfull discharges and 1.5 year recurrent interval discharges. A modified approach of calculating effective discharge termed as representative discharge (Qeff,rep) has been proposed and validated using data of four river sites collected in this study. Analysis of published data revealed that relation between effective discharge (Qeff), bankfull discharge (Qbf) and 1.5 year return period discharge (Q1.5) is strong. The relationship is stronger between Qeff and Qbf as compared to Qeff and Q1.5. For calculating the class based effective discharge flow data need to be distributed among various classes and the class sizes are determined by dividing the discharge range to the total number of classes of interest. This study showed that influence of class size on the estimation of effective discharge is significant and influences both magnitude of effective discharge and the nature of the histogram. The detailed analysis of the histogram suggested that unimodal (expected theoretically) histogram seems to be achieved for class sizes from 20 to 35 with average value being 25. The comparison of various approaches for estimating effective discharge with bankfull discharge showed that class based rating curve approach values are maximum and event based approach values are minimum. The ratio of effective discharge to bankfull discharge in this study ranges from 1.09 to 0.73 using various approaches. The assessment of modified approach revealed that it produces a reasonable similarity with measured values of bankfull discharge and class based rating curve method and moreover is unaffected by the variation in number of classes and hence reduces the subjectivity in calculating effective discharge. An effort was also made to investigate the changes in river channel planform due to variable flows in natural streams. The morphological data of 132 km long alluvial reach of River Indus in Pakistan for 32 years (1979 to 2010) were collected using satellite images and record of the relevant departments. This data indicated that overall changes in sinuosity of main river channel were not significant. However there have been instantaneous changes in sinuosity up to the order of ±5%, which could not sustain on long term basis. The average rate of migration of meanders in this reach was 68 meters per year for the same period. The analysis further indicates that there was a wide range of radius of curvature to channel width ratio (r/b) in this reach. The r/b values ranged from 1.3 to 8.4 and analysis indicated that maximum migration of meanders occurred at r/b value of about 2.6. An empirical model has been developed to predict the meander migration distance based upon the data set of 132 km long river reach of Indus River. The developed model is producing comparable results with the measured migration rates.