Prediction of Percentage Discharge Distribution in Non-Prismatic Compound Channels with Skewed Floodplains.

In rivers with non-prismatic compound cross-sections, due to the change in cross-section along the channel, mass exchange between the main channel and floodplains. Therefore, discharge distribution in non-prismatic compound channels is an important task for river and hydraulic engineers. In this paper, some results of experiments performed in non-prismatic compound channels with skewed and inclined floodplains have been explained. Two skew angles of 3.81o and 11.31o and three discharges were investigated. The effects of relative depth and relative distance on percentage discharge distribution in each sub-section of the skewed compound channels are presented. The experimental results show that the percentage discharge in each sub-section relies upon the parameters like relative depth, relative distance, skew angle, and floodplain side slope. By using the experimental results, multivariable regression models have been developed to estimate the percentage of discharge in the main channel and on the floodplains. Investigations indicate that the regression models presented in this research, in the validation range, can predict the percentage of discharge in each sub-section of the skewed compound channel fairy well. So that for the results used in this research, the coefficient of determination (R2) for predicting discharge regression model in the main channel is 0.96, on the diverging floodplain is 0.92, and on the converging floodplain is 0.91. Also, the mean absolute percentage errors (MAPE) between the calculated and measured value of percentage discharge in the main channel, on the diverging floodplain, and the converging floodplain are equal to 1.47%, 14.29%, and 21.7%, respectively. [ABSTRACT FROM AUTHOR]