Deterministic Modeling for Prediction of Sediment Yield from Watersheds of Hirak-ud Catchment

Abstract

In this paper, an attempt is made to develop a dimensionally homogeneous and statistically optimal model for predicting sediment production rate (SPR) from the ungauged watersheds of Hirakud catchment (Orissa portion) by regressing annual SPR on geomorphic parameters. For this purpose, SPR was made first dimensionless by dividing it with square root of the watershed area. It was then regressed with the uncorrelated geomorphic parameters as independent variables using multiple regression technique to obtain the desired prediction model. The procedure used for this was principal component analysis with varimax rotation. The model results in higher correlation coefficient and F-value. The validation result shows that model is able to predict the average annual SPR within the reasonable limits of the observed SPR