Streamflow measurements are of utmost importance in water resources engineering. The amount of discharge through a particular section of river is the basic data required for planning and design of various structures and water resource projects. The discharge in a river is generally estimated by developing a relationship between stage and discharge as continuous measurement of discharge is very costly, time consuming and sometimes even impractical during floods. The accuracy of estimated discharge depends upon the accuracy of method adopted to estimate the parameters of stage discharge relationship. In this paper, parameters of stage discharge relationship have been estimated using Pattern Search (PS), Levenberg Marquardt Algorithm (LMA) and classical rating curve (RC) method for the stage discharge data available in literature (Herschy, 2008). The stage discharge relationship obtained by the three approaches were compared based on statistical parameters such as root mean square error (RMSE), mean absolute relative error (MARE) and Correlation coefficient. Application of optimization techniques results in lower root mean square error and mean absolute relative error and better correlation as compared to conventional regression method during calibration as well as testing.
Rating Curve, Optimization, Pattern Search, Levenberg Marquardt
