Daily and monthly lumped parameter hydrologic models for analysis of small watersheds in Sri Lanka

Abstract

Water resource management has become vital with the increase of demand for fresh water and modelling of ungauged basins under the conditions of data scarcity has always been a challenge. The two-parameter (2-P) and fourparameter (4-P) models have been used in this study to simulate the response of two river basins Maha Oya and Kelani River in Sri Lanka in an attempt to investigate the potential of these lumped parameter models. Both daily and monthly rainfall and streamflow data sets have been used to evaluate the model performance. Two key performance indices, Pearson Correlation Coefficient (PC) and Mean Relative Error (MRE) have been used as objective functions to compare goodness of fit of the observed and simulated streamflow. The two models were calibrated and validated for 2011~2014 and 2008~2010 periods, respectively. Based on the comparison of the results, it was found that the 4-P model provides a greater accuracy for the evaluation of the streamflow of the two selected river basins with a PC > 0.75 in both basins with both daily and monthly data. Even the 2-P model provided results with PC > 0.70 when used with monthly and daily data and within the acceptable range.