A thesis submitted In partial fulfillment of the requirements for the degree of Master of Engineering, School of Engineering and Technology

Thesis (M.Eng.) - Asian Institute of Technology, 1992

The main objective of the study is to evaluate the relative performance of the three selected models----Regression model, NAM model and Unit Hydrograph model based on their applications on two small catchments----Lan Yang stream catchment in Taiwan (area= 824 km2) and Khlong Si Yat catchment in Thailand (area= 951 km2), An investigation to test the validity of Regression model and NAM model in simulating the annual flow, baseflow and peak flow and the validity of Unit Hydrograph model in simulating direct runoff volume and peak flow has been carried out. The models were calibrated using daily streamflow, rainfall, evaporation data (only for NAM model in Lan Yang stream catchment) and monthly evaporation data (for NAM model in Khlong Si Yat catchment). Independent data were used to assess their predictive capabilities. Some statistical criteria were used to evaluate the model performance, e.g. computed and observed mean, standard deviation, correlation coefficient, coefficient of efficiency, root mean square error and relative error of computed streamflow volume. Due to the high value of basin slope and average runoff coefficient and more representative average areal rainfall obtained in Lan Yang stream catchment, the rainfall data is the most important factor in the regression model, flatter basin slope, lower average runoff coefficient and poor drainage and slow permeability of soil characteristics caused the discharge data be the most important factor in Khlong Si Yat catchment. It is found that the performance of the regression is similar and applicable in the two catchments Due to more representative rainfall and evaporation data obtained in Lan Yang stream catchment, the predictive capability of NAM model is better than that in Khlong Si Yat catchment, but may be due to the lack of knowledge about the catchment characteristics and of simulating model parameters and inaccurate measurements of the response, the results of this model in the two catchments are not as good as the regression model. Due to smaller sample space and longer rainfall duration used in Unit Hydrograph model, the performance of the model is similar to NAM model and the predictive capability is also similar and not very good in the two catchments.