Distributed model for changes in river peak flow due to land development

Abstract

Land development for new townships, agriculture or recreational areas is almost always accompanied by environmental problems. Some of these are soil erosion, soil fertility degradation, river sedimentation and sometimes flash floods. Exposed soil is subjected to the impact of raindrops, sealing the soil surface, reducing infiltration and causing high surface runoff. A question that always comes to mind is, with land development, can the river still carry the runoff rate from a certain rainfall event or is flooding imminent? Increased runoff results in higher flows during rainfall events, which in turn increases the number of times that a river floods the adjacent land areas. Likewise, this increase in runoff and channel flows can drastically increase the erosion of river channel beds and banks, potentially destabilising bridges or local structures. There is a need then for a tool to correlate the land development with the river peak flow. For this purpose a Landuse- peak flow model (LPM) was developed, the model was based on curve number (CN) method, which was derived from landuse and hydrological soil groups in each sub basin for the watershed. The model was developed for Upper Bernam River basin, Malaysia for both wet and dry seasons. Using the measured peak flow data the models were verified and tested. To evaluate the model performance in simulating the peak flow changes the correlation coefficient (R2), mean absolute error (MAE), root mean square error (RMSE), Theil’s coefficient (U) and model efficiency (E) were used as statistical tests. The models obtained 0.99, 0.004, 0.84, 0.04 and 0.98 for the rainy season and 0.93, 0.11, 0.61, 0.07 and 0.92, respectively. It was found that increases in CN by 0.3, 0.5, 0.7 and 1.2 % cause increases in peak flow values by 2, 3, 4 and 7 %, respectively. The changes in peak flow are constant regardless of the changes in rainfall pattern. The models were applied to investigate the changes in peak flow from the proposed land development for the study area in the year 2020. Due to the implementation of this plan, peak flow would be expected to increase by 80 % and 76 % for the wet and dry seasons, respectively. The models can be used to simulate the changes in peak flow from future landuse scenarios.