FLOOD FORECASTING MODEL USING EMPIRICAL METHOD FOR A SMALL CATCHMENT AREA

Abstract

The two most destructive natural disasters in Malaysia are monsoonal and flash floods. Malaysia is located in the tropical area and received on average, around 2800 mm of rainfall every year. Due to this high amount, a reliable and timely flood forecasting system is necessary to provide early warning to minimize the destruction caused by flash flood. This study developed and checked the adaptability and adequacy of the flood forecasting model for 93 km2 catchment area, Kampung Kasipillay, in Kuala Lumpur. The Empirical Unit Hydrograph Model was used in this study and past rainfall data, water level and stagedischarge curve were used as inputs. A Rainfall-Runoff Model (RRM) which transforms the rainfall to runoff hydrograph, was developed using excel. Since some data, such as properties of the watershed, are not always complete and precise, some model parameters were calibrated through trial and error processes to fine-tune the parameters of the model to get reliable estimation. The simulated unit hydrograph model was computed in prior runs of the flood forecasting model to estimate the model parameters. These calibrated parameters are used as constant variables for flood forecasting model when the runoff hydrograph was regenerated. The comparison between the observed and simulated hydrograph was investigated for the selected flood events and performance error was determined. The performance error achieved in this study of 15 flood events ranged from -2.06% to 5.82%.e.

Keywords

• Flood Forecasting
• Empirical Unit Hydrograph Model
• RainfallRunoff Model (RRM)