Abstract

Floods are natural disasters which cause losses that often occur in coastal cities such as Jakarta, Indonesia. Paying attention to rainfall and water flow is in the form of predicting floods. The research goal is to determine the best concentration time method, predict water level and flood inundation, and reduce the impact of flooding for people in the downstream area of the river, Jakarta. The method used in this research is quantitative descriptive analysis, rational approach, time of concentration, artificial neural network (ANN), and the use of the HEC RAS application. The results show that the arrival time of the flood using the Kirpich method shows a concentration time of 12 hours 5 minutes, Travel Time of 12 hours 10 minutes, Velocity of 11 hours 21 minutes and Nakayasu of 9 hours 1 minute. Modelling using an artificial neural network shows that the modelling results can function to predict water levels in the future. The results of the artificial neural network that estimates the water level for the 2020 period are 5.1 m (January), 4.94 m (February), 3.89 m (March), 3.52 m (April), 3.71 m (May), 2.58 m (June), 2.59 m (July), 24.1 m (August), 2.97 m (September), 3.32 m (October), 3.07 m (November) and 3.68 m (December).