Time Lagged Back Propagation Neural Network with rainfall for flood forecasting

Flooding is a major problem in many countries while flood models have been proposed by various research groups on hydrological model in the past. Recently, Artificial Neural Network has been applied to the flood model. However, the learning factors of Neural Network for flood models need to include water level and rainfall which are two major impacts to the flooding. Therefore, this paper proposed Time-Lagged Back Propagation Neural Network Model with water flow and rainfall for flood forecasting. We ran our model with water flow and rainfall at the rivers around the lower part of northern region of Thailand. We compared the results between Back Propagation Neural Network Model and Time Lagged Back Propagation Neural Network Model with Gamma Memory. The study shows that Time Lagged Back Propagation Neural Network Model with Gamma Memory has better performance.