Artificial Neural Network Based Novel Flood Prediction Model: A Case Study in Rathnapura in Sri Lanka

Abstract

Most of the countries in the world face with natural disasters like floods, cyclones, landslides and droughts .Floods are major contributors to disruption of human lives ,economy and property damage. Also it can be strike with proper little warning or prediction. So, this study proposes a novel flood prediction model for Rathnapura district in Sri Lanka. Main reason for the flood in Rathnaura town is Kalu River. Rainfall of five meteorological stations namely Alupola station, Hapugasthenna station, Guruluwana station, Lelopitiya station and Rathnapura station are affected to the water level of the Kalu River. The methodology of this study is running under two main phases. In the first phase, K-mean clustering is used to cluster the water level of the Kalu River according to the rainfall of five meteorological stations. In the second phase, the Artificial Neural Network (ANN) model is successfully implemented for forecasting flood in Rathnapura town according to the rainfall of above mentioned five stations. Two model accuracy standards were employed. Such as mean absolute error and meansquare error. The novel ANN model gives the minimum error accuracies in both training and testing stages. Data set for this study is obtained from Department of Irrigation, Sri Lanka and it contains 1955 data. The new proposed model is useful to avoid or minimize the social and economic losses that may occur in the flood.