Implementation of Multiplicative Seasonal ARIMA Modeling and Flood Prediction Based on Long-Term Time Series Data in Indonesia

Author

Sri Supatmi, S.Kom.,M.T., D.Sc Irfan Dwiguna Sumitra, S.Kom., M.Kom. Ph.D

Abstrak

Timeseriesmodelingandpredictionhasfundamentalimportancein the various practical field. Thus, a lot of productive research works is working in this field for several years. Many essential methods have been proposed in publications to improve the accuracy and efficiency of time series modeling and prediction. This research aims to present the proposed prediction model namely Multiplicative seasonal ARIMA model (MSARIMA) based on non-stationary time series data to predicting the flood event. In this paper, we have described the performance of the ARMA model, the ARIMA model, and MSARIMA model to predict the flood event in the Region over Indonesia in 42 years (1976– 2017). We have used the two performance measures respectively (MAPE, and RMSPE) to evaluate prediction accuracy as well as to compare different models fitted to a time series data. The result has shown the proposed predicting model has the best performance accuracy than the others model in this research. Keywords: Flood