Open AccessStreamflow forecasting of Astore River with Seasonal Autoregressive Integrated Moving Average model
Author(s) -
Rana Muhammad Adnan,
Xiaohui Yuan,
Özgür Kişi,
Yanbin Yuan
Publication year - 2017
Publication title -
european scientific journal
Language(s) - English
Resource type - Journals
eISSN - 1857-7881
pISSN - 1857-7431
DOI - 10.19044/esj.2017.v13n12p145
Subject(s) - akaike information criterion , streamflow , bayesian information criterion , autoregressive integrated moving average , mean squared error , autoregressive model , statistics , mathematics , coefficient of determination , environmental science , bayesian probability , time series , geography , drainage basin , cartography
Simulation of streamflow is one of important factors in water utilization. In this paper, a linear statistical model i.e. Seasonal Autoregressive Integrated Moving Average model (SARIMA) is applied for modeling streamflow data of Astore River (1974 – 2010). On the basis of minimum Akaike Information Criteria Corrected (AICc) and Bayesian Information Criteria (BIC) values, the best model from different model structures has been identified. For testing period (2004-2010), the prediction accuracy of selected SARIMA model in comparison of auto regressive (AR) is evaluated on basis of root mean square error (RMSE), the mean absolute error (MAE) and coefficient of determination (R2 ). The results show that SARIMA performed better than AR model and can be used in streamflow forecasting at the study site.