Prediction of Iran's Water Resources by Singular Spectrum Analysis Approach

According to the severe shortage of water resources in Iran, water resources forecast is one of the most important issues in the national policies. The purpose of this study was to evaluate the efficiency of the Singular Spectrum Analysis model in forecasting the amount of groundwater resources in Iran versus ARIMA model. Singular spectrum analysis is a method which is suitable for analysis of nonlinear and stationary time series. For this purpose, water resources time series from 1983 to 2015 were used annually and the short-term and medium-term forecasts obtained from the two models were compared. According to the results, the SSA method was able to perform better in short and medium term predictions compared to the ARIMA model. Correspondingly the results showed 70% improvement in prediction of one step ahead up to 88% improvement in prediction of 3 steps ahead.