Hydrological Drought Forecasting Using Stochastic Models (Case Study: Karkheh watershed Basin)

Hydrological drought refers to a persistently low discharge and volume of water in streams and reservoirs, lasting months or years. Hydrological drought is a natural phenomenon, but it may be exacerbated by human activities. Hydrological droughts are usually related to meteorological droughts, and their recurrence interval varies accordingly. This study pursues to identify a stochastic model (or ARIMA) that can be used for mathematical description and monthly and seasonality forecasting of severity of hydrological drought in the Karkheh watershed basin, southwest Iran. The models were applied to forecast hydrological droughts using standardized discharge index (SDI) time series. ARIMA modeling approach involves the following three steps: model identification, parameter estimation, and diagnostic checking. In the model identification step for the period (1974-2000), considering the autocorrelation function (ACF) and partial autocorrelation function (PACF) results of SDI data series, different ARIMA models are identified. The model gives the minimum Akaike information criterion (AIC), Akaike information criterion corrected (AICC) and Bayesian information criterion (BIC) is selected as the best-fit model. The parameters estimation step indicates that the estimated model parameters are significantly different from zero. The diagnostic check step is applied to the residuals of the selected ARIMA models and the results indicate that the residuals are independent, normally distributed with K-S test and Porte Manteau test. For the model validation, the predicted results using the best ARIMA models are compared to the observed data with F and Z test for the period (2001-2008). The predicted data show reasonably good agreement with the actual data. The predicted results using the best ARIMA candidate models were compared with the observed data.