Warning of drought by using arima model in the Nakdong river basin in Korea

Abstract

    Drought tendency is significantly increasing and highly negative influencing causing of extreme impact of the over global climate change in the future and especial at river basins that is facing severe challenge for all water resources managers in the world. The early realization of drought event can help managers to have drought mitigation strategies, which is significantly able to damage reducing in river basins. Therefore, the main objective of this study is to provide methodology to assess and warn of drought by using of the Autoregressive Integrated Moving Average (ARIMA) model. Based on the rainfall in the basin, the Standardized Precipitation Index (SPI) series were determined by using fitting a probability density function, then the model was applied to simulate drought using this index series in the river basin. The simulated results obtained by using from the best model's parameters after they were calibrated and compared with the actual data, and it showed reasonably good agreement with the actual data. So the models can be used to warn droughts with reasonably accuracy. Besides, the results of the study also was indicated occurrence drought levels based on SPI index that were shown that the years 1990, 1992 represented severely dry and double of consecutive years of 2000-2001, 2003-2004, 2008-2009 represented extremely dry conditions in basin. Through results of the study provide useful methodology for managers to have drought mitigation strategies, which are significantly able to damage reducing, and suitable policies of water resources management in region