نوع مقاله : مقاله پژوهشی
نویسندگان
1 استاد دانشگاه تبریز، دانشکده کشاورزی، گروه مهندسی آب، تخصص: منابع آب
2 گروه مهندسی آب، دانشگاه تبریز
چکیده
کلیدواژهها
موضوعات
عنوان مقاله [English]
نویسندگان [English]
Drought prediction is of great importance in scientific planning and optimal management of water resources in each region. In this study, modeling and forecasting of drought index (SPEI) time series were carried out for three stations located in Ardabil province using ARMA and SARIMA models. For this purpose, meteorological data from Ardabil, Khalkhal and Parsabad stations in the statistical period (1992-2021) were used. The SPEI index was calculated at each station at different time scales, and the first 25 years of data were used for calibration and the last 5 years for validation. The most appropriate model for each station was selected. The results showed that at all three stations, the SPEI1, SPEI3 and SPEI6 time series had a strong seasonal trend, with a 12-month periodicity. The importance of the seasonal trend decreased with increasing time scale. So, no seasonal trend was observed in the SPEI12 series. At Ardabil station, the most accurate prediction among all time scales belonged to SPEI3, which was evaluated with the SARIMA(0,0,2)(3,1,0)12 model and the criteria RMSE=0.25, NS=0.94 and AIC=-144.94. At Khalkhal station, the most accurate prediction related to the SPEI3 series was evaluated with the SARIMA(1,0,2)(3,1,0)12 model and the criteria RMSE=0.3, NS=0.91 and AIC=-185.52. At Parsabad station, the best prediction related to the SPEI6 series was obtained with the SARIMA(3,0,0)(0,1,1)12 model and the criteria RMSE=0.33, NS=0.89 and ACI=-251.5. The results of this study can be useful in the management of water resources of Ardabil province.
کلیدواژهها [English]
)