Predicting Drought using Statistical Methods and Large-Scale Climate Signals

Management of water resources is essential in arid and semi-arid regions. Pre-knowledge about the amount of possible precipitation is important in planning water recourses, management of agriculture and droughts. Previous studies show that large-scale climate phenomena influence on the of climate and amount precipitation in different part of the world. In this study, first, among the 45 climate signals, 8 Index were selected as the most effective indicators; the total encompasses 81% of the variance in the principal component analysis (PCA) method. Subsequently, the correlation of large-scale climate signals in monthly Standard precipitation index (SPI) (one, 3, 6 and 12) of Maharlu-Bakhtegan basin simultaneously and the delay has been analyzed by using of cross correlation. Finally, multivariate regression equation was developed to predict. The results of cross correlation method showed that more of indices are significant with time lag with standardized precipitation index. Taylor diagram and error parameters showed that performance of regression equations for the scale of one month is better than the other scales.