摘要
Drought modeling is essential to water resources management and planning. In this study, Fourier spectral analysis is used to examine the cyclic structure for drought patterns and develop a long-term periodic model. A case study for historical precipitation data, obtained from the arid region of Kuwait for the period spanning from January 1967 to December 2009, are converted to drought measurements following the Standardized Precipitation Index (SPI) criterion. The SPI calculations are performed for two time scales of 12 and 24 months. The periodogram technique used for both time scales reveals periodicities of 12, 14, 19, 26, 31, 43, 64, 103 and 258 months. It is advocated here that the 26- and 258-month periods present in the data are attributed, respectively, to a Quasi-Biennial Oscillation pattern and a solar cycle over which the magnetic polarity of the sun first reverses then reverts to its former state. The detected periods are manipulated in the SPI model to produce drought forecasts, which suggest that until the end of year 2024 the climate is considered normal to very wet. This finding may be implemented to assess policy requirements related to water resources management.