-

An increase or decrease in the precipitation intensity compared to the normal conditions may lead to a higher risk of floods or a higher probability of incidence of drought. Knowledge of the probability distribution of precipitation is essential to the management of water resources. Such knowledge can be achieved using dynamical or statistical approaches. In the present article the probabilities of different drought states (very dry, dry, semidry, normal, semi wet, wet, and very wet) are calculated based on the annual and seasonal precipitation amounts in Ghazvin using the normal distribution and Markov chain methods. The results indicate that the probabilities of drought/wet states based on annual and seasonal precipitation data as calculated using the two methods are comparable. The statistical evaluation of the results using , however, reveals that the Markov chain method is superior to the normal distribution.
Keywords: Drought, Statistical models, Markov chain, Normal distribution