Long‐term probability of drought characteristics based on Monte Carlo simulation approach

In this study, an advanced stochastic procedure was applied to establish long‐term probability of different drought characteristics in 10 different parts of the world with various climatic conditions. Such purpose cannot be accomplished with a single historical data record, because the historical data record, no matter its length, represents only a realization of all such possible sequences of similar length. For this reason, a stochastic procedure was extensively used to generate alternative sequences of rainfall data in the selected 10 rainfall stations. Then, the generated synthetic rainfall sequences were served as a basis for drought monitoring using the Standardized Precipitation Index (SPI), which produce a set of realizations of possible drought characteristics. The long‐term probability distribution of non‐exceedance of drought duration, intensity, inter‐arrival time, density function of wet and dry periods, empirical joint probability of drought duration and severity, and transition probabilities of drought events were considered and developed as very useful guide on drought management and investigation. The results indicated the essential advantages of proposed methodology (Monte Carlo procedure) in long‐term drought characteristics investigations. Comparison of Probability Density Function (PDF) of SPI values based on historical and generated data at two selected stations (EGY_M and ZAF_PE stations).