Drought Modeling in Fars Province Using Time Series Analysis

Drought is one of the most important natural disasters that affects sectors with its high frequency particularly in dry and semiarid areas of Iran. Lack of raining will have various effects on agriculture, surface water and underground water resources, humidity of soil and rivers. Regarding the location of Fars Province on dry and semiarid climate areas and the incidence of repeated drought, especially in recent years, necessitates paying attention to drought simulation and forecasting. Quantitative and description of environmental process and the anticipation and estimation of its behavior at the time and situations without observation and sample requires having a pattern and model of that process. One of the methods to access the goal of modelalization in drought incidence is the use of time series model. In the present study, the precipitation data of 90 pluviometry and synoptic stations in Fars province were used with regard for 30 years. To determine the grade of drought severity, the standard precipitation index was also used due to its advantages in local analysis and establishing time relationship between drought incidences. Determination of desirable local model according to drought condition of Fars Province was conducted after drought severity monitoring and evaluation. The suitable model for each area was provided using time series analysis and based on provided local pattern in different areas of Fars Province. The ARIMA method was selected in drought incidence simulation using time series models on the basis of autocorrelation and partial-autocorrelation. All probable models were tested due to stability and parameters and different method evaluation and the most suitable model was determined and presented for drought simulation in each area.