Suggesting a probabilistic statistical method for identifying late winter frosts (A case study late winter frosts in the West of Iran)

In this research, the attempt is to introduce a probabilistic and statistical Method for identifying late winter frost in the west of Iran, applying statistical methods and Markov chain modeling. To do so, the acquired data about minimum temperature from 37 Synoptic stations and 2 Climatology stations of the 35-year period (1980-2014) were provided by the Meteorological Organization of the country. Then, by performing an interpolation, through Kriging method, for a period of 12784 days, in the time period under consideration, with 6 6 kilometer spatial resolution and a UTM video system, a network database was created. In this survey a late winter frosting event for each area of the surveyed region is a day in which the occurrence probability of a frost is 10% or less. Hence, for each day of the year and each cell (pixel) from the surveyed area, the occurrence probability of frost was estimated. The computational stages of the occurrence probability of frost included using the Markov chain in the MATLAB software and the maps were drawn in the Surfer software. The achievements reveal that in the southwest of the region, the late winter frost date is between 25 February and 5 March and in the north of the region between 21 April and 15 May. The severity of the late winter frosts in the southwestern area of the region is more than the northern part.