نوع مقاله : مقاله پژوهشی
موضوعات
عنوان مقاله English
نویسندگان English
Predicting frost periods is of great importance, as frost damage may reduce the yield potential of many horticultural and agricultural crops in the susceptible areas of the country. Two-state Markov chain models of first, second, and third orders (MC1, MC2, and MC3) were used in the current study for modeling daily frost occurrence and its sequences. The minimum daily temperature data from the Baft synoptic station during the cold months of 1994–2024 were used for the evaluation. To identify the optimum Markov chain model, the Akaike Information Criterion (AIC) was employed. In addition, the estimated conditional probabilities and probability of consecutive non-frost and frost days were given for 2 to 5 day sequences. Findings based on the AIC indicated that the first-order Markov chain model (MC1) was the best to estimate the occurrence of frost. The findings indicated that the consecutive probability average of frost days in 2- to 5-day sequences was, respectively, 30.12%, 22.92%, 17.70%, and 13.78%. Moreover, the chance of consecutive frost episodes of 2-5 days was higher during 3 December to 20 February than it was during the rest of the year.
کلیدواژهها English
