The objective of this study to investigating and predicting the hazardous effects of monthly Ferrin temperature on horticultural and agricultural products at the North bar of Iran. For this purpose, the first stage data were obtained for the whole station temperature over a period of 30 years. Then, using Anfis adaptive neural network model, the data were analyzed for prediction and prediction for the next 6 years. Then, to measure the land suitability of Iran's northern strip for cultivating based on the predicted data, two models of Vikor and Topsis were used. Both the Topsis and Vikor multivariate decision making models combined the minimum temperature of the stations well, but did not reflect well at the maximum temperature in the worst-priority stations. According to the findings of the study, with respect to the friction frain modeling, the maximum temperature showed the lowest defect compared to the minimum temperature. In Golestan province, the maximum temperature peaks and at least both are in weak increment, but in Gilan province, the maximum temperature peaks and at least both the maximum and maximum temperatures are higher. Mazandaran province showed maximum temperature and minimum temperature in both incremental and minimum temperature conditions.