Determination, Evaluation, and Comparison of Empirical Models for Predicting Radiation Frost Including the Long-Wave Radiation model (Case Study: Isfahan City)

Frost and chilling threaten agricultural production in various parts of the world which annually leads to significant economic losses to agricultural products. Even if all aspects of crop production are well managed, the occurrence of frost cooling can result in the complete loss of the entire agricultural production. According to the Agricultural Organization report of Isfahan Province, the extent of frost damage in this province during the years 2007 to 2017 was over 262 thousand hectares of land and 96 million tons of crops. Considering the annual frost damage, accurate predictions of frost radiation can provide the enough time for orchardists to prepare and utilize methods for combating frost and chilling protection. The aim of this research was to predict the minimum temperature using an empirical model based on long-wave radiation at nighttime events of frost radiation cooling in order to determine the start time of active protection methods. Meteorological data from the synoptic weather station in Isfahan over a 10-year statistical period were extracted and used for calculations and model determinations. In this regard, the Young model, correction factor, and multiple regression model were identified, examined, and compared. The accuracy of these models was evaluated using the goodness-of-fit tests. The results indicated that the FAO model and the multiple regression model developed in this study provide better predictions compared to other models. Furthermore, by using the wet bulb temperature at 18 Greenwich Mean Time at nights with characteristics of frost radiation cooling with the multiple-variable regression model, the minimum temperature for the next day can be predicted.