Prediction of soil frost penetration depth in northwest of Iran using air freezing indices

Abstract

Information about soil frost penetration depth can be effective in finding appropriate solutions to reduce the agricultural crop damage, transportations, and building facilities. Amongst proper methods to achieve this information are the statistical and empirical models capable of estimating soil frost penetration depth. Therefore, the main objective of this research is to calculate soil frost penetration depth in northwest of Iran during the year 2007–2008 to validate two different models accuracy. To do so, the relationship between air and soil temperature in different depths (5-10-20-30-50-100 cm) at three times of the day (3, 9, and 15 GMT) for 14 weather stations over 7 provinces was analyzed using linear regression. Then, two different air freezing indices (AFIs) including Norwegian and Finn AFI was implemented. Finally, the frost penetration depth was calculated by McKeown method and the accuracy of models determined by actual soil frost penetration depth. The results demonstrated that there is a significant correlation between air and soil depth temperature in all studied stations up to the 30 cm under the surface. Also, according to the results, Norwegian index can be effectively used for determination of soil frost depth penetration and the correlation coefficient between actual and estimated soil frost penetration depth is r = 0.92 while the Finn index overestimates the frost depth in all stations with correlation coefficient r = 0.70.