Risk Modelling of Late Spring Frost Damage on Fruit Trees, Case Study; Apple Tree, Mashhad Plain

پژوهشهای زراعی ایران

Article 15, Volume 9, Issue 2 - Serial Number 22, July 2011, Pages 273-284   

Document Type: Research Article

DOI: 10.22067/gsc.v9i2.11003

Authors

M Rahimi^* ; M Farajzadeh; Gh Kamali

Abstract

Mashhad plain is one of the most important regions of Apple cultivated areas. Occurring spring late frost creates a lot of damages on bud and decreasing the yield of Apple in this region. Assessment and risk modeling of frost damage would be useful to manage and decrease the damage. The study area is a part of Khorasan Razavi province which is located in Mashhad plain. This region is located in Northeast Iran (36º to 37 º N, 58 º 30' to 60 º E). The area of this region is about 13000 square km which is about one tenth of Khorasan province area. In order to modeling frost damage risk 12 affective parameters including climatological(Minimum temperature, temperature decreasing rate, temperature Increasing rate, Julian days of frost, cumulative degree days, Area under zero line, and frost duration) and geographical parameters (Elevation, Longitude, Latitude, Aspect, and slope) were selected. 3 damage full radiative frosts were selected in the period of Apple flowering time which was dated 20 April 2003, 8 April 2005, and 28 March 2005. Required meteorological data were collected from 9 meteorological standard stations inside and outside of study area. Linear multiple regression were used to modeling the relationship. The map for each parameter was plotted by using suitable interpolation method including IDW; Spline; Kriging. A grid map was defined with 5 by 5 kilometers to extract enough data for entering to the model. The regression equation was significant at the level of 99% significance. By using this equation the predicted amounts of frost risk damage were calculated for each point of grid and also the map was plotted. The regression equation of observed and predicted frost damage risk was provided by correlation of 0.93 and the error map also was prepared. According to this study in frost of 31 Farvardin 1388 South West parts of the plain estimated as the most frost risk areas by %53.19 and the southeast parts were estimated as the least frost risk areas of the plain by %34.5.