Yield forecasting in apple based on meteorological parameters

Three most important variables namely chill units(Cu), number of hours with temperature 14–20°C (Nh^14–20) and number of hours with RH ⁴°⁻⁶⁰ (Rh^40–60) during bloom were chosen on the basis of simple correlation with yield for regression analysis. Principal component(PC) regression gave better precision for the estimates than Ordinary Least Square(OLS) regression analysis. It revealed that the variable included in regression have explained about 97% of the total variation in yield and a unit change in the varibles Nh^14–20 and Rh^40–60 brought significant change in yield. Further evaluation of the selected regression line showed its execellent forecasting performance. These findings provide a strong base for future research on yield forecasting.