Assessing the reliability of four air temperature interpolation functions in estimating chill accumulation for deciduous fruit trees

Chill models, which are essential for estimating the chill accumulation necessary for the dormancy of various fruit trees, require continuous hourly temperature. To date, there has been no comprehensive study establishing the best method for filling gaps in hourly temperature records for Greek weather stations. Firstly, this study compares the performance of four temperature interpolation functions (the akima, cosine, hourly, and linear) against hourly air temperature observations from the Aristotle University of Thessaloniki weather station, for the dormancy period (October to March), during 1951-2023. Next, it tests these methods as input to two chill accumulation models (the Dynamic and the North Carolina Model. The hourly model was the most robust interpolation method during the day, even though at midday it was outperformed by the others. The hourly and linear interpolation methods exhibited the lower daily MAE values during the dormancy period. The hourly model was also the most accurate method for estimating chill accumulation with both chill models. the Dynamic and North Carolina model.