Model based on temperature parameters predicts optimal harvest date for ‘Whasan’ Asian pear

Abstract

The optimal harvest date for ‘Whasan’ pears (Pyrus pyrifolia ‘Whasan’), which is difficult to predict, is currently determined on the basis of field observations of fruit quality. To create better models for predicting pear maturation, we utilized flowering period data from the Pear Research Institute, South Korea; commercial harvest data; and daily meteorological information from 2007 to 2011 to assess five temperature-related parameters for inclusion in the models: (1) daily mean temperature; (2) daily maximum temperature; (3) daily minimum temperature; (4) average of the daily mean temperature and daily maximum temperature; and (5) average of the daily mean temperature, daily maximum temperature, and daily minimum temperature. We used these parameters and the minimum coefficient of variation method to estimate the base temperature and construct five temperature-based models for predicting harvest date. Data from 2015 to 2016 were used to evaluate the performance of these models, as assessed by average absolute deviation, average squared error, and average absolute percentage error. Using the average of the daily mean and maximum temperatures as the variable input temperature and 9.4 °C as the base temperature, we obtained the most accurate prediction of the optimal pear harvest date. Our model indicates that the optimal harvest date for ‘Whasan’ pear occurs at 2370 accumulated growing-degree days after full bloom.