Digital twins to quantify the impact of growing variability on the harvest quality of orange

The quality of citrus fruit is influenced by various growing conditions, including weather. However, the impact of weather differences between growing regions on citrus quality at harvest is not well understood. This study employs a mechanistic-driven digital replica of the growth process of a Valencia orange from fruit set until harvest to quantify this impact. Measured daily temperature, humidity, rainfall, solar radiation, and wind speed data from various orange growing regions of South Africa, including Citrusdal, Nelspruit, Letsitele, and Sundays River Valley (SRV), inform and drive this digital model. The results suggest that the differences in weather conditions between growing regions significantly affect fruit diameter (FD), fruit weight (FW), rind thickness (RT), rind weight (RW), total soluble solids (TSS), titratable acidity (TA), and °Brix/acid ratio of oranges at harvest. The differences between growing regions led to variations of up to threefold for FD, twofold for FW, RT, RW, TSS, °Brix/acid ratio and up to fourfold for TA upon harvest. Particularly, oranges produced from warmer Letsitele and Nelspruit regions are found to be larger, sweeter and less acidic compared to those from coastal SRV upon harvest. The study also reveals the impact of the fruit growth process on the temperature gradient within the fruit, which varies across growing regions. The maximal temperature difference between the fruit core and surface during the growth process ranges from 4 °C to ∼7 °C. These variations in fruit temperature gradient could lead to variations in temperature-driven quality decay of fruit from different climatic regions at the start of their postharvest journey. These findings provide valuable insights for the citrus industry, aiding to optimize practices, harvest planning, and postharvest logistics. The output of this digital twin will help identify areas needing extra precooling to extend postharvest shelf life and minimize quality decay. Real-world use allows growers to schedule harvests based on regional weather conditions.