Mathematical model of postharvest variation in tomato color based on optimized response surface methodology

Manual inspection and instrumentation form the traditional approach to determining tomato color but these methods only determine tomato color at a given moment and cannot predict dynamically how tomato color varies during storage and transportation. Such methods thus cannot help suppliers and retailers establish good management practices for the flexible control of tomato maturity, accurate judgment of market positioning in the industry, or during distribution and marketing. To address this shortcoming, this work first investigates how tomato color parameters (a* and h°) evolve through the various stages of maturity (green, turn, and light red) under different storage conditions. Based on experimental results, it develops an optimized response-surface model (RSM) by using differential evolution to predict how tomato color varies during storage.Tomatoes are more likely to change color at high temperatures and under conditions of high humidity. Temperature affects tomato color more strongly than humidity. The accuracy of the RSM was confirmed by a good agreement with experiments. All determination coefficients R2 of the RSMs for a* and h° are greater than 0.91. The mean absolute errors for a* and h° are 3.8112 and 5.6500, respectively. The root mean square errors for a* and h° are 4.6840 and 6.9198, respectively.This research reveals how storage temperature and humidity affect the postharvest variations in tomato color and thus establishes a dynamic model for predicting tomato color. The proposed RSM provides a reliable theoretical foundation for dynamic, nondestructive monitoring of tomato ripeness in the cold chain. © 2021 Society of Chemical Industry.