Shelf-life forecasting model for long-day yellow onions using pre-harvest and postharvest parameters

Overview

This solution is a data-driven forecasting model designed to predict the storage shelf life of long-day yellow onions. It integrates pre-harvest environmental factors (heat, soil moisture), plant-specific parameters (growing degree hours, leaf height, bulb diameter, number of green leaves, leaf bending), and postharvest management variables (curing time and temperature, skin color, aroma, storage temperature, and relative humidity) into a single predictive framework. The approach builds on a previously validated method used for wild blueberries under modified atmosphere packaging, where monitoring of external and fruit quality parameters enabled the extension of the marketing period by one month. For onion producers and packers, this means more reliable harvest planning, reduced postharvest losses, and the ability to extend the marketable window of stored bulbs.

Technical specifications

Key features:

• Integration of pre-harvest parameters including growing degree hours (GDH), foliage height, bulb diameter, green leaf count, and leaf bending measured at harvest
• Inclusion of environmental covariates such as heat units and soil moisture across two irrigation regimes and two GDH periods
• Postharvest curing modeled across a range of time-temperature combinations
• Storage condition modeling at 0–4 °C with relative humidity of 60–70%
• Use of a custom fractional factorial design (CFFD) combined with linear modeling and a predictive profiler to identify optimal storage settings
• Bulbs monitored over 6 months with measurements every 15 days, tracking weight loss, disease incidence, outer coat color, and aroma

How it works:

A custom fractional factorial experimental design captures the combined effects of growing conditions, curing treatments, and storage parameters. A linear model is fitted to the response variables (weight loss, disease presence, coat color, aroma), and a predictive profiler is used to identify the combination of factors that maximizes overall bulb quality and storage longevity. The approach allows prediction of quality outcomes even under non-optimal conditions.

Technology readiness level

The methodology is at an early-to-mid stage of development. The underlying modeling approach has been validated in a prior application on wild blueberries using modified atmosphere packaging, where a custom fractional factorial design with linear modeling and predictive profiling successfully extended the marketable period. For onions, the next phase of validation is planned: a CFFD combined with a linear model and predictive profiler will be applied to bulbs grown under two irrigation regimes and two GDH periods, with curing and storage treatments systematically varied. Validation requires participation of a local grower for one season to supply and store the bulbs. The technology is not yet commercially deployed and requires field validation before adoption.