Monitoring of Overwintering Leaf Age by Integrating Phenological, Temporal, and Thermal Data: An Indicator for Assessing Winter Wheat Seedling Condition

The seedling condition of winter wheat before overwintering affects its safe overwintering and final yield, and leaf age is an important indicator for evaluating the seedling condition. However, traditional leaf age surveys primarily rely on manual observation, which is not only time-consuming and labor-intensive but also makes large-scale and simultaneous observations difficult. Existing methods based on image segmentation are constrained by issues such as unclear leaf boundaries, leaf overlapping, and high computational complexity, limiting their large-scale application.

Against this backdrop, a research team led by Dr. Zhenhai Li from Shandong University of Science and Technology has proposed a method for monitoring winter wheat leaf age that integrates remote sensing phenological information, temporal variations in vegetation indices, and GDD (Fig. 1). Their study, made available online on May 28, 2026 in The Crop Journal, integrates remote sensing and meteorological data, providing new technical support for the monitoring and precise management of winter wheat seedling condition on a regional scale and laying the foundation for cross-regional application and long-term monitoring.

Based on MODIS remote sensing data and ERA5-Land reanalysis temperature data, the research team first extracted the emergence date of winter wheat and then used the emergence date, GDD, and NDVI change rate (β) to construct a physiologically-based GDD leaf age model (LA_GDD) and a random forest-based leaf age model (LA_RF) to estimate the leaf age of winter wheat before overwintering.

To evaluate model performance, the researchers used field-measured leaf age samples for validation and found that the LA_RF model, which integrates multi-source data, significantly improved the accuracy of leaf age monitoring. It achieved an R² of 0.67 and an RMSE of 0.56 leaves, which was significantly better than the LA_GDD model that relied solely on GDD. “The model demonstrated good adaptability and stability in major winter wheat planting areas of Shandong Province, providing a new technical solution for regional-scale winter wheat seedling monitoring,” says corresponding author Dr. Zhenhai Li.