In the quest to optimize agricultural productivity, researchers have made a significant stride in accurately estimating winter wheat yields under diverse management practices. A recent study published in *Agronomy* introduces a novel framework that integrates multi-source remote sensing data, offering a promising solution to the challenges faced by current yield estimation models.
The study, led by Hao Kong from the School of Surveying and Land Information Engineering at Henan Polytechnic University, addresses two critical issues: the limited adaptability of existing models to different management practices and the ineffective integration of multi-source remote sensing features. By combining UAV-based multispectral and thermal infrared remote sensing data, the researchers proposed a yield estimation framework based on multi-source feature fusion.
Three machine learning algorithms—Partial Least Squares Regression (PLSR), Random Forest (RF), and Extreme Gradient Boosting (XGBoost)—were employed to retrieve key biochemical parameters of winter wheat. The Random Forest model demonstrated superior performance, with impressive retrieval accuracies for chlorophyll, nitrogen, and phosphorus contents. “The RF model’s ability to handle complex data and provide accurate predictions makes it a valuable tool for precision agriculture,” said Kong.
Subsequently, yield estimation models were constructed by integrating the retrieved biochemical parameters with phenotypic traits such as plant height and biomass. The RF model again exhibited superior performance, with an R2 value of 0.66 and an RMSE of 867.28 kg/ha. SHapley Additive exPlanations (SHAP) analysis identified May chlorophyll content (Chl-5) and March chlorophyll content (Chl-3) as the most critical variables for yield prediction. “Understanding the dynamics of chlorophyll during key physiological stages is crucial for enhancing yield estimation accuracy,” explained Kong.
The study also quantified the impact of different management practices on yield. The straw return + 50% inorganic fertilizer + 50% organic fertilizer (RIO50) treatment under the combined organic–inorganic fertilization regime achieved the highest measured grain yield of 11,469 kg/ha. This finding provides valuable insights for farmers looking to optimize their fertilization strategies.
The commercial implications of this research are substantial. Accurate yield estimation under varied management practices enables farmers to make informed decisions, optimize resource allocation, and ultimately enhance productivity. “This study offers a technical basis for precise field management, which is essential for the development of smart agriculture,” said Kong.
The integration of multi-source remote sensing data and machine learning algorithms represents a significant advancement in the field of agritech. As the agriculture sector continues to embrace digital transformation, such innovations will play a pivotal role in shaping the future of farming. The study published in *Agronomy* by Hao Kong and his team from Henan Polytechnic University provides a robust framework for improving yield estimation accuracy and optimizing management practices, paving the way for more sustainable and productive agricultural systems.