[M-AG33] Satellite Land Physical Processes Monitoring at Medium/High/Very High Resolution

[MAG33-P03] Application of Semi-Physical Models for Corn and Soybean Yield Estimation in Illinois, USA

*Vishal Gautam¹, Abdul Gani², Shray Pathak ¹ (1.Indian Institute of Technology Ropar, Punjab, India , 2.Netaji Subhas University of Technology, New Delhi, India )

Keywords:Agricultural Productivity, Decision Making, Climate-Resilient Agriculture, Net Primary Production

Enhancing the accuracy of crop yield predictions is crucial for optimizing agricultural productivity, resource management, and strategic decision-making. The present study employs a semi-physical approach to predict corn and soybean yields in Illinois, USA, by integrating remote sensing data with biophysical modeling techniques. For assessing the plant growth, 25 years of historical data was considered from 2000 to 2024, of photosynthetically active radiation (PAR) and fraction of photosynthetically active radiation (fAPAR). Light use efficiency (LUE) was determined by using water stress, temperature stress, maximum light use efficiency and vegetative growth factor. By integrating these factors, net primary production (NPP) was calculated which serves as a critical parameter for the estimation of crop yield. A regression-based model is employed to estimate final yields of corn and soybean. Model performance was evaluated by using the coefficient of regression (R²) and root mean square error (RMSE). R² values of corn and soybean was determined as 0.59 and 0.42, which indicates a moderate correlation between the predicted and observed yields. RMSE values of 1.4377 and 0.4888 were found for corn and soybean which demonstrates reasonable accuracy in yield prediction. The findings highlight the effectiveness of combining remote sensing derived inputs and environmental stress factors in crop yield estimation. Thus, semi-physical modeling approach bridges the gap between empirical and mechanistic models, offering a robust framework for agricultural monitoring. The results indicate that integrating satellite data and stress factors enhances the accuracy and reliability of yield estimations, making this approach valuable for policymakers, agronomists, and researchers aiming to optimize agricultural production in varying climatic conditions. By advancing precision agriculture, this study contributes to the development of sustainable food production systems, ensuring more resilient and data-driven agricultural planning.