Innovative Approaches to Bengal gram Yield Mapping: Integration of Sentinel-1 SAR and Crop Simulation Models for Precision Agriculture

Abstract

Accurate spatial yield estimation is crucial for optimizing agricultural management and ensuring food security. This study integrates Sentinel-1A SAR remote sensing data and the DSSAT crop simulation model to predict Bengal gram (chickpea) yield in Nagaur district, Rajasthan, India. Sentinel-1A backscatter data were processed for crop area mapping, achieving an overall classification accuracy of 85.1% and a kappa index of 0.70, demonstrating the reliability of SAR for agricultural monitoring under diverse weather conditions. Leaf Area Index (LAI) was derived from SAR backscatter values and linked to DSSAT-simulated yields, generating spatial yield predictions. Validation using Crop Cutting Experiment (CCE) data showed a high agreement of 91.3% between predicted and observed yields, with low root mean square error (RMSE), confirming model accuracy. This research highlights the synergistic potential of SAR-based remote sensing and simulation models for large-scale yield forecasting, advancing precision agriculture. Future efforts may incorporate additional sensors and machine learning to further enhance prediction accuracy and adaptability to climate variability.