Integrating Synthetic Aperture Radar Data into a Monitoring Tool for Sugarcane in the Cauca Valley, Colombia

Citation: Mestre-Quereda, A., Lopez-Sanchez, J. M., Martinez-Marin, T., Cazcarra-Bes, V., Mosquera, C., Lozano, J., Quinones, J., Pelgrum, H., Skarli, A., van der Sande, C., Idrovo, A., Hoekman, D., & Vissers, M. (2026, June 17). Integrating Synthetic Aperture Radar Data into a Monitoring Tool for Sugarcane in the Cauca Valley, Colombia. Proceedings of the 16th European Conference on Synthetic Aperture Radar (EUSAR2026). European Conference on Synthetic Aperture Radar (EUSAR), Baden-Baden (Germany). https://doi.org/10.5281/zenodo.20734802

Abstract: A methodology to integrate SAR data with optical imagery to implement a reliable monitoring system for sugarcane under all weather conditions has been developed in the framework of the EU-funded DINOSAR project. An extensive field campaign was conducted to collect key biophysical parameters, e.g., stem count, plant height, and fresh biomass, for developing a model that establishes the relationship between these sugarcane traits and radar observables (backscatter and coherence). Building upon this model and the available models based on optical imagery, an integrated algorithm based on dynamical systems theory has been formulated and tested. This method synergistically combines SAR and optical data streams to generate a robust and accurate information source for sugarcane decision-support tools. Results of vegetation biomass estimation along a whole year demonstrate that such an integration approach outperforms methods based only on regressions (machine learning) or on only one type of input data (SAR or optical).