Supplementary Data — Multisource Earth Observation–Driven Hybrid Machine Learning Framework for Agricultural Sustainability Monitoring: Evidence from Northern Ghana

This dataset contains supplementary tabular outputs supporting the study:“Multisource Earth Observation–Driven Hybrid Machine Learning Framework for Agricultural Sustainability Monitoring: Evidence from Northern Ghana.” Files include:- Independent test samples with reference labels and model predictions- Area-by-class summary statistics Derived from Sentinel-1, Sentinel-2 SR Harmonized, and Dynamic World data processed in Google Earth Engine. Supplementary Data and Reproducibility MaterialsMultisource Earth Observation–Driven Hybrid Machine Learning Framework for Agricultural Sustainability Monitoring: Evidence from Northern Ghana This record provides the supplementary dataset and reproducibility documentation supporting the accompanying manuscript. The study develops a calibrated probabilistic sustainability-risk modeling framework under rare-event conditions (8.65% prevalence) using multisource Earth observation data.