Comprehensive Survey on Ensemble Machine Learning for Soil Health Monitoring

Recent years have seen rapid progress in soil monitoring technologies as the demand for real-time, scalable, and data-driven assessment of soil health continues to rise. Advances across sensing systems, remote observation tools, digital soil infrastructures, and machine learning have reshaped the way soil physical, chemical, and biological properties are measured and interpreted. Studies have explored the limitations of conventional point-based measurements, highlighted the growing influence of environmental pressures on soil degradation, and emphasized the need for harmonized monitoring frameworks at regional and national scales. Emerging approaches—including IoT-enabled soil sensing, Earth observation for large-scale mapping, microbiome-based indicators, and machine learning–driven predictive modeling—offer new opportunities but also introduce technical, logistical, and data-related challenges. Given the rapid evolution and fragmentation of these developments, a comprehensive survey is necessary to consolidate current knowledge, compare methodological trends, and outline research directions that support reliable, continuous, and integrated soil monitoring systems for sustainable land management.