The escalating global soil crisis, particularly acute in arid and semi-arid regions, underscores the urgent need for innovative, field-adapted methods to assess and enhance soil health. Soil Organic Carbon (SOC) is a pivotal indicator of soil quality and carbon sequestration, yet conventional laboratory techniques are often impractical due to their cost, delay, and environmental footprint. This study proposes a rapid, low-cost, and eco-friendly ‘lab-to-field’ approach using in-situ reflectance spectroscopy (Vis-NIR and MIR) to monitor SOC dynamics directly within almond orchards in Neve-Ya’ar, Israel—systems increasingly vulnerable to climate-induced stress.
Grounded in regenerative agriculture, the research examines the role of cover crops in improving soil structure, enhancing microbial activity, and promoting carbon storage while minimizing chemical inputs. Spatial and temporal SOC variations under different cover crop regimes are quantified using advanced machine learning models. Specifically, Artificial Neural Networks (ANNs) are employed to model complex non-linear relationships between spectral signatures and SOC levels. Their performance is benchmarked against traditional statistical methods such as Partial Least Squares Regression (PLSR) and Principal Component Regression (PCR).
By coupling precision sensing with predictive analytics, this research provides a scalable, real-time SOC assessment framework tailored to dryland farming systems. The outcomes aim to support farmers with actionable insights, enabling adaptive management practices that enhance soil resilience, sequester carbon, and sustain productivity in water-limited agroecosystems. This approach contributes directly to climate-smart agriculture and carbon market participation in dryland orchard systems.
