Soil health forecasting: Exploring pathways toward integration in digital agronomic advisory for sustainable soil management

Soil health is critical for sustainable agriculture and climate resilience, yet monitoring approaches remain static and provide limited guidance on how soil conditions evolve in response to management over time. Soil health forecasting, defined here as the forward estimation of soil functions under alternative management scenarios, may have the potential to complement existing assessment approaches by providing dynamic, management-relevant soil health information. This article examines whether advances in process-based and empirical models, and their integration through hybrid approaches, could support the development of predictive soil health information relevant to agronomic decision-making. Particular attention is given to the prerequisites required for operational use, model calibration and validation, uncertainty assessment, and demonstration of decision relevance. Sub-Saharan Africa is used as an illustrative context, where soil health constraints are substantial and digital agronomic advisory systems are expanding. Rather than presenting soil health forecasting as a readily applicable solution, the article proposes a staged research pathway linking modelling advances with empirical validation and progressive integration into digital agronomic advisory systems. It is expected that strengthening these foundations could help determine the extent to which predictive soil health information could enhance soil management decisions and improve soil stewardship, as well as contribute to more sustainable and climate-resilient agricultural systems in Sub-Saharan Africa. • Hybrid, AI-supported modelling may enable forward-looking soil health prediction. • Soil health forecasting could support contextualized agronomic decision-making. • Soil data and institutional gaps remain key constraints in Sub-Saharan Africa. • Staged research is needed to link models with digital agronomic advisory systems.