Statistical Modelling Relevance and Implications for Agricultural Extension

Agricultural extension systems in South Africa face increasing pressure to provide timely, location-specific, and risk-aware advisories to dryland grain farmers under intensifying climate variability. Statistical climate modelling has emerged as a critical bridge between climate science and farm-level decision-making, yet its application within extension practice remains uneven. This review synthesises peer‑reviewed literature published between 2010 and 2026 on climate data curation, statistical and spatial modelling, and crop–climate integration relevant to agricultural extension in South Africa and comparable semi‑arid regions. Trends in climate variability affecting dryland grain systems are examined alongside statistical methods used to link climate drivers with yield and risk outcomes. Approaches for embedding modelling outputs into extension advisory systems are assessed. Results indicate a shift from descriptive analyses towards integrated, multi‑model decision‑support frameworks combining downscaled climate data, regression and time‑series techniques, and process‑based crop models. Persistent gaps remain in local calibration, uncertainty communication, and extension capacity. A consolidated, extension‑oriented modelling framework is proposed, with priorities for research, policy, and capacity development identified to strengthen climate-smart agricultural extension.