Data-driven agriculture : big data applications across large-scale farming systems and lessons for Malta

Abstract

The digitalisation of agricultural systems represents one of the most consequential technological transitions in the history of food production. This study critically evaluates Big Data applications across thirteen agricultural domains and assesses their transferability to Malta’s small-scale, water-stressed Mediterranean farming system, a country classified as the most water-stressed in Europe. Employing a four-part methodology, a systematic literature review of sources published between 2010 and 2025, a structured comparative case study analysis of six jurisdictions (United States, Brazil, Netherlands, Australia, Israel, and Cyprus), the development of an original Integrated Big Data Agricultural Value-Chain Framework (IBDAVC), and the Maltese Digital Agriculture Readiness and Priority Model (MDARP), the study documents substantial productivity gains from Big Data adoption internationally: yield prediction accuracy within 5–10% of harvest values, irrigation water savings of 20–40%, fertiliser reductions of 12–18%, and disease detection accuracy exceeding 90%. Applied to Malta, this evidence reveals an acute and worsening crisis: annual groundwater abstraction already exceeds the sustainable recharge rate by approximately 11 million m³, nitrate concentrations in agricultural catchment runoff reach up to 471 mg/L, nearly ten times the EU Groundwater Directive limit, across the Wied il-Għasel watershed, and system dynamics modelling projects catastrophic groundwater decline under all climate scenarios without corrective action. Applied through the MDARP, the interventions are ranked by contextual fit: a Precision Agriculture Service Unit ranks highest (3.93/5), followed by shared soil and salinity mapping (3.70/5), with smart irrigation scheduling scoring 3.43/5 on the strength of exceptional environmental need and policy alignment. The study concludes that while Malta’s structural constraints, 69.7% of holdings under one hectare, only 55% of farmers receiving CAP subsidies, farmland costing 24 times the EU average, are genuine, appropriately designed shared-service models and targeted public investment can enable meaningful digital agricultural transformation, provided enabling layers of data infrastructure, analytics capability, governance architecture, and human capital are developed in parallel.