Abstract
Future agricultural landscapes will likely be shaped by the interplay between socioeconomic developments and natural conditions. However, existing theory-driven, process-based models often rely on idealized assumptions, limiting their capacity to capture real-world complexities fully. To complement these methods through an observational, data-driven approach, we developed a novel global dataset utilizing a statistical fixed-effects model. This paper presents a novel global dataset detailing projections of harvested area allocation for ten major crop groups across 197 countries and regions from 2020 to 2100. The dataset was generated using a statistical fixed-effects model calibrated on historical data. It includes annual projections under six distinct SSP-RCP scenarios (SSP1-2.6, SSP2-4.5, SSP3-7.0, SSP4-3.4, SSP4-6.0, and SSP5-8.5). For each scenario, the dataset provides future trajectories for key national agricultural management inputs-including nitrogen application rates, irrigation extents, and mechanization levels-and the resulting projected cropping shares. This dataset is designed to support assessments of food security, trade policy, and environmental impacts by providing a consistent, data-driven set of future agricultural landscape patterns.