Spatial dynamics of native vegetation, pasture, and grain cultivation in the Amazon, Cerrado and Pantanal biomes: Simulation of land use and land cover from 2024 to 2030

Project overview This project aimed to model scenarios of agricultural intensification and remnant conservation by simulating future deforestation in the Amazon, Cerrado, and Pantanal Brazilian biomes. It also assessed avoided deforestation to better understand the dynamics underlying land-use transitions. The project is a joint initiative between the Image Processing and Geoprocessing Laboratory (LAPIG–UFG) and The Nature Conservancy (TNC).Dataset description This dataset contains the outputs of spatial land-use and land-cover (LULC) modeling performed using the Dinamica EGO software (https://csr.ufmg.br/dinamica). Three future scenarios were simulated, representing different levels of native vegetation protection and land-use intensification. The dataset includes four LULC classes: Native Vegetation, Pasture, Mosaic, and Soy. All layers are provided at a 30-meter spatial resolution and are projected in ESRI:102033 (South America Albers Equal Area Conic). Scenarios The scenarios represent contrasting land-use governance and intensification pathways, including: Business-as-usual (BAU): existing legal framework and strengthens the Brazilian Forest Code; Collaborative Governance (CG): Increased protection of native vegetation and strictly blocks pasture expansion over native vegetation. Integrated Governance (IG): Intensified use of pasturelands with constraints, while strictly blocking pasture expansion over native vegetation. Temporal coverage Baseline year: 2023Projection horizon: 2030 Methodological notes Spatial simulations were conducted using Dinamica EGO, integrating land-use transitions derived from MapBiomas data (https://brasil.mapbiomas.org/en/), spatial predictors, and scenario-specific constraints. The resulting outputs represent potential future land-use patterns under different environmental and legall assumptions. The complete modeling framework, including scripts and parameters used to generate the scenarios, is available at:https://github.com/lapig-ufg/Scenarios-project Intended use and limitations This dataset is intended for scenario analysis, spatial planning, policy assessment, and research on land-use change and conservation strategies. Results should be interpreted as scenario-based projections and not as precise forecasts of future land-use outcomes. Acknowledgements This research was supported by LaMCAD/UFG.