Agent-based modeling and geotechnologies for land-use planning in Brazilian Cerrado

• Grant number: 24/06700-9
• Support Opportunities: Scholarships in Brazil - Doctorate
• Start date: September 01, 2024
• End date: August 31, 2026
• Field of knowledge: Physical Sciences and Mathematics - Geosciences
• Principal Investigator: Luiz Eduardo Oliveira e Cruz de Aragão
• Grantee: Luiz Gabriel da Silva
• Host Institution: Instituto Nacional de Pesquisas Espaciais (INPE). Ministério da Ciência, Tecnologia e Inovação (Brasil). São José dos Campos , SP, Brazil

Abstract

In Brazil, agropastoral activities are direct and indirect sources of greenhouse gases and their expansion often causes deforestation, however, they are also economically and socially important. The main agricultural frontiers in Brazil are currently found in the Cerrado, where soil and climate conditions and the existence of areas that can be legally deforested facilitate agricultural expansion. There is evidence that the eventual adoption of more restrictive deforestation policies in the Cerrado would not compromise Brazilian expectations for agricultural production, due to the possibility of intensifying production and relocating it to locations with greater suitability. However, there is little evidence of the feasibility of these options, as their adoption depends on individual farmer decisions. Agent-based modeling allows the simulation of phenomena based on individual agent decisions and, for modeling geospatial phenomena it can be coupled with geotechnologies. This thesis aims to quantify the effects of land-use restriction policies on agropastoral activities and on native vegetation in the Brazilian Cerrado, using geotechnologies coupled with agent-based modeling to simulate land-use change. It will use a scenario evaluation approach, with each scenario reflecting a different policy. The generated scenarios will be compared with each other and with a reference one, based on metrics relevant to agropastoral activities and to the landscape. This research aims to contribute to the area of agent-based modeling for geospatial phenomena in large regions, through the design, implementation and validation of the model, and with decision-making for the development of land-use policies currently under discussion.