Deforestation and environmental imbalance in the Amazon
DEVELOPMENT OF GEOSPATIAL SKILLS FOR PUBLIC CONTROL
Instruction and mentoring to develop geospatial skills for the team of employees at the Federal Audit Court (TCU).
The control and inspection of environmental crimes in the Brazilian Amazon depends in part on the availability of geotechnologies and the technical capacity to incorporate them into work practices. Using freely available tools and data, this project aimed to strengthen the geospatial skills of the team of employees at the Federal Audit Court (TCU) for public control.
METHODOLOGY
The large-scale analysis of CAR (Rural Environmental Registry) data is conducted using advanced open-source geoprocessing and modeling tools. Technologies such as PostgreSQL/PostGIS and Dinamica EGO integrate land use data, deforestation, enforcement actions, and other geospatial datasets to generate accurate insights.
In this project, to address the inherent challenges of big data, we adapted Dinamica EGO for parallel processing and high-performance computing using AWS infrastructure. For land use mapping, we developed a set of models using an integrated version of Sentinel and Planet imagery, forming a cloud-free mosaic with land use classification, employing machine learning algorithms.
RESULTS
More than ten TCU employees participated in the course and, among the feedback received, the relevance of the customized content to meet the organization's demands and the link between theoretical and practical content stands out. In total, 16 classes were taught, totaling 34 hours of synchronous moments.