OEMC project use case: Tropical deforestation monitoring and characterisation tool

We propose a 10-m deforestation monitoring and characterization tool to support specific tropical country cases. We will use the free and open access data on deforestation and forest disturbances (Hansen et al., 2013) as a reference to enable the characterisation of land uses that follows after deforestation (using Sentinel-1 and Sentinel-2 data) and machine learning. Emphasis will be put on identifying deforestation linked to expansion of commodity crops into forest areas (Linkage to the European Union Deforestation Regulation - EUDR).  Different national stakeholders will be served incl. specific country case studies (e.g. INPE-Brazil), international NGOs (WRI) requiring different types of data and information.  Stakeholder needs:  Free and open access to maps on land use that follows after deforestation to ensure wide usability by diverse stakeholders. Provision of codes via a dedicated web interface or API to enable automated data access and integration into existing workflows. Planned implementation: The proposed 10-meter resolution Tropical Monitoring and Characterization Tool will be implemented at a pantropical scale, focusing on mapping land use following deforestation. Sentinel-1 and Sentinel-2 data, combined with machine learning algorithms, will enable the land use characterization with specific focus on identification of commodity crop expansion into forest areas. Hansen et al. (2013) deforestation and disturbance data will serve as a proxy to forest loss due to its coverage on diverse forest types. The project will prioritize specific tropical country use cases, such as Brazil (in collaboration with INPE) and international stakeholders like WRI, ensuring diverse needs are addressed. The tool aims to support national governments in monitoring and reporting on deforestation mitigation measures, aligned with policies like the European Union Deforestation Regulation (EUDR), and similar regulations in the UK and the US. The implementation will involve collaboration with environmental organizations to enhance conservation planning and resource prioritization. Feedback from stakeholders will guide iterative improvements, ensuring the tool meets operational and scientific needs. Open and reusable data generated by the tool will foster collaboration within the research community, advancing studies on the land use changes and impacts of deforestation. Implementation steps include:  Collection of reference data: Gather and standardize reference data on land use following deforestation across the tropics, emphasizing commodity crop expansion into forest areas. Model adaptation: Adapt and scale the existing machine learning model for Africa to cover the pantropical region, incorporating Sentinel-1, Sentinel-2, and ancillary data such as elevation, temperature, and precipitation. Pantropical mapping: Implement mapping of land use following deforestation across the tropics for the years 2001 to 2023, using a 10-meter spatial resolution and a combination of radar and optical data to enhance characterization accuracy. Dissemination and open access: Publish the results through scientific manuscripts, and make the output maps and codes openly accessible to support stakeholders, including national governments, NGOs, and private companies, in meeting regulatory requirements such as the EUDR.