Semantics and machine reasoning enable interoperable, user-friendly biodiversity and ecosystem services modeling

Interoperability between diverse data and models based on the FAIR Data Principles, combined with AI, offers a path forward to make biodiversity and ecosystem services modeling faster and more transparent while improving the reuse of existing knowledge. This is particularly relevant for natural capital accounting under the System of Environmental Economic Accounting, which requires the integration of diverse environmental and economic data alongside models to quantify the benefits ecosystems provide to society. This talk will describe how the Artificial Intelligence for Environment and Sustainability (ARIES) modeling platform uses AI to foster interoperability between a wide range of scientific data and models. ARIES integrates (1) semantics that work across diverse scientific disciplines to logically and parsimoniously describe data and model elements, supporting machine reasoning with (2) open data and models that can be linked together using (3) open-source software tools. A web explorer interface brings data and models to nontechnical stakeholders with visualizations and transparent reporting of provenance (the origin of data, models, and algorithms), while a technical modeler interface offers tools for scientists to contribute data and models to a growing knowledge base, using rules to guide their appropriate reuse. This approach has been applied to natural capital accounting in partnership with the U.N., and can be extended to the Sustainable Development Goals and Post-2020 Global Biodiversity Goals to support faster, easier, and more consistent compilation of such indicators.