A fuzzy tree-based framework for vegetation state monitoring

The climate change emergency strongly affects vegetation growth in terrestrial ecosystems: large scale vegetation-climate interactions reveal an increased frequency of extreme weather and climate events, with significant impacts on ecosystems at different spatiotemporal scales. Vegetation monitoring is a critical element to assess the changes and treats to the environment also aimed at sustainable conservation of wildlife. A framework is proposed to aggregate vegetation indices described by fuzzy sets to assess vegetation health. Several fuzzy rules have been defined grouped by feature estimation (cover, vigor, water stress, etc.) and then triggered according to a decision tree schema to obtain a robust interpretation of vegetation status. The control and flow of the activation rules is driven and optimized by an agent-based modeling. Case studies highlight the applicability of the proposed framework.