IMSA- Biodiversity Conservation and Solar Farms: Spatiotemporal Analysis of AI-derived Observational Data

Project Overview: Considering global expansion of solar energy facilities, we question if we can support both power generation and biodiversity conservation. Answer is yes; it is coined ‘ecovoltaics.’ Because birds are highly mobile, occupy a broad range of ecological niches, and serve as sensitive indicators of environmental change, understanding their abundance and behavior is a great start. We recorded 10,000+ hours of daytime videos at 11 solar farms across 9 U.S. states and extracted footage of moving objects using Argonne’s custom AI model. Analysts sort the AI-extracted video clips using their human intelligence and generate highly accurate bird observational data to reliably answer the questions: (1) How consistent or variable are bird abundance and behavior across seasons and regions? (2) Can we observe any patterns between bird abundance/behavior and other facility attributes (e.g., ground cover, surrounding landscape, facility size…)? If you are interested in participating in addressing these questions, this project is for you.  You could be the first to witness interesting bird behavior at solar farms!

Description of Student Project: The student will (1) review and interpret AI-extracted video crips under mentor’s guidance, (2) document/sort/organize video clips birds (and insects), (3) analyze the data and identify patterns of bird abundance and behavior across the seasons and regions, and (4) document the work and summarize findings. The mentor will provide suggested literature, relevant materials/resources, and instructions/training for data review and interpretation. At the end of the project, the student will have gained experience in designing scientific research, identifying relevant information on a research topic, preparing and analyzing data, transforming knowledge into useful/actionable information to set a point of departure for research, in a collaborative setting. The student will also experience an example of how AI could contribute to ecological research.