NSF Award
2223511
Plant phenology – the timing of plant life-cycle events, such as leaf growth, flowering, and fruiting – plays a fundamental role in shaping terrestrial ecosystems. The timing of plant phenology not only affects the fitness of individual plants, it also impacts the fitness and behaviors of organisms dependent on plants, which in terrestrial ecosystems includes nearly all animals, either directly or indirectly. Thus, changes in plant phenology can trigger dramatic, and sometimes devastating, consequences for ecosystems and human economic interests and health. Plant phenological data are therefore indispensable for understanding ecosystem function, detecting ecosystem changes, and predicting the impacts of ongoing climate and land use changes. Given the importance of plant phenology, continuing local, regional and national data collection efforts have generated large volumes of phenological data. However, these data are surprisingly heterogeneous, difficult to integrate, and thus remain largely inaccessible for broader research. At the same time, community science and specimen digitization infrastructure have produced massive, rapidly expanding collections of herbarium specimens and in situ plant photographs, which contain a wealth of virtually untapped historical and contemporary phenological information. This project will use machine learning approaches to extract phenological data from plant photographs and digitized specimens. These data will then be integrated with phenological monitoring resources to create an open access, global plant phenology database – Phenobase. During this project, one postdoctoral researcher and several graduate and undergraduate students will be trained in programming and data science skills. <br/><br/>The goal of this project is to support community needs for generating and delivering high-precision, harmonized and semantically integrated plant phenological data at unprecedented taxonomic, geographic, and temporal scales, along with new tools to help scientists and the public engage with these data. To achieve this goal, this project will develop a global, standardized knowledge base by integrating different phenology observation networks around the world; expand this knowledge base by using computer vision (CV) techniques to generate new, high-quality phenological data from the rapidly growing collection of community-submitted plant photographs on iNaturalist and Budburst; add critical historical data by using similar CV techniques on herbarium specimens available through iDigBio and GBIF; develop tools for data query, access, and visualization delivered via the Web and as software packages; and foster compelling, community-driven use cases showcasing the use of Phenobase for new research and for public good. These approaches will not only meet current growth in imaging, but scale to meet continuing, exponential growth into the future. By weaving together phenologically relevant outputs from monitoring projects from around the globe, including the efforts of millions of community scientists, Phenobase will support and empower phenological research that is currently impossible. Results derived from this project can be found at http://plantphenology.org/.<br/><br/>This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.
