Creating Real-Time, Continent-Wide Forecasts Of Bird Migrations

BirdCast would unite big data with ornithology.

Machine Learning
Computer visualizations of tree swallow migration and roosting during fall migration, based on weather radar data, developed by scientists at Cornell, Oregon State and Tulane universities.BirdCast

Via Alexis Madrigal, we've learned of BirdCast: a project from Cornell University's Lab of Ornithology and other institutions that hopes to create sophisticated predictions of bird migrations in real time. BirdCast aims to bring current advances and innovations in computing power and data analysis to birdwatching on the continental scale, by creating computer models that will "reconstruct and predict the behavior of ~400 species of migrating birds across North America;" creating a standardized and open-source methodology for managing bird-related data (observations, weather data, radar readings, satellite imagery, human population, and more); and translating all this information into web-based visuals that will appeal to and inform the general public about conservation, ecology, and computer science.