History
1999-2001
The BirdCast project began in 1999 as a collaboration among EMPACT, EPA Region 3 (Delaware, District of Columbia, Maryland, Pennsylvania, Virginia, and West Virginia), EPA’s Office of Pesticide Programs, the National Audubon Society, Cornell Lab of Ornithology, Clemson University Radar Ornithology Laboratory, the Academy of Natural Sciences in Philadelphia, and GeoMarine, Inc. EPA’s EMPACT program funded the original BirdCast project, with public operation beginning on April 1, 2000 and ending May 31, 2001. As one of the primary objectives of the project (below), predicting and observing bird migration based on forecast models using weather data to predict bird density aloft and then observe migration on five (2000) or 10 (2001) radars, respectively, were novel and far advanced for their time.
The Four Original Objectives of BirdCast
1
To maintain a website that shared information about bird migration and the steps that people could take to mitigate the danger and stress that migrating birds face when passing through an area.
2
To predict and monitor bird migrations on a daily basis using weather radar, continually interpreted by trained scientists and presented using text summaries, charts, and radar maps.
3
To collect and disseminate volunteers’ reports of bird sightings, and allow website visitors to view this database and display reports in chart or graph form.
4
To raise public awareness about the sensitivity of migratory bird populations through press releases, land manager outreach, promotional materials, conferences and conventions.
Despite rather dramatic success in achieving the primary objectives of the project (below), the transfer of this technology from a regional to continental approach did not garner the funding necessary to continue expanding research and applications. To read more about BirdCast’s origin story, you can download this document.
2010-2018
In 2010, after a decade of technological evolution and innovation that saw the advance of the Internet, major developments in machine learning and artificial intelligence, breakthroughs in pursuing and managing big data research, and the advent of cloud computing, ornithologists and computer scientists from Cornell University and Oregon State University proposed a respective and novel approach that would restart the BirdCast project to the National Science Foundation’s (NSF) “Cyber-Enabled Discovery and Innovation (CDI)” Initiative. This was the government’s “bold five-year initiative to create revolutionary science and engineering research outcomes made possible by innovations and advances in computational thinking.” This proposal resulted in a jointly awarded CDI grant from NSF in 2011. The NSF grant covered the development of BirdCast from 2011-2016 and evolved it from its earlier incarnation to the project it would become. This second stage of development also included collaboration with the University of Massachusetts Amherst along with contributions from individual scientists and experts at Microsoft, NASA, and Amazon among others.
When determining how to create the maps that make BirdCast what it is today, researchers at Oregon State University and University of Massachusetts Amherst proposed two innovative machine-learning techniques: Collective Graphical Models (CGMs) and Semi-Parametric Latent Process Models (SLPMs). These techniques would allow researchers to identify the complex conditions governing the dynamics of migration behavior, such as choices of migratory pathways, factors influencing migration initiation, and the speed and duration of nocturnal movements. The BirdCast team also planned to develop new interoperable data infrastructure for synthesizing bird observations, flight calls, radar data, and covariate data from multiple sources including satellite imagery, weather, and human population data. Research furthered plans for understanding relationships between radar and eBird data and inspired new NSF awards (for example, see Dark Ecology and BirdVox).
2018 – Present
The NSF funding ended in 2016, leaving the project in an excellent position to expand with continued support from Leon Levy Foundation, Amazon Web Services, Lyda Hill Philanthropies, and Amon G. Carter Foundation, among others. Novel web-based data visualizations for communicating the migration predictions generated by BirdCast to the general public had been a hallmark of the project since 2012; but beginning in spring 2018, the project began providing forecast and live bird migration maps that elegantly represented the original vision and spirit of the project’s 1999 intent and the grandeur of bird movements, but this time at the continental scale. After major research and development efforts in machine learning, cloud-based computing, and big data analytics, these visualizations and products represented the culmination of a 20-year long vision; so, too, they represented the beginnings of new inspiration for the next generation of bird migration research, outreach and education, and application. A cadre of dedicated aeroecology postdoctoral research associates at the Cornell Lab of Ornithology, supported by the Edward W. Rose Postdoctoral Fellowship, collaborated with researchers at UMASS, Oxford University, and by 2019 Colorado State University (CSU) Aeroeco Lab. More recently, these collaborations have included Purdue University and University of Illinois Urbana-Champaign.
For more information about radar ornithology and how to understand the application of radar technology for studying bird migration, see this primer.
Scientific Team
BirdCast is made possible by the participating scientists at the below institutions, and many other contributors.
