Mapping the Web of Life | Google’s AI for Biodiversity
Researchers at Google are using AI to build a map of plant, animal, fungi and other species across the world. These advanced Species Distribution Models (SDMs) help scientists, policymakers, and conservationists identify critical habitats, supporting protection efforts where they matter most.
While SDMs are not new, our approach is: Google’s SDMs are powered by Graph Neural Networks, a complex ML architecture, which can capture interactions among species, and also bring together data from diverse modalities including satellite, species traits, and species observations. This vast web of information leads to increased accuracy in identifying where species live and thrive, including down to fine spatial scales, which are often the most relevant for decision making.
Hear directly from Neil D. Burgess, Chief Scientist of the United Nations Environment Programme World Conservation Monitoring Centre (UNEP-WCMC), Jenna Wraith, Head of Sustainable Futures & Principal Data Scientist at the Queensland Cyber Infrastructure Foundation (QCIF) and EcoCommons, and Drew Purves, Research Scientist at Google, about the value that machine learning brings to species distribution modeling.
To protect life on Earth, we first need to know where to find it.
Additional Resources:
Learn more about Google’s biodiversity research: https://deepmind.google/blog/mapping-...
Read about the work of the UNEP-WCMC at unep-wcmc.org
Learn about Dr. Jenna Wraith’s work at QCIF and EcoCommons at qcif.edu.au and ecocommons.org.au/
Dive deeper into graph neural networks at https://distill.pub/2021/gnn-intro/
Read our paper on the Graph Neural Net Model https://arxiv.org/abs/2503.11900
Subscribe to Google Research to watch other videos about our AI work: / googleresearch
#GoogleResearch #AIForScience #Biodiversity #Conservation #MachineLearning #Sustainability #EarthEngine #GraphNeuralNetworks #NatureTech #SpeciesDistributionModels #SDM #GNN