The National Geographic Society and Microsoft’s AI for Earth program are partnering to support novel projects that create and deploy AI tools to improve the way we monitor, model, understand, and ultimately manage Earth’s natural resources for a more sustainable future.
The grants given by the partnership will support projects that use cloud computing to create and deploy open-source models and algorithms that make key analytical processes more efficient in the field. This partnership is focused on supporting projects that will build tools such as applications, application programming interfaces (APIs), or packages to be shared. Microsoft will help the successful applicants make their models and tools available for use by other environmental researchers and innovators.
Extreme weather events, wildlife trafficking, rising sea levels, increased agricultural and urban development demands, higher global temperatures, and increased ocean acidity are among the threats to the natural systems and biodiversity we rely on. Proposed work should address one or more of the topics below and create generalizable, scalable tools that can be used by other environmental researchers and conservationists.
Biodiversity: Biodiversity around the world is rapidly declining. AI can help us understand and address challenges such as these:
- Human-wildlife conflict
- Invasive species introduction and spread
- Habitat loss, including agricultural and urban encroachment
- Species discovery, identification, and distribution
- Wildlife poaching and trafficking
Climate Change: Countries and communities around the world are engaged in climate resilience, adaptation, and mitigation efforts. AI can help in areas such as these:
- Understanding and quantifying carbon sequestration
- Monitoring and understanding afforestation/reforestation efforts
- Identifying fire risks and ways to mitigate damage
- Extreme weather and climate modeling
- Sustainable land-use change
- Resilience to impacts caused by extreme events (e.g., droughts, floods, other natural disasters)
All models supported through this grant must be open source, and grant recipients must be willing to publicly share both the underlying data and their developed solutions—including, for example, applications, APIs, and Python/R packages—for use by other environmental researchers. More specifically, grant recipients must do the following:
- Release source code for projects publicly under the MIT license and feature their work on the Microsoft AI for Earth Technical Resources page
- Make the training data on which their solutions are developed publicly available in a standard digital format
- Develop and deploy their solutions on Microsoft Azure