Earth-2: Nvidia’s Climate Digital Twin

Nvidia has launched Earth 2, a digital twin of the planet’s climate, to simulate and visualize weather and climate at an unprecedented scale.

The Weather Company and the Taiwan Weather Administration are among the first to leverage the new Nvidia Earth-2 APIs.

Fueled by advanced AI and processing technology, the platform can show weather patterns with incredible detail down to 2 kilometers. Nvidia touts this as a game changer in how we understand and forecast weather globally.

Announced as part of Nvidia’s CUDA-X microservices, the Earth-2 digital twin enables almost anyone to create AI-powered simulations of global atmosphere conditions, cloud covers, and extreme weather events like typhoons and hurricanes.

Nvidia says Earth-2’s application programming interfaces help users develop warnings and forecasts in mere seconds compared to minutes or hours in traditional CPU-driven modeling.

Nvidia Founder and CEO Jensen Huang said: “Earth 2 cloud APIs strive to help us better prepare for, and inspire us to act to moderate, extreme weather.”

With diffusion modeling, a new generative AI model called CorrDiff creates clear weather images 12 and a half times faster than traditional computing, according to Nvidia.

The AI can fix mistakes in detailed forecasts and extract deep insights for decision-making.

The Weather Company, parent company of the Weather Channel, plans to combine its in-house weather and data tools with Nvidia’s Omniverse to create digital models showing the real impact of weather conditions.

Taiwan’s Central Weather Administration hopes to more accurately predict where typhoons will hit using the advanced AI models. The intelligence helps prioritize early evacuations in collaboration with Taiwan’s National Science and Technology Center for Disaster Reduction.

As a critical link in the global supply chain, flooding risk analysis and evacuation preparedness are top priorities to the weather administration, according to Administrator Chia-Ping Cheng.

On the digital twin front, Nvidia also announced:

  • Project Ceiba, a supercomputer collaboration with Amazon Web Services aiming to make AI powered digital twins more realistic.
  • An expanded partnership with Siemens to improve industrial metaverse applications with AI driven digital twin technology.

Digital twins are virtual replicas of physical objects, systems, processes or services.
Integrated with AI, digital twins are dynamic, predictive models that can learn and adapt over time.

Global interest in digital twinning has grown alongside AI and machine learning.
The evolution toward intelligent digital twins is a significant shift toward autonomous, predictive, and prescriptive digital tools.