What if you could build a digital twin of the Earth? That’s the idea behind Earth-2, NVIDIA’s generative AI platform for climate simulation and weather prediction. It’s designed to make forecasting faster, more accurate, and more energy efficient. Earth-2 is critical as climate change accelerates and extreme weather events become more disruptive and costly.
“We’re trying to use AI across the Earth system modeling stack to improve the accuracy of climate simulations and their relevance for society.” — Dr. Michael Pritchard, Director of Climate Simulation Research at NVIDIA Earth-2.
This video is also available on YouTube.
Learn more about the Earth-2 Platform for Climate Change Modeling.
Try Earth-2 Interactive Visualization Demo
Update: May 14, 2025. Check out this incredible new paper published by Dr. Pritchard and his team: Climate in a Bottle: Towards a Generative Foundation Model for the Kilometer-Scale Global Atmosphere.
Authors: Noah D. Brenowitz, Tao Ge, Akshay Subramaniam,Aayush Gupta, David M.Hall, Morteza Mardani, Arash Vahdat, Karthik Kashinath, Michael S. Pritchard
Venue: (NVIDIA, Santa Clara, CA, USA)
Date: May 14, 2025
Abstract
AI emulators offer a path to compressing, boosting limited ensembles, and improving the latency of interacting with petabyte-scale climate prediction data. However, prevailing auto-regressive paradigms offer limited flexibility, and are challenging to train on climate time horizons due to drifts, instabilities and component-coupling challenges. Conditionally generative models offer an appealing alternative. In this context we demonstrate a generative diffusion-based framework—Climate in a Bottle (cBottle)—for emulating global km-scale climate simulations and reanalysis on the equal-area HEALPix grid. cBottle consists of two model stages: a globally-trained coarse-resolution image generator that generates 100km (50k-pixel) fields given monthly average sea surface temperatures and solar conditioning, followed by a locally-trained 16x super-resolution stage that generates 5km (12.5M-pixel) fields; global super-resolution is made affordable using an overlapping patch-based multi-diffusion. Overall, cBottle shows promise as an emulator across a battery of climate model diagnostics, including diurnal-to-seasonal scale variability, large-scale modes of variability, tropical cyclone statistics, and trends of climate change and weather extremes. Moreover, cBottle is a step towards a foundation model, by bridging multiple data modalities (reanalysis and simulation) with corresponding utility beyond emulation to tasks such as zero-shot bias correction, climate downscaling, and channel in-filling. The code is available at https://github.com/NVlabs/cBottle.
#Earth2 #NVIDIA #ClimateTech #AI #GTC2025 #WeatherForecasting #GenerativeAI #DisasterPreparedness #ClimateChange #drmichaelpritchard #michaelpritchardnvidia #climateinabottle #climatemodelingnvidia #cbottle #cbottleNVIDIA
Share this post