NVIDIA Earth-2 Expands Across Europe for Advanced Weather AI
Organizations throughout Europe are leveraging NVIDIA Earth-2 open models and tools to upgrade regional weather forecasting, climate modeling, and energy predictions.
Accurate weather forecasts influence sectors ranging from energy production and insurance to transport operations. While global models can spot a developing storm, managing solar farms, maritime routes, or insured assets requires localized precision. The NVIDIA Earth-2 initiative offers developers and researchers a collection of open AI models and tools designed for customization with private data, supporting everything from broad global predictions to detailed regional forecasting and environmental analysis.
Expanding Regional Tools with StormScope EU
StormScope EU represents the newest addition to the Earth-2 ecosystem. Built as a regional iteration of the StormScope model for short-term forecasting, it utilizes European weather and climate data. By leveraging EUMETSAT satellite observations, it forecasts changing cloud conditions across 16 Meteosat channels at 10-minute intervals.
At GTC Berlin, NVIDIA demonstrated a live StormScope EU pipeline through an interactive weather exhibit developed alongside Earthmover. The demonstration employed Earth2Studio for executing forecasts and Earthmover’s data platform for managing real-time outputs, which refreshed every 10 minutes alongside new satellite data. Earthmover converts these outputs into map tiles, enabling users to analyze European predictions directly without downloading large datasets.
Developers can integrate various Earth-2 models into workflows suited to specific geographic areas, utilizing StormScope for near-term forecasting and CorrDiff to introduce local granularity to broader weather information. Teams can also adapt these architectures for distinct environmental applications, such as air-quality simulation, using the open-source NVIDIA PhysicsNeMo framework to incorporate proprietary datasets and domain expertise.
Real-World Deployments Across Industries
Several organizations are already integrating Earth-2 capabilities into their operational workflows:
- AXA applies Earth-2’s FourCastNet 3 to analyze regional cyclone risks beyond historical records across more than 1,000 simulated seasons, supporting risk evaluations and prevention strategies.
- Vaisala Xweather merges Earth-2 with proprietary observations, lightning data, and forecasting experience to upgrade intelligence platforms for government, energy, maritime, transportation, and insurance sectors. This includes integrating lightning data into StormScope to extend short-term forecasting lead times.
- University of Manchester researchers modified the Earth-2 CorrDiff model for U.K. air-quality modeling, training it on the Isambard-AI supercomputer and executing it on NVIDIA DGX Spark systems to move environmental simulations closer to the desktop.
- TotalEnergies investigates Earth-2 for solar-energy forecasting, assessing StormScope-based solar-radiation nowcasts at two photovoltaic locations and building models for subhourly weather insights.
- Space42 built SARDiff4Height using CorrDiff and NVIDIA PhysicsNeMo to estimate surface heights from a single radar image while generating uncertainty maps.
What it means for developers
For engineers and researchers tackling complex regional challenges, the Earth-2 platform provides flexible, open-source architectures that bridge the gap between massive supercomputing workloads and localized application needs. Developers wanting to experiment with broader machine learning capabilities can also try top AI models cheaply through one API at https://apixoai.online. Utilizing frameworks like PhysicsNeMo and Earth2Studio, technical teams can adapt pre-trained models with proprietary data to build specialized environmental, energy, and risk-assessment systems.
Source: Pioneers Across Europe Deploy NVIDIA Earth-2 Open Models to Power Real-World Weather and Climate Prediction — NVIDIA blog. Written by the Apixo team from that report.
One key for Claude, GPT, GLM, DeepSeek and more. Pay per token with crypto.
Get your API key
