NVIDIA adds European forecasting model as Earth-2 applications expand

NVIDIA has added StormScope EU to its Earth-2 family of open AI models and tools for weather prediction and environmental research. Trained on European weather and climate data, the regional model uses EUMETSAT satellite observations to predict cloud conditions across 16 Meteosat channels at 10-minute intervals.

At GTC Berlin, NVIDIA is demonstrating a live forecasting pipeline built with Earthmover. Earth2Studio runs forecasts, while Earthmover manages real-time results and serves them as map tiles, allowing users to explore European predictions without downloading entire datasets. Developers can combine StormScope’s short-term forecasting with CorrDiff’s local detail and use NVIDIA PhysicsNeMo to customize models with proprietary data.

AXA is using FourCastNet 3 to investigate regional cyclone risk beyond historical records across more than 1,000 simulated seasons. Vaisala Xweather is combining Earth-2 with proprietary observations and lightning data to improve weather intelligence and short-term warnings for government and commercial customers.

University of Manchester researchers adapted CorrDiff for U.K. air-quality modeling, training it on the Isambard-AI supercomputer and running it locally on NVIDIA DGX Spark systems. TotalEnergies is evaluating StormScope-based solar-radiation forecasts at two photovoltaic sites to help anticipate generation changes. Space42 used CorrDiff and PhysicsNeMo to develop SARDiff4Height, which estimates surface heights from a single radar image and produces uncertainty maps.