Case study

ZAMG takes a more precise look at the weather with HPC

Challenge

More precise forecasting
In Austria, if you want to know the weather, you turn to ZAMG. The oldest state weather service in the world and a subsidiary of the Austrian Federal Ministry of Education, Science, and Research, the Central Institution for Meteorology and Geodynamics (ZAMG) issues comprehensive weather data around the clock for all of Austria.
"We want to set new standards in forecasting," explains Michael Staudinger, director of ZAMG. "We can't make the weather better, but we can improve our forecasts."
Today, the mathematical and physical weather models at ZAMG calculate forecasts with a resolution of 2 km, accounting for smaller terrain structures and atmosphere more precisely than ever before. The volume and diversity of data processed each second is immense: satellite and radar imagery along with data about temperatures, precipitation, wind, and clouds.
Lacking computing power
ZAMG had been running its own high-performance computing (HPC) system since 2001. The programs and computers in its internal supercomputing cluster could no longer provide the computing power needed to generate hourly weather forecasts at a local level.