Skip to main content

Supercomputing - Accelerating climate research

November 02, 2015 - The next-generation Earth system model will simulate climate systems at unprecedented resolution over an unprecedented time scale in order to understand climate change, Earth system feedbacks and potential tipping points. The Accelerated Climate Model for Energy project, led by Oak Ridge National Laboratory’s Peter Thornton, aims to answer the most pressing climate prediction questions, including how precipitation patterns will change in a warming climate and how the melting Antarctic ice sheet will contribute to sea level rise. Using Titan, the nation’s fastest supercomputer for open science research, the team has carried out a series of first-ever simulations in which human activity (fossil fuel emissions and agricultural commodity prices) was synchronously predicted with climate system responses (atmospheric composition and ecosystems). The fully coupled model, which will be ready for production around mid-2017, will be capable of simulating 80 years (1970-2050) of climate variability in less than three weeks.