SCREAM wins Gordon Bell Local weather Prize at SC23
Sandia researcher Mark Taylor, the mission’s lead computational scientist. (Photograph by Loni Anderson) Click on on the thumbnail for a high-resolution picture.
“Now we have created the world’s first cloud-resolved mannequin to simulate a worldwide local weather 12 months in a single day,” mentioned Mark Taylor, a Sandia researcher and lead computational scientist for the Exascale Earth Vitality Methods Mannequin, or E3SM, an eight-lab mission. By Lawrence Livermore Nationwide Laboratory and supported by the Division of Vitality’s Workplace of Science to develop superior local weather fashions. “We’re ushering in a brand new period of precision.”
The E3SM mannequin simulates vital elements of the Earth’s local weather system and predicts the potential influence on circumstances in america within the coming many years, together with excessive temperatures, droughts, floods, and rising sea ranges. Taylor, who led Gordon Bell’s presentation, detailed the crew’s record-setting demonstration of SCREAM, the E3SM Cloud Decision Easy Atmospheric Mannequin, on Oak Ridge Nationwide Laboratory’s supercomputer, able to 1.2 exaflops – 1.2 quintillion – calculations per second.
Taylor mentioned clouds play an important position within the Earth’s local weather system, influencing climate patterns and precipitation. “Conventional Earth system fashions wrestle to precisely characterize clouds as a result of they can’t simulate the small atmospheric overturning circulation accountable for cloud formation. As an alternative, they depend on complicated approximations of those processes. This next-generation software program has the potential to dramatically scale back main systemic overturns,” he mentioned. : “Precipitation errors have been present in present fashions, attributable to their extra real looking and express therapy of convective storms and atmospheric actions accountable for cloud formation.”
Incorporating state-of-the-art parameters for fluid dynamics, microphysics, moist turbulence, and radiation, SCREAM is an entire atmospheric common circulation mannequin developed for high-resolution simulations on exascale machines. SCREAM, led by LLNL’s Peter Caldwell, achieved these new outcomes via shut collaboration between atmospheric scientists and computational scientists. The Gordon Bell Prize for Local weather Modeling, in keeping with the ACM, goals to acknowledge the contributions of modern parallel computing to fixing the worldwide local weather disaster. It was awarded for the primary time this 12 months on the Worldwide Convention on Excessive-Efficiency Computing, Networking, Storage and Evaluation, SC23, in Denver, and was accompanied by a $10,000 prize introduced by Gordon Bell, a pioneer within the area of high-performance computing. The Sandia-led mannequin was chosen based mostly on its potential to influence local weather modeling and associated fields.
/Normal launch. This materials from the unique group/creator(s) could also be chronological in nature, and is edited for readability, type and size. Mirage.Information doesn’t take company positions or events, and all views, positions and conclusions expressed herein are solely these of the creator(s).View in full right here.
