San José State University researchers will join a new $30 million National Science Foundation Science and Technology Center dedicated to understanding turbulence, a problem that has resisted clean solutions for more than a century. The center, called TEMPEST, will apply a combination of physics, mathematics, and AI to turbulent systems ranging from fusion plasmas to hypersonic flight.
Led by Michigan State University, TEMPEST brings together physicists, mathematicians, engineers, computer scientists, and AI researchers from institutions across the country. The center's goal is to develop predictive models for turbulence that work across the enormous range of spatial and temporal scales found in real-world systems - scales that no single computational method can currently handle.
Two SJSU researchers, two approaches
Liam Stanton, associate professor of applied mathematics at SJSU, will focus on multiscale transport and turbulent mixing in extreme environments, particularly fusion energy and high energy-density plasmas. His work connects microscopic particle interactions with continuum-scale transport, a gap that matters because fusion experiments span an extraordinary range of physical scales.
"Fusion experiments involve physics spanning an extraordinary range of scales," said Stanton. "Understanding how microscopic plasma interactions ultimately affect turbulent mixing and energy transport at much larger scales is a problem no single researcher or computational method can solve. TEMPEST gives us an opportunity to connect those scales by bringing experiments, mathematics, large-scale simulation, and AI together."
Mike Wood, assistant professor with appointments in SJSU's Moss Landing Marine Laboratories and Department of Computer Science, brings a different angle. He will use machine learning and neural networks to develop turbulence closures - approximations that simulations rely on when they cannot explicitly resolve every turbulent scale. Even the most powerful computers today cannot resolve all the detail in complex flows, so simulations must approximate the effects of unresolved turbulence. Wood's research will use high-resolution simulation data to improve those approximations.
"Turbulence is everywhere - from the oceans and atmosphere here on Earth to supernovae in space," said Wood. "What's exciting about TEMPEST is the opportunity to bring together researchers from across the country who study turbulence in very different physical systems. By combining these perspectives, we aim to uncover common principles that improve our ability to understand and predict turbulent flows across an extraordinary range of scales and applications."
Moving from calibration to prediction
A central goal of the research is to shift from models that reproduce experiments after the fact to models that predict turbulent transport before an experiment runs. That distinction matters for fusion, where improved predictive capability could directly shape the computational tools used to design inertial and magneto-inertial fusion experiments.
The center will test its methods on non-canonical turbulence - systems where extreme conditions, cross-scale interactions, and multiple coupled physical processes break conventional modeling assumptions. Application areas include fusion plasmas, astrophysical flows, and hypersonic aerospace environments.
In addition to Michigan State and San José State, TEMPEST includes researchers from Baylor University, Auburn University, the University of Rochester, Yale University, Texas A&M University, and the Georgia Institute of Technology. Los Alamos National Laboratory, Sandia National Laboratories, Lawrence Livermore National Laboratory, Pacific Fusion, and General Atomics participate as unfunded partners.
Student involvement and workforce training
SJSU students will work directly on interdisciplinary projects in applied computational fluid dynamics, scientific computing, machine learning, and data science. The center also plans to make research data and software broadly available and to develop educational programs aimed at training an AI-fluent scientific workforce.
"Participating in a center of this caliber represents a significant milestone for SJSU," said Crist Khachikian, vice president for Research and Innovation. "Science and Technology Centers are among the National Science Foundation's largest and most competitive investments, and our researchers have earned a place alongside colleagues from the nation's leading research universities."
Why this matters for science and research professionals
For researchers working in computational science, this center signals a shift in how large-scale scientific problems get solved. The combination of physics-based multiscale modeling with machine learning is becoming a standard approach for problems that resist traditional methods - and the funding structure means those methods will now be developed and tested across national laboratories and major scientific facilities. For professionals in fields adjacent to turbulence research, the center's open data and software policies mean the tools and methods developed here will be available for broader use, not locked inside individual institutions.
Your membership also unlocks:
