Scientists make ‘rare advance’ in tackling the oldest unsolved problem in physics

Jupiter's red spot, a massive storm system, is an example of turbulent flow. NASA Goddard Space Flight Center via CNN Newsource

Scientists have made a leap forward in understanding the pattern and structure of turbulence — a natural phenomenon observed in fluids such as moving water, ocean currents, chemical reactions, blood flow, storm clouds, plumes of smoke and even the plasma of stars.

While turbulent flow is chaotic and irregular, as a fluid’s movement causes larger vortices or eddies to form and break down into smaller ones, physicists have long attempted to study and model the process using mathematical equations and computers. However, even with modern supercomputers, a direct and accurate simulation of all but the simplest turbulent flows remains out of reach, and a complete understanding of turbulence has eluded researchers for some 200 years.

Now, an international team of scientists has pioneered a new approach to simulating turbulence that deploys a quantum computing-inspired method, described in a study published January 29 in the journal Science Advances.

The ability to accurately model and predict the phenomenon could have many practical applications in science and engineering, potentially improving the design of airplanes, cars, propellers, artificial hearts and making weather prediction more accurate, said the study’s lead author Nik Gourianov, a researcher in the department of physics at the University of Oxford.

“Turbulence was and still is an unsolved problem in the sense that we cannot exactly simulate realistic flows on computers, i.e. we still need a wind tunnel to design an aircraft wing. But advances such as ours ‘chip away’ at the problem and push the frontier,” Gourianov said.

Most prior approaches to simulating turbulence relied on a deterministic strategy, which with a specific set of initial conditions always produces the same results, Gourianov explained. Instead, the new research modeled the fluctuations in turbulence probabilistically, an approach that takes into account random variation.

The team applied a quantum computing-inspired algorithm to turbulent flows, allowing them to compute in a few hours what would take a classical algorithm several days to do on an entire supercomputer.

Quantum computers process information in a fundamentally different way from classical computers. Traditional computers do calculations using bits: data that exists in one state at a time, a one or a zero. Quantum computers use quantum bits (or “Qbits), which can be zeros, ones or any combination of both. The study authors used a mathematical tool called tensor networks that can be used to simulate a quantum system.

James Beattie, a postdoctoral research associate and fellow in the department of astrophysical sciences at Princeton University in New Jersey, said that, by representing data with many variables in a simpler way, the team had been able to speed up complex calculations necessary to begin to understand turbulence. Beattie was not involved in the research.

“The simulation they are running is a fluid simulation of two different chemicals mixing and reacting. By using this representation, it means that this rather complex calculation can use significantly less memory, allowing it to be run on a laptop,” Beattie added.

“Seeing advances like this (a million times better utilization of memory and a thousand times speed-up in computation) is rare, making this an exciting advancement in the modeling of turbulence,” he said.

The mystery of turbulence

While the latest study is “amazing progress,” it is not the full story, Beattie added, noting that it doesn’t address issues of scale, or how turbulent vortices of different sizes relate to one another.

“Turbulence, as the authors say, is a multi-scale problem, i.e., turbulence can span from thousands of lightyears to less than a foot,” he said via email. “We want to know how these scales talk to each other.”

“This is an aspect that makes simulating turbulent fluids so challenging — we want to resolve many, many scales in the simulation, which take up lots and lots of memory and computation, which means putting these simulations on large supercomputers,” Beattie said.

The new research is “highly impressive,” said Yongxiang Huang, a researcher and associate professor at State Key Laboratory of Marine Environmental Science & College of Ocean and Earth Sciences at Xiamen University in southeastern China.

Gourianov and team have come up with a novel method that significantly reduces memory usage and computational complexity, Huang, who was not involved in the study, said. However, he agreed it did not paint a complete picture, which he said was extremely difficult because of the broad range of scales involved.

Turbulence has been described as the oldest unsolved problem in physics. The German theoretical physicist Werner Heisenberg allegedly said on his deathbed, “When I meet God, I am going to ask him two questions: Why relativity? And why turbulence? I really believe he will have an answer for the first.”

Gourianov said that the computational advantage of the new technique revealed by the study opens up new, previously inaccessible areas of turbulence physics for scientific investigation, although the findings don’t really mean the mystery of turbulence had been unraveled. That, he said, would require drastically new algorithms or computing hardware relative to what is available now.

“Many (exceptionally talented and gifted) scientists have looked at this problem, yet we are still not even close to solving it,” Gourianov said.