🏔️ K2: Turbulence

Understanding and modeling turbulent fluid flow — one of the deepest unsolved problems of classical physics. Nine paths to the summit. Together, we climb.

Andrey Kolmogorov — the sherpa guide for K2 – Turbulence
Our Sherpa guide welcomes us: Hi! I'm Andrey Kolmogorov 👋, a towering Soviet mathematician 🇷🇺 who fundamentally transformed modern mathematics by establishing the rigorous axiomatic foundation for probability theory 🎲. My work turned probability from a loose set of statistical intuitions into a strict, logical branch of mathematics based on measure theory 📐. Let me tell you a funny story! 😄 It's about my overconfidence — and a brilliant 23-year-old who made me cancel a whole conference! 🎤 I was holding a major seminar where I confidently declared a long-held belief: that standard multiplication was fundamentally inefficient and couldn't possibly be sped up ✖️ — I strongly conjectured the standard method was the "best possible" algorithm. A young student in the audience, Anatoly Karatsuba 🧑‍🎓, listened to my lecture, went home, and developed an entirely new, significantly faster algorithm ⚡ (now known as the Karatsuba algorithm), and showed me the proof the very next day! 📄 I was so astonished 😲 — and so frustrated that my conjecture had been completely demolished overnight 💥 — that I canceled the rest of the conference! 🚪
K2 — the real mountain, the savage mountain
The "Savage Mountain": It is widely considered the world's most dangerous mountain to climb ☠️ due to its extreme technical difficulty, steep icy slopes 🧊, and severe, unpredictable storms 🌨️. The 1902 expedition to K2 is famous for involving the infamous occultist Aleister Crowley 🔮. Ravaged by illness and bad weather 🤒, the team bickered constantly 😤. At one point, a fever-delirious Crowley pulled a revolver 🔫 on a fellow climber who wanted to turn back. Despite the drama 🎭, he was the first to correctly identify the Abruzzi Spur as the best route 🧭. The Abruzzi Spur later became the standard path to the summit ⛰️ for almost all climbers.

Turbulence is the swirling, chaotic motion of fluids that surrounds us everywhere — in rivers and oceans, in the atmosphere and inside jet engines, in the blood in our arteries and the gas between the stars. Despite governing so much of the physical world, it remains one of the great unsolved challenges of classical physics: we can write down the equations of fluid motion, yet we still cannot fully predict or explain the tangle of eddies they produce.

Below are nine paths to the summit — distinct approaches to understanding and modeling turbulence, from Kolmogorov's classical scaling laws to modern machine learning. Within each path, the base camps are the coherent subtopics we will study in depth, each with its stepping stones and the key references that light the way.

Choose Your Path

Path 1. Classical Phenomenology and Scaling Laws

Kolmogorov's energy cascade, universal scaling laws, intermittency, and two-dimensional turbulence.

Path 2. Statistical Closures and Field-Theoretic Approaches

Closure approximations, the direct-interaction approximation, and renormalization-group field theory.

Path 3. Dynamical Systems and Chaos in Turbulence

Routes to chaos, strange attractors, the Lorenz system, and shell models of the cascade.

Path 4. Coherent Structures, Vortex Dynamics, and Geometry

Vortex stretching, coherent structures, and exact coherent states in turbulent geometry.

Path 5. Canonical Flows, Experiments, and Measurement Techniques

Wall-bounded flows, jets and wakes, and the experimental techniques that measure them.

Path 6. Modeling for Prediction: RANS, LES, Hybrids, and Uncertainty Quantification

Predicting turbulence with RANS, LES, hybrid methods, and uncertainty quantification.

Path 7. Lagrangian Turbulence and Mixing

Following fluid particles: dispersion, passive-scalar mixing, and Lagrangian coherent structures.

Path 8. Geophysical and Rotating/Stratified Turbulence

Rotating and stratified flows, and the turbulence of oceans, atmospheres, and planets.

Path 9. Machine Learning and Data-Driven Turbulence

Machine learning and data-driven modeling, reduced-order models, and deep-learning closures.
← Back to Home