Imagine a lump of clay in the hands of a sculptor. Slowly, gradually, the shape changes – here some excess is trimmed away, there something missing is added. In the end, a beautiful statue emerges. That's more or less how geometric flows work – except instead of clay, we have mathematical spaces, and instead of the sculptor's hands, we have equations that slowly «reshape» geometry, making it more refined.
Not long ago, on the picturesque island of Corsica, the BRIDGES summer school gathered mathematicians from all over the world to discuss these fascinating processes. And today, I'll show you how numbers teach space to dance.
When space decides to change
In everyday life, we tend to think of space as something fixed. A room stays a room, a street stays a street. But mathematicians have long known: space can be flexible, alive, capable of transformation.
A geometric structure is like the «passport» of a space. It tells us how to measure distances, which lines count as straight, how to decide what's parallel and what's perpendicular. It's like the rulebook of a game played inside the mathematical universe.
Picture yourself looking at a map of Copenhagen. A regular map is flat – that's one geometry. But if you project the same map onto a globe, you get a completely different geometry. Distances change, «straight» lines bend. That's how different geometric structures work.
Geometric flows are a way of smoothly moving from one structure to another. As if our Copenhagen map were slowly, step by step, morphing from flat to spherical.
Two paths to perfection
Mathematicians have discovered two main approaches to finding the «ideal» geometry:
The first path – elliptic, is like solving a puzzle. You try to find the final answer right away: what should the perfect shape be? It's like carving a statue of David straight from marble, already knowing exactly where every contour belongs.
The second path – parabolic, is about gradual improvement. You start with what you have and slowly correct its flaws. Each day you remove a bit of excess, add a touch of what's missing. It's a more natural process – like evolution in nature.
Geometric flows belong to this second type. They follow the principle of «a little better each day».
The Ricci flow – the first star of the mathematical stage
The most famous geometric flow bears the name of Italian mathematician Gregorio Ricci. Its work is simple and elegant: it finds the spots where space is «too curved» and gradually smooths them out.
Think of an oddly shaped balloon. The Ricci flow slowly «redistributes» the air from places where there's too much to places where there's not enough. In the end, the balloon becomes perfectly round.
The Ricci flow equation looks deceptively simple: the rate of change of geometry is proportional to its curvature. The stronger the curvature – the faster it gets corrected.
But there's a catch. The Ricci flow is like a car with no steering wheel – it moves forward, but it could «drift» in any direction. Mathematically speaking, it doesn't notice rotations or shifts of the space itself. That creates serious technical hurdles when proving that a solution both exists and is unique.
DeTurck's trick – taming the untamable flow
In 1983, mathematician Dennis DeTurck came up with a brilliant trick. He figured out how to «anchor» the flow to a fixed coordinate system, removing the extra degrees of freedom.
Imagine photographing a pair of dancers. Without a tripod, the pictures blur – the camera moves along with them. A tripod locks the camera in place, letting us see the dance clearly. DeTurck's trick is that tripod for geometric flows.
This method not only made the Ricci flow «well-behaved», but also proved two fundamental facts:
- Existence: a solution always exists, at least for some initial stretch of time
- Uniqueness: that solution is the only one
These results laid the foundation for exploring more complex flows.
G₂-structures – the geometry of seven dimensions
Now let's step into true mathematical exotica. If ordinary spaces can still be pictured (at least in three dimensions), G₂-structures live in seven. It's like trying to explain color to someone born blind – our intuition simply doesn't reach that far.
They're named after one of the exceptional Lie groups – G₂. These structures are special: they define distances, angles, and orientation in seven-dimensional space all at once.
What's their magic? Imagine a special «compass» that, at every point in seven-dimensional space, shows you how to measure everything around correctly. That universal compass is a G₂-structure.
Every G₂-structure comes with a «torsion tensor» – a mathematical quantity that shows how far the structure strays from being perfect. It's like a calibration indicator for our seven-dimensional compass.
The three dances of G₂-structures
Mathematicians have designed several ways to improve G₂-structures using flows. Each flow has its own choreography:
The Laplacian flow – a classical dance
This is the most natural flow for G₂-structures. It acts like heat diffusion: «hot» spots (highly curved) cool down by spreading energy to «cold» spots (less curved).
Mathematically, it's driven by the Laplace operator – the same one that describes heat spreading through a rod or gas diffusing in a room.
The flow has a catch: it preserves certain key properties of G₂-structures, but only works if the starting point is already reasonably good.
The isometric flow – a dance to fixed music
This flow doesn't change distances in the space – it's like dancing where the tempo is fixed, but the steps may vary.
The isometric flow directly «repairs» the torsion tensor, making the G₂-structure more refined. It's less finicky than the Laplacian flow and works for any initial conditions.
The modified co-flow – a dance with double rhythm
The most intricate of the flows doesn't act on the G₂-structure itself, but on its «companion» – a special four-dimensional form. Like controlling a marionette with strings – you move your hands, and the puppet dances.
This flow too comes with guaranteed existence and uniqueness of solutions.
A universal formula for perfection
Recent work has revealed something remarkable: all reasonable G₂-flows can be expressed through a single universal formula. It has just six parameters, and depending on their values, you get different flows.
It's like discovering that all pieces of music can be written as combinations of a handful of basic rhythms. By tweaking the coefficients, you get the Laplacian flow, the isometric flow, and countless new ones yet to be explored.
Even more, mathematicians have identified conditions under which any such flow can be «tamed» by a method similar to DeTurck's trick.
Questions without answers
The geometry of G₂-flows is still a young field, full of mysteries. We know that solutions exist for short periods of time – but what happens afterward?
Some flows may last forever, gradually approaching an ideal geometry. Others may «blow up» in finite time – mathematically speaking, certain quantities shoot off to infinity.
There's also a practical angle. G₂-structures play a central role in string theory – one of the leading candidates for a «theory of everything» in physics. Perhaps understanding their flows will bring us closer to grasping the fabric of our universe.
Beauty in numbers
When I talk about this research, people often ask: «But what's the point»? The answer is simple: because it's beautiful.
Picture a seven-dimensional space, slowly dancing to the music of mathematical equations, moving ever closer to perfection. It's poetry in numbers, a symphony in formulas.
And who knows – maybe this is how reality itself works. Maybe our universe is one vast geometric flow, searching for its perfect form over billions of years.
After all, the most abstract mathematical ideas often end up in the most unexpected places. GPS in your phone works thanks to Einstein's relativity. Cryptography protects your bank transfers thanks to number theory. And G₂-flows... who knows what they'll bring us tomorrow?
Numbers don't lie. But sometimes, the stories they tell are so astonishing, they sound like science fiction.
