Thierry Aubin

In the vast landscape of modern mathematics, few have navigated the intricate terrain between geometry and analysis with the skill and insight of Thierry Aubin. His work addressed a fundamental quest: the search for "ideal" or "canonical" forms within the bewildering variety of abstract shapes, or manifolds. This pursuit, seemingly abstract, holds the key to understanding the deep structure of space itself. Aubin's genius lay in his ability to translate profound questions about geometric shape into the concrete world of partial differential equations, developing powerful analytical tools to solve problems that had stumped mathematicians for decades.

This article explores the enduring legacy of Thierry Aubin's contributions. The journey is structured into two main parts. In the first chapter, ​​Principles and Mechanisms​​, we will delve into the mathematical heart of his work, dissecting his breakthroughs on the two monumental challenges that defined his career: the Yamabe problem and the Calabi conjecture. We will uncover the nature of these problems and the brilliant analytical strategies he deployed to overcome them. The second chapter, ​​Applications and Interdisciplinary Connections​​, will broaden our view to reveal how these solutions in pure mathematics unexpectedly rippled across science, building astonishing bridges to Einstein's theory of General Relativity and providing the essential geometric framework for modern String Theory.

Imagine you are a cartographer from a bygone era, tasked with creating the most perfect map of the world. You're not concerned with political boundaries, but with the very fabric of the globe itself. Your goal is to represent the Earth's curved surface on a flat piece of paper in such a way that the distortion, while unavoidable, is somehow uniform everywhere. This is the essence of the grand challenges that Thierry Aubin dedicated his life to solving—not for the 2-dimensional Earth, but for abstract spaces of any dimension. His work lies at a beautiful crossroads where the geometry of shapes meets the analysis of functions, transforming questions about "ideal forms" into profound problems in the world of differential equations.

The world of geometry is filled with a dazzling variety of shapes, or ​​manifolds​​ as mathematicians call them. A central question has always been: can we find a "best" or "most canonical" version of a given shape? For two-dimensional surfaces, a stunningly complete answer is given by the Uniformization Theorem, which tells us that any surface can be smoothly deformed into one of three types: one with constant positive curvature (like a sphere), constant zero curvature (like a flat plane), or constant negative curvature (like a saddle). But what about in higher dimensions?

The full picture becomes far too complex. So, Hidehiko Yamabe proposed a more modest, yet still profound, question in 1960. He asked: can we at least find a metric within a given ​​conformal class​​ that has ​​constant scalar curvature​​? Let's unpack this. A conformal class is a family of shapes that are related by uniform, directionless stretching at every point. Think of a perfectly round balloon versus one that has been squeezed and distorted; they are in the same conformal class. The scalar curvature at a point is a single number that represents the "average" curvature there—how much a tiny sphere's volume deviates from a flat Euclidean sphere's volume. Yamabe's question, then, is whether we can always find a way to stretch or shrink a given manifold so that this average curvature becomes the same everywhere.

From Geometry to a Devilish Equation

The first step in tackling such a problem is to translate the geometry into the language of analysis—the language of equations. The transformation of scalar curvature under a conformal change of metric, say from an old metric $g$ to a new one $\tilde{g} = u^{\frac{4}{n-2}}g$ (where $n \ge 3$ is the dimension), is governed by a remarkable formula. This formula connects the new scalar curvature $R_{\tilde{g}}$ to a special operator called the ​​conformal Laplacian​​, $L_g$, acting on the stretching function $u$. The relationship is surprisingly elegant:

where $L_g = -a_n \Delta_g + R_g$, with $a_n = \frac{4(n-1)}{n-2}$ being a constant depending on dimension, $\Delta_g$ the standard Laplacian operator (a measure of how a function differs from its average value nearby), and $R_g$ the original scalar curvature.

If we want the new scalar curvature $R_{\tilde{g}}$ to be a constant, let's call it $\lambda$, then our geometric problem becomes the task of finding a positive function $u$ that solves the following partial differential equation, now known as the ​​Yamabe equation​​:

This looks like a standard problem from physics or engineering, but it hides a venomous sting in its tail: the exponent $\frac{n+2}{n-2}$.

The Tyranny of the Critical Exponent

That seemingly innocuous exponent is no random number. It is the infamous ​​critical Sobolev exponent​​. Its "criticality" stems from a peculiar symmetry. If you take a solution and "zoom in" on it—a mathematical operation called scaling—the structure of the Yamabe equation, because of this specific exponent, remains unchanged. You can think of it like a fractal that looks the same at different magnifications.

In the world of variational calculus, where one often finds solutions by minimizing an "energy" functional, this symmetry is catastrophic. It means that the standard tools for proving that a minimizer exists completely break down. A sequence of functions that drives the energy ever lower might fail to converge to a proper solution. Instead, all its energy can concentrate into an infinitesimally small region, forming a "bubble" that then vanishes from the manifold, leaving nothing behind,. This loss of compactness was the formidable barrier that stumped mathematicians for years. Proving the existence of a solution to the Yamabe equation was not just about finding a function; it was about proving that these energy bubbles couldn't just disappear.

Aubin's Breakthrough: Caging the Bubbles

This is where Thierry Aubin entered the scene with a moment of true genius. He recognized that these elusive bubbles were not just abstract concentrations of energy; they were, in essence, tiny, scaled-down copies of the round sphere $\mathbb{S}^n$, which is the quintessential solution to the Yamabe problem. He reasoned that a bubble could only form if a minimizing sequence contained enough "energy" to create one. The total energy available is given by the infimum of the Yamabe functional, a quantity now called the ​​Yamabe invariant​​, denoted $Y(M,[g])$.

Aubin's brilliant strategy was to prove that for a large class of manifolds, the Yamabe invariant is strictly less than the Yamabe invariant of the standard sphere: $Y(M,[g]) \lt Y(\mathbb{S}^n, [g_{\mathrm{round}}])$. If the manifold simply doesn't have enough energy to form a full sphere-bubble, then the bubbling phenomenon is energetically forbidden! This forces any minimizing sequence to be well-behaved and converge to a genuine solution,.

He accomplished this with an ingenious test-function argument. But here, the story takes another fascinating turn, revealing the subtle complexities of geometry. Aubin's method, which relied on analyzing the local geometry through the ​​Weyl tensor​​ (a measure of how a manifold deviates from being conformally flat), worked perfectly for dimensions $n \ge 6$. In these higher dimensions, the effect of the Weyl tensor was strong enough to guarantee the strict inequality he needed. However, for dimensions 3, 4, and 5, the influence of the Weyl tensor was too weak or was canceled out by other effects in his calculations. The gap in low dimensions would later be filled by Richard Schoen using different, powerful techniques related to general relativity's Positive Mass Theorem. But Aubin's breakthrough had paved the way, solving the problem for a vast range of cases and introducing the key idea that ultimately led to its complete resolution.

Aubin's influence extended to another monumental problem, this time in the elegant and more rigid world of ​​Kähler manifolds​​. These are complex manifolds—spaces where coordinates are complex numbers—endowed with a special geometric structure that harmonizes their complex and Riemannian properties. They are the natural higher-dimensional analogues of the surfaces studied in complex analysis.

In the 1950s, Eugenio Calabi posed a conjecture of breathtaking scope. He asked whether one could essentially prescribe the Ricci curvature of a Kähler manifold. The ​​Ricci curvature​​ is a more detailed measure of curvature than the scalar curvature. Calabi's conjecture asserted that for any plausible candidate Ricci form (one that is compatible with the manifold's underlying topology), one can find a unique Kähler metric that realizes it.

A particularly crucial case of this conjecture is the search for ​​Kähler-Einstein metrics​​, where the Ricci curvature form is simply a constant multiple of the metric form itself: $\operatorname{Ric}(\omega) = \lambda \omega$. Such metrics represent a state of perfect geometric equilibrium and have profound implications across mathematics and theoretical physics.

Just as with the Yamabe problem, this geometric question translates into a formidable nonlinear PDE, a type known as a complex Monge-Ampère equation. The existence of solutions was far from clear. Aubin, once again deploying his mastery of nonlinear analysis, provided the first major breakthrough. He developed a powerful technique called the ​​continuity method​​. The idea is to start with a version of the problem you know you can solve, and then continuously deform it into the hard problem you actually want to solve. If you can prove that a solution exists at every single step along this path by establishing robust a priori estimates, then you can show a solution exists at the very end.

Using this method, Aubin successfully proved the existence of Kähler-Einstein metrics for the entire class of compact Kähler manifolds with a negative first Chern class ($c_1(M) 0$), which corresponds to the case $\lambda 0$. This result, independently and nearly simultaneously proven by Shing-Tung Yau (who then went on to solve the $c_1(M)=0$ case, completing the proof of the Calabi conjecture), was a landmark achievement. It demonstrated that a vast family of complex manifolds admit these canonical, "Einstein" metrics.

Thierry Aubin's work stands as a testament to the power of analysis in service of geometry. He showed that the quest for the "best" shapes, for hidden uniformity in the fabric of space, could be pursued and won in the arena of differential equations. His legacy is not just in the theorems he proved, but in the beautiful and deep mechanisms he uncovered, revealing the intricate dance between the form of a space and the functions that live upon it.

The profound questions that Thierry Aubin tackled were not mere mathematical curiosities, isolated in the abstract realm of pure thought. Like a physicist probing the nature of a material by striking it and listening to its ring, his work on geometric analysis provided a set of powerful tools to "strike" the very fabric of space. The resulting "ring" was not a simple tone, but a rich harmony of insights that have resonated across vast and seemingly disconnected fields of science, from Einstein's theory of gravity to the strange, extra-dimensional world of string theory. Aubin’s legacy is not just in the answers he found, but in the beautiful and unexpected connections he helped to uncover.

Our geometric intuition is often forged on two-dimensional surfaces. We know from the celebrated Uniformization Theorem that any closed surface, be it a sphere, a donut, or a multi-holed pretzel, can be conformally stretched—without tearing—into a shape of perfectly constant Gaussian curvature. A sphere becomes perfectly round, a donut becomes perfectly flat (like a video game screen that wraps around), and a pretzel acquires a perfectly uniform negative curvature, like a fragment of Escher's hyperbolic worlds. The mathematical equation governing this transformation is a well-behaved partial differential equation (PDE) whose primary component is the familiar Laplacian operator, making the problem analytically manageable.

One might naturally ask: can we do the same in higher dimensions? Can we take any closed 3-dimensional, 4-dimensional, or n-dimensional space and find a conformally related shape of constant curvature? This is the essence of the Yamabe problem. However, the moment we step beyond two dimensions, the world changes dramatically. Gaussian curvature, a single number at each point that captures the full curvature of a surface, gives way to a complex bestiary of curvature tensors. The most natural "average" curvature to work with is the scalar curvature. The quest thus becomes: can we find a metric with constant scalar curvature in any conformal class?

This seemingly small change—from Gaussian to scalar curvature—transforms the underlying PDE from a relatively tame one into a formidable beast. The equation one must solve, the Yamabe equation, is what mathematicians call "critically nonlinear." The nonlinearity appears with an exponent, $u^{\frac{n+2}{n-2}}$, which is precisely at the knife-edge of what standard analytical techniques can handle. This "critical Sobolev exponent" isn't a mere technicality; it’s a sign that the problem has deep physical meaning related to scale invariance, and its solution would require entirely new ideas.

The modern attack on the Yamabe problem, pioneered by Neil Trudinger and advanced by Aubin, was through the calculus of variations. Imagine all possible shapes in a given conformal class as a landscape. Finding a metric with constant scalar curvature is equivalent to finding the lowest point in this landscape—the configuration with the minimum possible "total scalar curvature," properly normalized. The value at this minimum is a fundamental characteristic of the shape's conformal class, called the Yamabe invariant, $Y(M,[g])$.

The central difficulty is that a sequence of shapes might "descend" toward the minimum, but instead of settling into a nice smooth shape, it could develop an infinitely sharp spike at a single point, a phenomenon called "bubbling" or "concentration." Trudinger showed that there is a universal energy barrier for this to happen: the Yamabe invariant of the perfect n-sphere, $Y(\mathbb{S}^{n})$. If one could prove that the Yamabe invariant of a given manifold was strictly less than that of the sphere, $Y(M,[g]) \lt Y(\mathbb{S}^{n})$, then bubbling would be energetically forbidden, guaranteeing that a smooth, constant scalar curvature solution exists.

This is where Aubin made his brilliant contribution. He devised a method to "probe" the energy landscape. He constructed ingenious "test functions," which acted like tiny, concentrated bubbles of curvature placed at a point on the manifold. By calculating the energy of these test functions, he could get an estimate for the Yamabe invariant. His calculations showed that for dimensions $n \ge 6$, if the manifold was not "locally conformally flat"—if its Weyl curvature tensor, which measures the deviation from being locally sphere-like, was non-zero at some point—then one could place a bubble there and achieve an energy strictly less than the sphere's. Aubin's local, analytical probe was powerful enough to solve the Yamabe problem for a huge class of manifolds.

Yet, a stubborn gap remained. In dimensions 3, 4, and 5, and for any manifold that was locally conformally flat but not globally a sphere (like the product of a sphere and a circle), Aubin's local curvature expansion wasn't strong enough to produce the desired strict inequality. The problem was stalled.

The resolution, provided by Richard Schoen, is one of the most breathtaking instances of interdisciplinary thought in modern mathematics. Schoen’s idea was to re-imagine the "bubbling" point not as a pathology, but as a portal. He showed that if a manifold's Yamabe invariant were to equal the sphere's, it implied the existence of a mathematical structure that looked like a complete, asymptotically flat universe with zero scalar curvature, emerging from the puncture point. Such objects are the domain of Einstein's theory of general relativity, and they possess a globally defined quantity: their total mass-energy, or ADM mass. The celebrated Positive Mass Theorem (PMT), proven by Schoen and Yau, states that for any such universe with non-negative local energy density (i.e., non-negative scalar curvature), its total mass must be non-negative. Moreover, the mass can be zero only if the universe is completely empty—flat Euclidean space.

Schoen connected the ADM mass of this fictitious universe directly to the Yamabe problem. He demonstrated that the very condition that had stymied Aubin's method—$Y(M,[g]) = Y(\mathbb{S}^{n})$—would force the ADM mass of this associated universe to be zero. By the rigidity of the PMT, this meant the manifold had to be conformally identical to the sphere. In other words, the only way for a manifold to fail the test for existence was to already be the trivial case! This beautiful contradiction, bridging geometric analysis and general relativity, finally closed the Yamabe problem for all manifolds.

Aubin's work also played a crucial role in another monumental quest: the search for "canonical" metrics on complex manifolds. These are spaces that locally look like $\mathbb{C}^n$ and possess a rich geometric structure. The goal here is to find a Kähler-Einstein (KE) metric—one that is not only compatible with the complex structure (Kähler) but also satisfies a stringent equilibrium condition: its Ricci curvature is directly proportional to the metric itself, $\operatorname{Ric}(\omega) = \lambda \omega$.

The existence of such a perfect metric is governed by a topological invariant called the first Chern class, $c_{1}(M)$. This class acts like a topological "charge" that divides the landscape of complex manifolds into three distinct realms.

​​The Unobstructed Realms: $c_{1}(M) \le 0$​​

The cases of negative or zero Chern class are the domains of the celebrated Aubin-Yau theorem.

• When $c_{1}(M) \lt 0$, the manifold has negative topological charge. Here, Aubin and Yau proved that a unique Kähler-Einstein metric with negative curvature ($\lambda \lt 0$) always exists. A simple one-dimensional example is a compact Riemann surface of genus $g \ge 2$ (a pretzel with two or more holes), which always admits a unique hyperbolic metric of constant negative curvature.
• When $c_{1}(M) = 0$, the manifold is topologically neutral. This is the stage for the famous Calabi Conjecture, which proposed that in this case, a unique Ricci-flat ($\lambda=0$) KE metric should exist within every possible Kähler class. Yau's celebrated proof of this conjecture, a technical tour de force in solving the complex Monge-Ampère equation, relied on the foundational analytical estimates pioneered by Aubin. This result guarantees, for instance, that complex tori and the enigmatic K3 surfaces can all be endowed with these special Ricci-flat metrics.

​​The Crown Jewels: Calabi-Yau Manifolds and String Theory​​

The Ricci-flat metrics produced by the Aubin-Yau theorem for $c_1(M)=0$ manifolds are far from being mathematical toys. They are the protagonists in one of the grandest stories of modern physics. When a Kähler metric is Ricci-flat, it acquires a magical property. The group of symmetries one observes when parallel transporting vectors around loops on the manifold—the holonomy group—shrinks. For a generic complex n-dimensional manifold, this group is the unitary group $U(n)$. But for a Ricci-flat one, the existence of a now-parallel, nowhere-vanishing holomorphic volume form forces the holonomy to be contained in the special unitary group, $SU(n)$.

These special-holonomy manifolds, known as Calabi-Yau manifolds, turned out to be precisely what physicists were looking for in the 1980s. String theory, a candidate for a "theory of everything," requires that spacetime have ten dimensions. To match our observed four dimensions, the theory posits that the extra six are curled up into a tiny, compact space. For the theory to be consistent with observations (like the existence of fermions and the preservation of supersymmetry), this six-dimensional space cannot be arbitrary; it must be a 3-complex-dimensional, Ricci-flat Kähler manifold—in other words, a Calabi-Yau manifold. The geometry of these spaces, guaranteed to exist by the work of Aubin and Yau, dictates the physics of our world, from the spectrum of elementary particles to the values of fundamental constants.

​​The Obstructed Realm: $c_{1}(M) \gt 0$​​

When the topological charge is positive, $c_{1}(M) \gt 0$, the situation becomes vastly more complex. Here, even the powerful methods of Aubin and Yau hit a wall. Existence of a Kähler-Einstein metric is no longer guaranteed. Obstructions, rooted in the manifold's algebraic symmetries, can appear. The presence of certain holomorphic vector fields can generate a non-zero "Futaki invariant," which acts as a definitive barrier to a KE metric's existence. The modern resolution, known as the Yau-Tian-Donaldson correspondence, reveals another profound connection: the existence of a KE metric is equivalent to a purely algebraic notion called "K-polystability." A problem that began in the world of differential geometry finds its ultimate answer in the language of algebraic geometry.

Thierry Aubin's journey into the world of curvature and nonlinear equations reveals a recurring theme in science. The determined pursuit of fundamental, often abstract, questions about the nature of space can lead to unexpected and spectacular revelations. His work provided the crucial analytical bedrock for solving both the Yamabe problem and the Calabi conjecture, but its true impact lies in the bridges it helped build—from the local analysis of PDEs to the global structure of the cosmos, and from the abstract beauty of complex geometry to the concrete physics of string theory. It is a testament to the fact that in the search for truth, the deepest paths are often the ones that connect the stars.