Stationary Varifolds

Nature often seeks efficiency, forming shapes that minimize energy, most famously illustrated by the minimal area of soap films. For centuries, mathematicians studied these "minimal surfaces" using the language of smooth functions and differential geometry. However, this classical approach falters when confronted with the complex reality of these shapes—what happens when films meet at junctions, form multiple layers, or exist as a fragmented "dust"? These scenarios expose a gap in our mathematical language, demanding a more powerful framework to describe objects that are surface-like but not necessarily smooth.

Stationary varifold theory, a cornerstone of geometric measure theory, rises to this challenge. It provides a revolutionary way to think about generalized surfaces, not through explicit equations, but through how they measure space and the orientation of their tangent planes. This approach is robust enough to handle the intricate singularities and multi-sheeted structures that classical methods cannot. This article navigates the foundational concepts and profound implications of this theory. The chapter on "Principles and Mechanisms" will define what a varifold is, explore the crucial concept of stationarity as a generalization of minimal energy, and uncover powerful analytical tools like the Monotonicity Formula. Subsequently, "Applications and Interdisciplinary Connections" will reveal how this abstract machinery solves concrete problems, from proving the smoothness of nearly-flat surfaces to establishing fundamental theorems in topology and general relativity.

Imagine a soap film. It's a beautiful, shimmering surface that magically finds the shape with the least possible area for a given boundary. Mathematicians have been fascinated by these "minimal surfaces" for centuries. But what if the film breaks and reforms into a more complex shape? What if it's not one smooth sheet, but several sheets meeting at an edge, like the junction of three soap bubbles? Or what if, in some bizarre physical process, you ended up with a "dust" of tiny, disconnected surface fragments? Our classical idea of a smooth surface, described by neat equations, simply breaks down. We need a new language, a more powerful and flexible way to think about "surface-like" objects.

This is where the idea of a ​​varifold​​ comes in. It's a bit of a strange beast at first, but the underlying idea is wonderfully intuitive. Instead of describing a surface by writing down equations for its coordinates, we describe it by how it "measures" space. A varifold is a type of ​​Radon measure​​, which is a formal way of saying it's a machine that tells you "how much surface" is in any given region of space. But it does more than that. For every point on the object, it also records the orientation of the surface element at that point—not in the sense of 'up' or 'down', but the direction of its tangent plane.

Think of it like this: A simple description of a forest might be a line on a map showing its boundary. A more sophisticated description, like a varifold, would be a detailed satellite image. For every pixel (a small region of space), it tells you not just if there's a tree there, but also how many trees are there (​​multiplicity​​) and which way they are leaning (the ​​tangent plane​​). An ​​integral varifold​​ is one where this multiplicity is a whole number—one, two, three layers of surface, and so on. This framework is powerful enough to describe a smooth sheet of paper, two sheets crossing each other, or even a stack of a hundred sheets lying on top of each other.

Nature is wonderfully "lazy." From the path of a light ray to the shape of a water droplet, physical systems tend to settle into a configuration of minimum energy. For a soap film, this means minimizing its area. A smooth surface that achieves this is called a ​​minimal surface​​, and it has the mathematical property that its ​​mean curvature​​ is zero at every point. The mean curvature is essentially a measure of how "bent" the surface is on average. A zero mean curvature means the surface is perfectly balanced, like a stretched rubber sheet that's flat in all directions, at least infinitesimally.

But what about our weird varifolds, which might not be smooth? We need a more general idea of "minimality." This is what we call ​​stationarity​​. Imagine you have a varifold. We can "deform" it slightly by applying a smooth vector field—think of a gentle breeze blowing on our surface dust. The ​​first variation​​, denoted $\delta V(X)$, tells us the initial rate of change of the total area of the varifold $V$ as we deform it according to the vector field $X$.

A varifold is called ​​stationary​​ if its first variation is zero for any smooth, compactly supported deformation we can think of. What does this mean? It means the varifold is at a critical point in the landscape of all possible shapes. Any tiny jiggle doesn't change its total area, to a first-order approximation. It's perfectly balanced. For a smooth surface, this condition of being stationary is exactly equivalent to having zero mean curvature. Thus, stationarity is the perfect and profound generalization of minimality to the wild world of varifolds.

To truly appreciate the power of stationarity, let's take a walk through a gallery of these remarkable objects.

• ​​The Classics (Smooth and Familiar):​​ The simplest stationary varifold is just a flat $m$-dimensional plane (or several planes stacked on top of each other) in a higher-dimensional space. Their curvature is zero everywhere, so of course they are stationary. A more exciting example is the ​​catenoid​​ in $\mathbb{R}^{3}$, the shape you get by revolving a catenary curve. It was the first non-trivial minimal surface discovered, and it's the shape a soap film makes when stretched between two circular rings. It is a beautiful example of a complete, smooth stationary varifold.

• ​​The Singular Wonders:​​ This is where the varifold theory truly shines. Consider two infinitely long lines crossing at the origin in a 2D plane. Can this be a stationary varifold? Let's check. If we deform this shape far away from the origin, the total length doesn't change because the lines are straight. The only interesting place is the junction. It turns out that the tension forces balance perfectly, and the total length is stationary. You can prove this rigorously by showing the first variation is zero for any deformation. Similarly, three half-lines meeting at $120^\circ$ angles (a ​​Y-junction​​ or Steiner tree) also form a stationary varifold. This is exactly the configuration soap films form at their junctions, a beautiful piece of physics captured perfectly by the mathematics of stationarity.

• ​​An Icon of Singularity: The Simons Cone:​​ The theory goes much deeper. In dimensions 8 and higher, a truly spectacular singular stationary varifold exists: the ​​Simons cone​​. It's the cone over the product of two spheres, $C = \{(x,y) \in \mathbb{R}^4 \times \mathbb{R}^4 : |x| = |y|\}$. At its apex, the origin, it has a singularity. This object is of monumental importance because, for dimensions $n \ge 8$, it is not only stationary but genuinely ​​area-minimizing​​. This shattered the long-held belief that area-minimizing surfaces must be smooth. It proved that singularities are not just a theoretical curiosity but an unavoidable feature of area minimization in higher dimensions.

Every great theory has a secret weapon, a tool of such profound power that it unlocks deep truths. For stationary varifolds, this weapon is the ​​Monotonicity Formula​​. It looks technical, but its physical meaning is breathtaking.

Let's say you have a stationary varifold. Pick any point $x_0$ in space. Now, draw a small ball of radius $r$ around it and measure the total "area" or "mass" of the varifold inside that ball, let's call it $\mu(B_r(x_0))$. Now, compare this to the area of a perfectly flat disk of the same dimension $m$ and radius $r$, which is $\omega_m r^m$ (where $\omega_m$ is just a constant, the volume of a unit m-ball). This ratio, $\Phi(r) = \frac{\mu(B_r(x_0))}{\omega_m r^m}$, is called the ​​density ratio​​. It tells you how dense the varifold is near $x_0$ compared to a flat plane.

Here is the magic: The monotonicity formula states that for any stationary varifold, this density ratio $\Phi(r)$ is a ​​non-decreasing​​ function of the radius $r$. As you look at larger and larger balls, the density can only stay the same or go up. It can never go down.

Think about what this means. It's a fundamental rule of structure for these minimal shapes. Imagine you're flying over a city at night. If the city lights represented a stationary varifold, the average brightness in your circular field of view could never decrease as you fly higher (increasing your radius). This is a rigid constraint that comes directly from the stationarity condition. The proof involves a clever choice of vector field in the first variation formula and beautifully reveals that the "increase" in density is precisely related to how much the varifold deviates from being a perfect cone.

The monotonicity formula is not just an elegant curiosity; it is the engine that drives a powerful mathematical microscope. It allows us to "zoom in" on a varifold at any point and understand its infinitesimal structure.

What happens if we take a sequence of smaller and smaller radii, $r_j \to 0$, and "blow up" the image of the varifold in each ball $B_{r_j}(x_0)$ so that it fills a ball of a standard size? This process is called a ​​blow-up​​. The monotonicity formula guarantees that the mass of these rescaled varifolds doesn't run off to infinity. And because of this mass bound, a fundamental result called the ​​compactness theorem for varifolds​​ tells us that we can always find a sequence of blow-ups that converges to a well-defined limit varifold.

What does this limit look like? It's always a ​​cone​​—a varifold that is invariant under scaling from the origin. This limit is called a ​​tangent cone​​.

• If you zoom in on a smooth point of a minimal surface, what do you expect to see? A flat plane! And indeed, at any ​​regular point​​ (a point where the varifold is locally a smooth surface), the tangent cone is unique and is simply the multiplicity-one tangent plane to the surface. This confirms our intuition.
• But what if you zoom in on a singular point, like the intersection of two lines? The tangent cone is the intersection of two lines itself. At the tip of the Simons cone, the tangent cone is the Simons cone. The tangent cone is a perfect "snapshot" of the singularity's geometry.

The existence of the density $\Theta^m(\mu, x_0) = \lim_{r \to 0} \frac{\mu(B_r(x_0))}{\omega_m r^m}$ is a direct consequence of the monotonicity formula. This density tells us something beautiful about the tangent cone: it's equal to the multiplicity of the cone! For example, at a smooth point of a single sheet, the density is 1. At a smooth point of two stacked sheets, the density is 2. At a Y-junction formed by three sheets, the density at the junction is 3. This gives us a powerful quantitative tool to study singularities. In fact, a deep result in the theory states that if the tangent cone at a point is a plane of multiplicity one, that point must be a regular (smooth) point!

We've been using the idea that stationary varifolds are "critical points" for area. In calculus, critical points can be local minima, local maxima, or saddle points. The same is true here. Being stationary does not automatically mean a varifold minimizes area, not even locally.

A stationary varifold that is a true local minimum of area is called ​​stable​​. A classic example that illustrates this difference is the catenoid. A soap film stretched between two rings will form a catenoid, but only if the rings are close enough. If you pull the rings too far apart, the catenoid is still a minimal surface (it has zero mean curvature, so its varifold is stationary), but it is no longer the shape of least area. The soap film will snap and reform into two separate flat disks, one on each ring. The area of the two disks is now less than the area of the "overstretched" catenoid. The catenoid has become a saddle point for the area—it's stationary, but unstable.

An even simpler example is a pair of crossing lines. It's stationary, as we saw. But is it area-minimizing? Clearly not! For the boundary consisting of the four points $(\pm 1, 0)$ and $(0, \pm 1)$, the cross shape is a stationary configuration of length 4. However, it is not a minimizer, as another configuration connecting the same points—for instance, two segments joining $(-1,0)$ to $(0,-1)$ and $(1,0)$ to $(0,1)$—has a shorter total length of $2\sqrt{2} \approx 2.828$. This shows the crossing lines are stationary but unstable.

This distinction is crucial. Stationarity is a local, differential condition ($H=0$). Area-minimization is a global, integral condition (its area is less than or equal to any competitor's). The entire theory of stationary varifolds is a grand exploration of the consequences of this simple, beautiful balancing act, and it reveals a world of geometric forms far richer and more complex than we might have ever imagined.

Now that we have acquainted ourselves with the principles of stationary varifolds, you might be asking, "What is this all for?" It is a fair question. The machinery of varifolds, first variations, and monotonicity formulas can seem quite abstract, a playground for mathematicians. But like all great physical theories, the power of this framework is not in its abstraction, but in its ability to give precise answers to deep and often very concrete questions about the world. Its purpose is to solve problems.

The theory of stationary varifolds is, in essence, a vastly generalized language for talking about "minimal surfaces"—the shapes that nature chooses to minimize energy, like soap films spanning a wire loop. It gives us a rigorous way to handle the wild behavior these surfaces can exhibit: they can have singularities, they can branch, they can be composed of multiple sheets. By providing a unified framework, the theory allows us to see connections between problems that, on the surface, look utterly different. We shall see that the same ideas used to understand the junction of three soap films in your kitchen sink can be used to prove that the universe cannot have negative mass. This is the journey we are about to take.

One of the most profound achievements of this theory is its ability to deduce smoothness from very weak, measure-theoretic assumptions. This is a common theme in physics and mathematics: you start with a "weak" law, an integral principle, and you end up with a "strong" conclusion, a smooth object satisfying a differential equation.

Imagine you have a foggy photograph of what you believe is a sheet of paper. You can't see the details, but you can measure its total area and verify that, on average, it looks like a single, flat sheet. Allard's regularity theorem is the magical developing fluid for this photograph. It tells us that if a stationary varifold has a density close to $1$ at some point (meaning it looks like a single sheet in terms of mass) and it is "almost flat" (meaning its tilt-excess is small), then it must be a perfectly smooth, $C^{1,\alpha}$ surface in a neighborhood of that point. The near-constancy of the density ratio, a direct consequence of the monotonicity formula, is the key that unlocks this powerful conclusion. A weak assumption of stationarity blossoms into the strong, geometric reality of a smooth graph.

But what happens when the density is not close to $1$? What if the fog in our photograph is thicker, hiding not one but several sheets of paper, perhaps intersecting or branching? This is where the story gets even more interesting. Allard's theorem, with its reliance on a single-graph representation, can no longer be applied directly. The local structure is too complex. This is the frontier that was conquered by F. J. Almgren's monumental work. He developed a theory of "multi-valued functions" to handle situations where the density is an integer $Q \ge 2$, providing a framework to describe and control the behavior of multiple interacting sheets.

This deep theory has beautiful, tangible consequences. In our familiar three-dimensional space, the theory of stationary 2-varifolds—our soap films—predicts exactly the kinds of singularities we observe. By performing a "blow-up" analysis, where we zoom in on a singular point, we find that the limiting shape, the tangent varifold, must be a stationary cone. The stationarity of the cone in $\mathbb{R}^3$ is equivalent to the stationarity of its "link" on the unit sphere $\mathbb{S}^2$. For a 2-varifold, this link is a 1-varifold, and its stationarity requires it to be a network of geodesics (great circles) that are perfectly balanced at every junction. This is the mathematical reason behind the famous rule that soap films meet in threes along curves at angles of $120^\circ$, or in fours at a central point with tetrahedral symmetry. The abstract theory predicts the concrete, beautiful geometry of physical reality. Other stationary cones, such as multiple planes passing through the origin, are also possible singularities.

It is one thing to describe the properties of a stationary varifold once you have one. But how do you prove these objects exist in the first place? Imagine you have a complex, donut-shaped manifold. Can you always find a minimal surface "cutting it in half"? For decades, this was an incredibly difficult question. The Almgren-Pitts min-max theory provides a breathtakingly general answer.

The core idea is to imagine continuous "sweepouts" of the manifold. Think of a family of surfaces, parameterized by time, that starts as nothing, expands to sweep through the entire space, and then shrinks back to nothing. For each such sweepout, we find the slice with the maximum area. The "width" of the manifold is then defined as the minimum possible value of this maximum area, taken over all possible sweepouts in a given class. It’s a min-max problem, a search for a saddle point on an infinite-dimensional landscape of shapes.

The genius of the theory is to show that this min-max value, the width, is not just a number but is realized as the area of an actual geometric object. The proof involves a "pull-tight" procedure. Starting with a sweepout whose maximal area is very close to the width, one applies a deformation that intelligently flows each slice in a direction that reduces its area, unless the slice is already stationary. This deformation is meticulously constructed to be continuous and to preserve the overall homotopy class of the sweepout. You keep "tightening" the sweepout, lowering its maximal area, until you can't anymore. The family of surfaces you are left with must contain slices that are "arbitrarily close" to being stationary. A final compactness argument then allows you to extract from this process a limiting object: a stationary integral varifold whose mass is precisely the width.

The technical complexity is immense. For instance, the critical set from this process might contain multiple, disjoint stationary varifolds. A further, highly sophisticated combinatorial selection argument is needed to zoom in on a sequence of shrinking annuli where the surfaces are "almost minimizing," guaranteeing that the limit is a single, connected minimal surface. This masterpiece of mathematical construction guarantees the existence of minimal hypersurfaces in any closed Riemannian manifold.

With such a powerful existence theorem in hand, we can attack problems in other fields. In topology, one might ask if a given homology class—a formal way of describing a non-trivial "cycle" in a manifold—can be represented by a minimal surface. Geometric measure theory provides the answer. One can start with any surface in the class, and then seek a mass-minimizing object within the same class. The Federer-Fleming compactness theorem guarantees a minimizer exists as an integral current, which is automatically a stationary varifold. Now, the regularity theory kicks in. For ambient dimensions $n \le 7$, the theory guarantees that this area-minimizing object has no singularities. Further arguments using the maximum principle show it is embedded and has multiplicity one, yielding a smooth, embedded minimal hypersurface representing the original topological class.

The most spectacular application, however, lies in the heart of physics: Einstein's theory of general relativity. The Positive Mass Theorem states that the total mass (or more accurately, the ADM energy) of an isolated gravitational system, like a star or a black hole, cannot be negative, provided its matter content satisfies a reasonable energy condition (nonnegative scalar curvature). This is a fundamental statement about the stability of gravity.

In their celebrated proof, Schoen and Yau used the theory of minimal surfaces. They showed that if the mass were negative, one could construct a complete, stable minimal hypersurface within the space. A stable minimal surface is one that minimizes area up to second order. The key is that the regularity theory for these surfaces is dimension-dependent. Work by Schoen, Simons, and Yau showed that the singular set of a stable minimal hypersurface has a dimension of at most $n-8$, where $n$ is the dimension of the ambient space. Therefore, for physically relevant spacetime dimensions and their spatial slices (e.g., $n=3$, so dimension is $3+1$), or more generally for $n \le 7$, any such stable minimal surface is guaranteed to be smooth. Once a smooth surface is known to exist, one can use a clever integral identity (the stability inequality) that relates the geometry of the surface to the scalar curvature of the surrounding space. The assumption of negative mass leads to a contradiction. The fact that an abstract theory about idealized soap films provides the key to proving a fundamental theorem about gravity is a stunning testament to the unity of science and mathematics.

The theory of stationary varifolds is not static; it is a living field with an expanding set of tools to tackle new problems.

For example, the beautiful monotonicity formula we have discussed is modified when our surface lives not in flat Euclidean space, but in a curved Riemannian manifold. The curvature of the ambient space acts as a kind of friction or assistance to the area growth. The formula acquires an exponential correction factor that depends on the bounds of the sectional curvature. For a stationary varifold in a manifold with sectional curvature bounded by $|\text{sec}| \le K$, the quantity that becomes non-decreasing is no longer simply $\frac{\mu_V(B_r(p))}{r^m}$, but rather $e^{\Lambda r^2} \frac{\mu_V(B_r(p))}{r^m}$ for some constant $\Lambda$ depending on $K$. This shows how the local variational principle of stationarity intimately interacts with the global geometry of its environment.

Another elegant tool is the reflection principle, used to solve free boundary problems. Suppose you have a stationary varifold in a half-space, which meets the boundary plane orthogonally (the "free boundary" condition). This is a difficult problem because of the boundary. The trick is to reflect the entire situation across the boundary plane. The original varifold plus its reflection creates a new, symmetric varifold defined on the whole space. The brilliance of the method is that the free boundary condition on the original varifold is precisely the condition needed to ensure that the new, doubled-up varifold is stationary everywhere, with no boundary. The problem is transformed from a difficult boundary-value problem into a simpler interior one, with the only change being that the density along the reflection plane is now doubled. This kind of symmetry argument is a recurring theme in physics, and here it finds a beautiful and powerful application in geometry.

From the fine-grained analysis of singularities to the grand existence theorems and their profound applications in topology and physics, the theory of stationary varifolds provides a deep and unified lens through which to view the world of shapes. It is a testament to the power of finding the right language to describe a physical idea—in this case, the simple, intuitive notion of area minimization.