Flat extensions of principal connections and the Chern-Simons $3$-form

This paper introduces the concept of flat extensions for principal connections, linking their existence on trivial bundles over closed 3-manifolds to the vanishing of Chern-Simons invariants, and applies this framework to derive global obstructions for conformal, Lorentzian, and equiaffine immersions of 3-manifolds into Euclidean 4-space.

The Gist

The Big Picture: The "Flat Map" Detective Story

Imagine you are a cartographer trying to draw a map of a strange, curved world (like a mountain or a sphere) onto a perfectly flat sheet of paper (like a table).

In mathematics and physics, there is a famous problem: Can you flatten a curved surface without tearing, stretching, or crumpling it?

• If you try to flatten a basketball onto a table, you have to cut it or stretch it. It's impossible to do perfectly.
• However, sometimes, if the curvature of the object is "just right," you can fit it into a higher-dimensional flat space without distortion.

This paper is about a new mathematical tool (a "detector") that tells us when a curved 3D world can be perfectly embedded (immersed) into a flat 4D space.

The Characters in Our Story

1. The Curved World (The 3-Manifold): Think of this as a 3D universe. It could be a shape like a sphere, a donut, or a twisted knot. It has its own internal geometry (curvature).
2. The Flat World (Euclidean 4-Space): This is a perfect, flat, 4-dimensional room. No curves, no bumps.
3. The Connection (The "Compass"): To navigate the curved world, we need a rulebook for how to move in a straight line. In math, this is called a connection. It's like a compass that tells you how to walk without turning, even if the ground is curved.
4. The Chern-Simons Invariant (The "Curvature Score"): This is a special number calculated from the compass rulebook.
   • If the score is zero (or a whole number), it suggests the world might be "flat enough" to fit into the 4D room.
   • If the score is a weird fraction (like 0.5), it's a red flag. It means the world is too twisted to fit perfectly.

The New Idea: "Flat Extensions"

The authors introduce a clever trick called a Flat Extension.

Imagine you have a complex, curved puzzle piece (your 3D world). You want to know if it fits into a giant, flat puzzle box (4D space).

• The Old Way: You try to force the piece in and see if it breaks.
• The New Way (Flat Extension): You ask, "Can I build a bigger version of this puzzle piece that is perfectly flat, but still contains my original piece inside it?"

If you can build this "bigger, flat version" (mathematically called a flat extension), then your original piece is guaranteed to fit into the 4D room.

The Analogy of the Shadow:
Think of the 3D world as a shadow cast by a 4D object.

• If the shadow is cast by a perfectly flat 4D object, the shadow has a specific "signature" (the Chern-Simons score).
• The paper proves that if your shadow (the 3D world) has a "Flat Extension" (meaning it comes from a flat 4D object), then its signature must be zero (or an integer).
• If the signature is not zero, then the shadow cannot come from a flat 4D object. Therefore, you cannot embed your 3D world into 4D space without distortion.

The "Blindness" Trick

One of the coolest parts of the paper is a mathematical property the authors call "Partial Blindness."

Imagine you are wearing special glasses that only let you see the "curved" parts of a shape, but they make the "flat" parts invisible.

• The authors found that the Chern-Simons "score" is blind to certain parts of the geometry.
• If you have a shape that is "flat" in a specific higher-dimensional sense, the score ignores the extra dimensions and only looks at the core 3D part.
• This allows them to calculate the score easily. If the "flat extension" exists, the score must be zero.

Real-World Applications (Why should we care?)

The paper uses this theory to solve three specific puzzles:

1. The Riemannian Case (The Classic Problem):

   • Question: Can a curved 3D space (like the surface of a planet, but in 3D) be placed inside 4D Euclidean space?
   • Result: They recovered a famous old result: If the "Chern-Simons score" isn't a whole number, the answer is NO. You can't flatten it.

2. The Lorentzian Case (Time and Space):

   • Question: What about universes where time is different from space (like in Einstein's relativity)? Can they fit into a 4D "spacetime" room?
   • Result: They found that if you try to fit a specific type of curved spacetime into a 4D room with a "negative" direction (like a time machine), the score must be zero. If it's not, it's impossible.

3. The Equiaffine Case (Volume and Shape):

   • Question: Can a 3D shape be placed in 4D space such that its "volume" is preserved perfectly, even if the shape is twisted?
   • Result: They proved that for a specific shape called Real Projective Space ($RP^3$), the answer is NO. Even though it looks nice and symmetric, its "Curvature Score" is 0.5. Because 0.5 is not a whole number, it cannot be perfectly embedded in 4D space without breaking the volume rules.

The Takeaway

This paper is like finding a new metal detector for geometry.

• Before, mathematicians had to do heavy lifting to check if a shape could fit into a higher dimension.
• Now, they have a simple "score" (the Chern-Simons invariant).
• If the score is "wrong" (not an integer or not zero), the shape is too twisted to fit into the flat 4D world.
• If the score is "right," it means the shape has a "Flat Extension," and it can fit.

It's a beautiful bridge between abstract algebra (Lie groups) and the physical reality of how shapes can exist in our universe.

Technical Summary

1. Problem Statement

The paper addresses the geometric interpretation and obstructions related to Chern–Simons (CS) invariants on 3-manifolds. While CS invariants are well-understood in Riemannian geometry (specifically regarding isometric immersions into Euclidean space) and CR-geometry, their meaning in broader differential geometric contexts (such as Lorentzian geometry or equiaffine structures) has received less attention.

The central problem is to establish a general criterion linking the vanishing or integrality of the Chern–Simons invariant of a principal connection $\theta$ to the existence of a specific geometric structure: a flat extension of that connection. The authors aim to unify various known obstruction results (like the original Chern–Simons obstruction for $E^4$ immersions) and derive new ones for Lorentzian and equiaffine geometries.

2. Methodology and Framework

A. Flat Extensions of Connections

The authors introduce the concept of a flat extension. Let $G$ be a Lie subgroup of a larger Lie group $\tilde{G}$, with Lie algebras $\mathfrak{g} \subset \tilde{\mathfrak{g}}$. Assume $\tilde{\mathfrak{g}}$ is equipped with an $\text{Ad}$-invariant non-degenerate symmetric bilinear form $\langle \cdot, \cdot \rangle$.

• Definition: A connection $\theta$ on a principal $G$-bundle $P \to M$ admits a flat extension of type $(\tilde{G}, G)$ if there exists a bundle homomorphism $F: P \to \tilde{G}$ (viewing $\tilde{G}$ as a principal $G$-bundle over $\tilde{G}/G$) such that:
  $$\theta = F^*(\mu_{\tilde{G}}^\top)$$
  where $\mu_{\tilde{G}}$ is the Maurer–Cartan form of $\tilde{G}$, and $\mu_{\tilde{G}}^\top$ is its projection onto $\mathfrak{g}$.
• Intuition: This implies $\theta$ arises as the pullback of a component of a flat (Maurer–Cartan) form on a larger group.

B. Algebraic Decomposition and "Partial Blindness"

A core technical tool is the decomposition of the $\tilde{\mathfrak{g}}$-valued Maurer–Cartan form $\psi = \psi^\top + \psi^\perp$, where $\psi^\top \in \mathfrak{g}$ and $\psi^\perp \in \mathfrak{g}^\perp$.

• Theorem 5.1 (Partial Blindness): If $\psi$ satisfies the Maurer–Cartan equation ($d\psi + \frac{1}{2}[\psi, \psi] = 0$) and $[\psi^\perp, \psi^\perp] \in \Omega^2(\mathfrak{g})$, then:
  $$CS(\psi) = CS(\psi^\top)$$
  This means the Chern–Simons 3-form of the flat form $\psi$ depends only on its $\mathfrak{g}$-component. Consequently, if $\theta$ admits a flat extension, $CS(\theta)$ is the pullback of $CS(\mu_{\tilde{G}})$.

C. Cohomological Normalization

The authors carefully normalize the bilinear form $\langle \cdot, \cdot \rangle$ (typically a multiple of the trace form or Killing form) to ensure that the cohomology class $[CS(\mu_{\tilde{G}})]$ lies in $H^3(\tilde{G}, \mathbb{Z})$. This is crucial for determining whether the resulting invariant is integer-valued or vanishes.

3. Key Contributions and Results

A. General Theorem on Obstructions

Theorem 6.2 is the main result. It states that if a connection $\theta$ on a closed oriented 3-manifold $M$ admits a flat extension of type $(\tilde{G}, G)$:

1. If $[CS(\mu_{\tilde{G}})] \in H^3(\tilde{G}, \mathbb{Z})$, then the Chern–Simons invariant $\int_M \sigma^* CS(\theta)$ is an integer for any global section $\sigma$.
2. If $CS(\mu_{\tilde{G}})$ is exact (i.e., the class is zero), then the invariant vanishes.

This provides a unified mechanism to generate obstructions: the existence of a specific immersion implies the existence of a flat extension, which in turn forces the CS invariant to be an integer or zero.

B. Application 1: Riemannian Geometry (Recovery of Chern–Simons)

• Setup: $G = SO(3)$, $\tilde{G} = SO(4)$. The pair $(\mathfrak{so}(4), \mathfrak{so}(3))$ is a symmetric pair.
• Result: An isometric immersion $M \to \mathbb{R}^4$ induces a flat extension via the Gauss map.
• Conclusion: The Chern–Simons invariant of a Riemannian 3-manifold must be an integer. If the normalization is chosen such that the invariant is in $\mathbb{R}/\mathbb{Z}$, the existence of an isometric immersion implies the invariant vanishes. This recovers the classical result by Chern and Simons.

C. Application 2: Lorentzian Geometry

• Setup: $G = SO_0(2, 1)$, with two possible $\tilde{G}$: $SO_0(3, 1)$ (Minkowski space $\mathbb{R}^{3,1}$) and $SO_0(2, 2)$ (split signature $\mathbb{R}^{2,2}$).
• Distinction:
  • For $\tilde{G} = SO_0(2, 2)$, $H^3(\tilde{G}, \mathbb{R}) = 0$. Thus, an isometric immersion into $\mathbb{R}^{2,2}$ implies the CS invariant vanishes.
  • For $\tilde{G} = SO_0(3, 1)$, $H^3(\tilde{G}, \mathbb{R}) \cong \mathbb{R}$ (non-trivial). An isometric immersion into $\mathbb{R}^{3,1}$ implies the CS invariant is an integer.
• Significance: This distinguishes the topological constraints imposed by immersions into different Lorentzian signatures.

D. Application 3: Equiaffine Geometry

• Setup: $G = SL(3, \mathbb{R})$, $\tilde{G} = SL(4, \mathbb{R})$. This concerns 3-manifolds with a torsion-free connection preserving a volume form.
• Result: An equiaffine immersion into $\mathbb{R}^4$ induces a flat extension.
• Conclusion: The Chern–Simons invariant associated with the equiaffine connection must vanish.
• Example: The authors prove that $\mathbb{RP}^3$ with its standard Levi-Civita connection and volume form admits no global equiaffine immersion into $\mathbb{R}^4$, as its CS invariant is $1/2 \notin \mathbb{Z}$ (and specifically non-zero).

4. Significance and Impact

1. Unification: The paper provides a single theoretical framework (flat extensions) that explains and generalizes disparate results in Riemannian, Lorentzian, and affine geometry.
2. New Obstructions: It establishes new topological obstructions for the existence of immersions in Lorentzian and equiaffine settings, extending the classical work of Chern and Simons beyond the Riemannian case.
3. Geometric Interpretation: The work clarifies the geometric meaning of the "extra" components in the Maurer–Cartan form of the extended group. In the Riemannian case, the orthogonal component corresponds to the second fundamental form (shape operator). In the equiaffine case, these components relate to the shape operator and second fundamental form of the equiaffine immersion, which are shown to be independent but constrained by volume compatibility.
4. Computational Rigor: The authors provide explicit normalizations for the bilinear forms (Appendix A) to ensure the integrality of the invariants, a necessary step for applying these results as strict topological obstructions.

In summary, the paper demonstrates that the existence of certain geometric immersions is equivalent to the existence of a flat extension of the underlying connection, which forces the Chern–Simons invariant to satisfy specific arithmetic conditions (vanishing or integrality), thereby providing powerful tools to rule out the existence of such immersions on specific manifolds.