Volume Stability for Hyperbolic Manifolds and Applications to General Relativity

Technical Summary: Volume Stability for Hyperbolic Manifolds and Applications to General Relativity

Problem Statement
The paper addresses the quantification of the rigidity property for closed hyperbolic three-manifolds. Specifically, it investigates the stability of the hyperbolic volume bound under scalar curvature constraints. Let $(M, h)$ be a closed hyperbolic three-manifold normalized such that the scalar curvature $R(h) = -6$ (equivalently, $\text{Ric}_h = -2h$). The classical Schoen volume rigidity conjecture states that for any smooth metric $g$ on $M$ satisfying $R(g) \ge -6$, the volume satisfies $\text{Vol}_g(M) \ge \text{Vol}_h(M)$, with equality if and only if $g$ is isometric to $h$ (up to diffeomorphism).

The authors study the stability of this result in the regime of "small data," defined by a small volume deficit. The central question is: If a sequence of smooth metrics $g_i$ satisfies $R(g_i) \ge -6$ and $\text{Vol}_{g_i}(M) - \text{Vol}_h(M) \le \delta_i$ with $\delta_i \to 0$, in what topology does $(M, g_i)$ converge to $(M, h)$? The authors note that the volume functional is too weak to control strong convergence (e.g., $C^k$ or Gromov-Hausdorff) globally, as defects can concentrate on sets of vanishing volume. The goal is to establish a sharp convergence result outside exceptional sets of negligible volume.

Methodology
The proof relies on the normalized Ricci flow with surgery (Hamilton-Perelman) and the monotonicity of Perelman's entropy functional. The argument proceeds in three main stages:

1. Long-Time Evolution and Thick-Thin Decomposition:
   For each initial metric $g_i$, the authors construct a normalized Ricci flow with surgery $g_i(t)$. They utilize the preservation of the scalar curvature lower bound ($R(g_i(t)) + 6 \ge 0$) and the monotonicity of the rescaled Perelman entropy $\bar{\lambda}[g] = \lambda[g] \text{Vol}(g)^{2/3}$.
   By integrating the entropy dissipation and volume deficit estimates, they show that for a sequence of large times $t_i \to \infty$, the metrics $g_i(t_i)$ become "good" in the sense that the soliton defect (the deviation from being a gradient Ricci soliton) and the scalar curvature defect vanish in an $L^2$ sense.
   Using Perelman's long-time existence and thick-thin decomposition theorems, they demonstrate that on the "thick" part of the manifold, the flow converges smoothly to a finite-volume hyperbolic metric. Since the underlying topology is fixed as the hyperbolic manifold $M$, Mostow rigidity implies this limit is isometric to $h$. This establishes $C^0$ convergence on large domains for the time-evolved metrics, modulo sets of vanishing volume.

2. Pullback to the Initial Slice via Gradient Flow Estimates:
   The primary difficulty lies in transferring the convergence from the large time $t_i$ back to the initial time $t=0$. The authors cannot simply integrate the metric evolution equation because standard $L^2$ estimates on the deformation tensor $\partial_t g$ are insufficient to control the accumulated distortion over long times (a naive Cauchy-Schwarz estimate would introduce a factor of $t_i$).
   To overcome this, they employ a Lojasiewicz-Simon inequality near the hyperbolic metric. This inequality relates the entropy deficit to the $L^2$ norm of the soliton defect. By combining this with the gradient flow structure of the normalized Ricci flow (modulo diffeomorphisms and scalings), they derive an $L^1_t L^2_x$ path-length estimate for the deformation tensor.
   Specifically, they show that the accumulated metric distortion along "surviving worldlines" (trajectories that do not pass through surgery regions or thin parts) is small in an $L^1$ sense. This allows them to construct a bilipschitz comparison between the initial metric and the evolved metric on a "good" set, provided the set of points with large accumulated distortion is removed.

3. Control of Exceptional Sets:
   The authors carefully track the volume of "bad" sets, which include:

   • Regions where the scalar curvature defect accumulates significantly.
   • Regions where the accumulated metric distortion exceeds a threshold.
   • Surgery regions and thin parts of the thick-thin decomposition.
     Using Chebyshev-type inequalities and the integral bounds derived from the volume deficit and entropy monotonicity, they prove that the volume of these exceptional sets tends to zero as $i \to \infty$.

Key Contributions and Results

• Theorem 1.1 (Main Volume Stability Theorem):
  For a sequence of smooth metrics $g_i$ on a closed hyperbolic 3-manifold $M$ with $R(g_i) \ge -6$ and $\text{Vol}_{g_i}(M) \to \text{Vol}_h(M)$, there exist subsets $Z_i \subset M$ with $\text{Vol}_{g_i}(Z_i) \to 0$ and smooth domains $K_i \subset M$ with $\text{Vol}_h(M \setminus K_i) \to 0$, such that there are diffeomorphisms $\psi_i: K_i \to M \setminus Z_i$ satisfying:
  $$\|\psi_i^* g_i - h\|_{C^0(K_i, h)} \to 0$$
  In tensorial terms, $(1-\epsilon_i)\psi_i^* h \le g_i \le (1+\epsilon_i)\psi_i^* h$ on $K_i$. The convergence is $C^0$ (tensorial), not necessarily higher order, and holds outside regions of vanishing volume.

• Theorem 1.2 (Reduced Hamiltonian Stability):
  The authors apply the volume stability result to the Fischer-Moncrief reduced Hamiltonian in General Relativity. For vacuum CMC (Constant Mean Curvature) initial data $(g_i, k_i)$ on a hyperbolic 3-manifold, the reduced Hamiltonian is $H(g, k) = (-\tau)^3 \text{Vol}_g(M)$, where $\tau = \text{tr}_g k < 0$.
  They prove that if a sequence of such data has reduced Hamiltonians converging to the ground state value (corresponding to the Lorentz cone over the hyperbolic metric), then the normalized spatial metrics converge to the hyperbolic metric in $C^0$ outside sets of vanishing volume. This provides a rigorous formulation of the asymptotic picture where the Lorentz cone is the ground state of the reduced Einstein dynamics.

Significance and Claims
The paper claims to provide a sharp volume-stability theorem for hyperbolic 3-manifolds, quantifying the rigidity of the hyperbolic volume bound. The significance lies in:

1. Topological Convergence: It establishes that near-minimizers of the hyperbolic volume (under scalar curvature bounds) are geometrically close to the hyperbolic metric in the $C^0$ topology, except on negligible sets. This resolves the question of the topology of convergence for such sequences.
2. General Relativity Application: It offers a rigorous mathematical foundation for the Fischer-Moncrief asymptotic picture. It confirms that near-minimizing vacuum CMC data in the hyperbolic topological class converge to the Lorentz-cone geometry (the ground state) in a volume-dominated sense.
3. Methodological Rigor: The work bridges the gap between the global well-posedness of Ricci flow with surgery and the stability of geometric invariants. It explicitly handles the difficulty of pulling back convergence from long-time slices to the initial data using Lojasiewicz-Simon inequalities and path-length estimates, avoiding the need for strong curvature bounds that are unavailable in the stability regime.

The authors explicitly state that their result is a stability statement for near-minimizers, not a global convergence theorem for arbitrary large data. They note that for large data, the existence of global CMC foliations is not guaranteed due to potential curvature focusing and trapped surface formation, consistent with the Klainerman-Rodnianski continuation criteria. Thus, the theorem characterizes the behavior of the system near the ground state of the reduced Hamiltonian.