Threshold Rigidity for Separated Sub-Static Splittings
This is the first milestone of the project. After requiring AI explore more results on previous article, it has successfully made a PhD level, publishable result.
Threshold Rigidity for Separated Sub-Static Splittings
Shengrong Wu
National University of Singapore
Automatic Pipeline with ChatGPT
Abstract
We study the separated sub-static ansatz \[ g=\beta(y)^2dt^2+h(y), V(t,y)=A(t)\beta(y), \] on \(\mathbb{R}\times N^{n-1}\) under the inequality \[ S(V,g):=V\operatorname{Ric}_g+(\Delta_gV)g-\nabla^2_gV\ge 0. \] For this ansatz, we show that the full tensor inequality is equivalent to the one-dimensional condition \[ A''(t)+\kappa_* A(t)\ge 0 \text{for all } t\in \mathbb{R}, \] where \[ \kappa_*=\inf_{y\in N}\lambda_{\min}^h\bigl(\mathcal{T}_h(\beta)(y)\bigr), \mathcal{T}_h(\beta)=\beta^2\operatorname{Ric}_h-2\beta\nabla^2_h\beta+(\beta\Delta_h\beta+|\nabla_h\beta|_h^2)h. \] This yields an exact criterion for bounded positive nonconstant line factors: such an \(A\) exists if and only if \(\kappa_*>0\).
We derive sharp local corollaries. Strict negativity of the traced slice tensor or of a single tensor direction implies nonexistence of bounded positive line factors, while the borderline case forces constancy. On compact slices, the existence of a bounded positive nonconstant line factor is equivalent to pointwise positive-definiteness of \(\mathcal{T}_h(\beta)\). In dimension three we obtain a critical-point determinant criterion: at every critical point of \(\beta\), nonconstant bounded line factors require \[ \beta K_h>|\lambda_1-\lambda_2|, \] where \(K_h\) is the Gauss curvature and \(\lambda_1,\lambda_2\) are the eigenvalues of \(\nabla^2_h\beta\).
We also prove a conformal reformulation of \(\mathcal{T}_h(\beta)\), and use it to derive a boundary-control theorem in a geodesic collar. Finally, we record a corrected formal collar expansion clarifying what additional pointwise input would be needed to connect the threshold invariant directly to Wang’s Neumann-data framework.
MSC 2020: 53C21, 53C25, 58J60. Keywords: sub-static manifold, splitting theorem, threshold invariant, Bakry-Emery tensor, warped product, rigidity.
Introduction
A Riemannian triple \((M^n,g,V)\) with \(V>0\) is called sub-static if \[ S(V,g):=V\operatorname{Ric}_g+(\Delta_gV)g-\nabla^2_gV\ge 0. \] This condition appears naturally in the static Einstein setting and in recent comparison, rigidity, and splitting results for weighted and conformally compact geometries [WangRiccati, BF, CFMR, GSW]. A central structural question in the sub-static splitting program is whether the presence of a line forces the potential to be independent of the line direction.
In Wang’s Riccati approach [WangRiccati], the unresolved part of the splitting problem reduces to the separated ansatz \[ g=\beta(y)^2dt^2+h(y), V(t,y)=A(t)\beta(y), \tag{ansatz} \] on \(M=\mathbb{R}\times N^{n-1}\), with \(\beta>0\) on the slice and \(A>0\) on \(\mathbb{R}\). The remaining question is whether the sub-static inequality forces \(A\) to be constant. The purpose of this paper is to give an exact criterion for that question inside the separated ansatz.
The key observation is that the slice block of \(S(V,g)\) factors into a purely geometric tensor and a one-dimensional term. This leads to the threshold tensor \[ \mathcal{T}_h(\beta):=\beta^2\operatorname{Ric}_h-2\beta\nabla^2_h\beta+(\beta\Delta_h\beta+|\nabla_h\beta|_h^2)h \] and the threshold invariant \[ \kappa_*:=\inf_{y\in N}\lambda_{\min}^h\bigl(\mathcal{T}_h(\beta)(y)\bigr). \] Our main result states that, for the separated ansatz, the condition \(S\ge 0\) is equivalent to the scalar ODE inequality \[ A''+\kappa_* A\ge 0. \] The existence problem for bounded positive nonconstant line factors is therefore completely determined by the sign of \(\kappa_*\).
This reformulation has several immediate consequences. It upgrades the local trace and tensor obstructions to sharp nonexistence/rigidity statements, yields a compact-slice criterion in terms of pointwise positive-definiteness of \(\mathcal{T}_h(\beta)\), and in dimension three produces a clean determinant test at critical points of \(\beta\). We also record a conformal identity for \(\mathcal{T}_h(\beta)\) and a boundary-control theorem in a geodesic collar, followed by a corrected formal asymptotic lemma relevant to Wang’s boundary expansions.
The strongest clean contribution of the paper is the exact threshold criterion together with its corollaries. The later conformal and boundary sections are intended as a geometric interface with the broader splitting program, rather than as a full solution to the general conformally compact case.
Main results
The paper is built around one exact structural theorem and several sharp corollaries. The structural theorem is the Exact Threshold Criterion (Theorem 3.2 below): for the separated ansatz, the tensor inequality \(S(V,g)\ge 0\) is equivalent to the scalar inequality \(A''+\kappa_* A\ge 0\). This result is stronger than a rigidity statement alone because it simultaneously gives an existence theorem in the positive-threshold regime.
The local consequences are recorded in the Sharp Trace Obstruction and Sharp Tensor Obstruction propositions. They sharpen the usual convexity argument by distinguishing the strict and borderline cases: strict negativity of a local obstruction yields nonexistence of bounded positive line factors, while equality yields constancy. In the compact case, the Compact Slices Corollary turns the global existence problem into a pointwise positive-definiteness test for \(\mathcal{T}_h(\beta)\).
The three-dimensional results show that the threshold theory has concrete geometric content beyond the scalar ODE reduction. At critical points of \(\beta\), positivity of the slice tensor becomes an explicit inequality relating Gauss curvature and Hessian anisotropy. The conformal and collar results then reinterpret the same threshold through the compactified metric and show how boundary geometry can force rigidity.
Context and comparison with the literature
The present paper should be read as a contribution to the separated-ansatz step in the sub-static splitting program, rather than as a full new splitting theorem for arbitrary sub-static manifolds. In Wang’s line-based approach [WangRiccati], the unresolved issue is the constancy of the line factor once the separated form (ansatz) has been obtained. Our exact threshold criterion addresses precisely that point.
This perspective is complementary to other approaches in the literature. Colombo, Freitas, Mari, and Rigoli [CFMR] obtain rigidity statements from stable minimal hypersurfaces, while Borghini and Fogagnolo [BF] develop weighted comparison geometry under the sub-static condition. The threshold invariant extracted here is different in flavor: it is neither purely variational nor purely comparison-theoretic, but instead an eigenvalue bound arising directly from the slice block of \(S(V,g)\).
The conformal identity in the Conformal Reformulation section suggests that these viewpoints are not unrelated. After conformal compactification, the threshold tensor becomes a Bakry–'Emery-type curvature quantity. This does not by itself solve the compactified problem, but it indicates that the threshold invariant is geometrically natural and not an artifact of the separated coordinates.
Organization
Section 2 derives the slice decomposition. Section 3 proves the exact threshold criterion and the bounded-solution criterion. Section 4 explains how the threshold interacts with local tests and records model families. Section 5 treats scaling and one-dimensional slices. Section 6 proves the sharp local obstructions. Section 7 treats the surface (dimension-three) case. Section 8 contains the conformal reformulation and the collar theorem. Section 9 records the corrected formal expansion relevant to Wang’s boundary data. The appendices collect the warped-product formulas, the ODE lemmas, and expanded model computations.
The Separated Ansatz and Slice Decomposition
Throughout, \((N^m,h)\) denotes a connected smooth Riemannian manifold of dimension \(m=n-1\ge 1\). We write \(R_h\), \(\operatorname{Ric}_h\), \(\Delta_h\), \(\nabla_h\), and \(\nabla^2_h\) for the scalar curvature, Ricci tensor, Laplacian, gradient, and Hessian of \((N,h)\). We fix \(\beta\in C^\infty(N)\) with \(\beta>0\), and \(A\in C^\infty(\mathbb{R})\) with \(A>0\). All tensors are smooth.
The two slice tensors used throughout are \[ \begin{aligned} \Theta_h(\beta) &:= \beta\operatorname{Ric}_h-2\nabla^2_h\beta+ \\ \bigl(\Delta_h\beta+\beta^{-1}|\nabla_h\beta|_h^2\bigr)h, \\ \mathcal{T}_h(\beta) &:= \beta\Theta_h(\beta) \\ = \beta^2\operatorname{Ric}_h-2\beta\nabla^2_h\beta+ \bigl(\beta\Delta_h\beta+|\nabla_h\beta|_h^2\bigr)h. \end{aligned} \] The \(h\)-trace of \(\Theta_h(\beta)\) is \[ \mathcal{Q}_h(\beta)=\beta R_h+(n-3)\Delta_h\beta+(n-1)\beta^{-1}|\nabla_h\beta|_h^2. \] The starting point is the following decomposition.
Lemma 2.1 (Slice decomposition).
Under the separated ansatz (ansatz), \[ S(\partial_t,\partial_t)=0, \qquad S(\partial_i,\partial_t)=0, \] and on the slice directions one has \[ S(\partial_i,\partial_j)=A\Theta_h(\beta)_{ij}+\beta^{-1}A''h_{ij}. \tag{Sij} \] In particular, \[ \operatorname{tr}_h\bigl(S|_{TN}\bigr)=A\mathcal{Q}_h(\beta)+(n-1)\beta^{-1}A''. \tag{traceid} \] Proof.
For \(g=\beta^2dt^2+h\), the nonzero Christoffel symbols involving the \(t\)-direction are \[ \Gamma^t_{it}=\beta^{-1}\partial_i\beta, \qquad \Gamma^i_{tt}=-\beta\nabla_h^i\beta, \qquad \Gamma^k_{ij}=\Gamma^k_{ij}(h). \] With \(V=A(t)\beta(y)\), the Hessian components are \[ \begin{aligned} \nabla^2_gV(\partial_t,\partial_t) &=A''\beta+A\beta|\nabla_h\beta|_h^2, \\ \nabla^2_gV(\partial_i,\partial_t) &=0, \\ \nabla^2_gV(\partial_i,\partial_j) &=A\nabla^2_h\beta_{ij}. \end{aligned} \] Hence \[ \Delta_gV=\beta^{-1}A''+A\Delta_h\beta+A\beta^{-1}|\nabla_h\beta|_h^2. \] The Ricci tensor of \(g=\beta^2dt^2+h\) satisfies \[ \begin{aligned} \operatorname{Ric}_g(\partial_t,\partial_t)&=-\beta\Delta_h\beta, \\ \operatorname{Ric}_g(\partial_i,\partial_t)&=0, \\ \operatorname{Ric}_g(\partial_i,\partial_j)&=\operatorname{Ric}_h{}_{ij}-\beta^{-1}\nabla^2_h\beta_{ij}. \end{aligned} \] Substituting these identities into \(S(V,g)=V\operatorname{Ric}_g+(\Delta_gV)g-\nabla^2_gV\) gives \(S_{tt}=0\), \(S_{it}=0\), and \[ \begin{aligned} S_{ij} &=A\beta\bigl(\operatorname{Ric}_h{}_{ij}-\beta^{-1}\nabla^2_h\beta_{ij}\bigr) \\ +\bigl(\beta^{-1}A''+A\Delta_h\beta+A\beta^{-1}|\nabla_h\beta|_h^2\bigr)h_{ij} -A\nabla^2_h\beta_{ij} \\ &=A\bigl(\beta\operatorname{Ric}_h{}_{ij}-2\nabla^2_h\beta_{ij} \\ +(\Delta_h\beta+\beta^{-1}|\nabla_h\beta|_h^2)h_{ij}\bigr) +\beta^{-1}A''h_{ij}, \end{aligned} \] which is (Sij). Taking the \(h\)-trace yields (traceid). \(\square\)
Remark.
Lemma 2.1 shows that all sub-static information for the separated ansatz lives in the slice block. In particular, the constancy problem for the line factor \(A\) is entirely encoded by the slice geometry \((N,h,\beta)\).
The Exact Threshold Criterion
We begin by noting that finiteness of the threshold invariant is automatic whenever a separated sub-static solution exists.
Lemma 3.1 (Automatic finiteness of \(\kappa_*\)).
Assume the separated ansatz (ansatz) with \(A>0\) and \(S(V,g)\ge 0\). Then \[ \kappa_*=\inf_{y\in N}\lambda_{\min}^h\bigl(\mathcal{T}_h(\beta)(y)\bigr)>-\infty. \] Proof.
Multiplying (Sij) by \(\beta>0\) gives \[ A(t)\mathcal{T}_h(\beta)(y)+A''(t)h(y)\ge 0 \qquad\text{for all } t\in\mathbb{R},\ y\in N. \tag{T-inequality} \] Fix \(t_0\in\mathbb{R}\). For every \(y\in N\) and every \(h\)-unit vector \(X\in T_yN\), \[ 0\le A(t_0)\mathcal{T}_h(\beta)(X,X)+A''(t_0). \] Since \(A(t_0)>0\), \[ \mathcal{T}_h(\beta)(X,X)\ge -\frac{A''(t_0)}{A(t_0)}. \] Taking the infimum over all \(y\) and all \(h\)-unit vectors \(X\) yields \[ \kappa_*\ge -\frac{A''(t_0)}{A(t_0)}>-\infty. \qquad\square \] Theorem 3.2 (Exact threshold criterion).
Under the separated ansatz (ansatz) with \(A>0\), the sub-static condition \(S(V,g)\ge 0\) is equivalent to \[ A''(t)+\kappa_* A(t)\ge 0 \qquad\text{for all } t\in\mathbb{R}. \tag{ode} \] Proof.
By Lemma 2.1, \(S\ge 0\) is equivalent to \[ A(t)\mathcal{T}_h(\beta)(y)+A''(t)h(y)\ge 0 \quad\text{as bilinear forms on }T_yN,\quad \forall t\in\mathbb{R},\ y\in N. \] Assume first that (ode) holds. For every \(y\in N\) and every \(h\)-unit vector \(X\in T_yN\), \[ \mathcal{T}_h(\beta)(X,X)\ge \lambda_{\min}^h(\mathcal{T}_h(\beta)(y))\ge \kappa_*. \] Hence \[ A(t)\mathcal{T}_h(\beta)(X,X)+A''(t)\ge A(t)\kappa_*+A''(t)\ge 0, \] which proves (T-inequality), and therefore \(S\ge 0\).
Conversely, assume (T-inequality). Fix \(t\in\mathbb{R}\). Then for every \(y\in N\) and every \(h\)-unit vector \(X\in T_yN\), \[ 0\le A(t)\mathcal{T}_h(\beta)(X,X)+A''(t). \] Taking the infimum over \(y\) and \(X\) gives \[ 0\le A(t)\kappa_* + A''(t), \] which is (ode). \(\square\)
Corollary 3.3 (Bounded-solution criterion).
Assume the separated ansatz (ansatz) with \(A>0\) bounded on \(\mathbb{R}\). Then a bounded positive nonconstant line factor exists if and only if \(\kappa_*>0\).
Proof.
If \(\kappa_*>0\), choose \(0<\omega^2<\kappa_*\) and \(0<\varepsilon<1\), and set \[ A(t)=1+\varepsilon\sin(\omega t). \] Then \(A\) is bounded, positive, and nonconstant, and \[ A''(t)+\kappa_* A(t) =\kappa_*+\varepsilon(\kappa_*-\omega^2)\sin(\omega t) \ge \kappa_*-\varepsilon(\kappa_*-\omega^2)>0. \] So Theorem 3.2 gives \(S\ge 0\).
Conversely, suppose \(A\) is bounded, positive, and nonconstant with \(S\ge 0\). By Theorem 3.2, \(A''+\kappa_* A\ge 0\). If \(\kappa_*<0\), write \(\kappa_*=-\mu\) with \(\mu>0\). Then \[ A''(t)\ge \mu A(t)>0 \qquad\text{for all } t. \] Thus \(A'\) is strictly increasing. Fix \(t_0\in\mathbb{R}\). If \(A'(t_0)\ge 0\), then \(A'(t)>0\) for all \(t>t_0\), so \(A(t)\to+\infty\) as \(t\to+\infty\). If \(A'(t_0)<0\), then \(A'(t)\le A'(t_0)<0\) for every \(t<t_0\), hence \[ A(t)\ge A(t_0)-A'(t_0)(t_0-t)\to+\infty \quad(t\to-\infty). \] So \(A\) cannot be bounded. If \(\kappa_*=0\), then \(A''\ge 0\), so \(A\) is convex; a convex function bounded above on all of \(\mathbb{R}\) is constant. Both alternatives contradict the assumption that \(A\) is bounded and nonconstant. Therefore \(\kappa_*>0\). \(\square\)
Proposition 3.4 (Uniform spectral-gap criterion).
For a fixed slice \((N,h,\beta)\), the following are equivalent.
There exists a bounded positive nonconstant \(A:\mathbb{R}\to(0,\infty)\) for which the separated ansatz is sub-static.
\(\kappa_*>0\).
There exists \(\kappa>0\) such that \(\mathcal{T}_h(\beta)\ge \kappa h\) pointwise on \(N\).
If \(N\) is compact, these are further equivalent to:
- \(\mathcal{T}_h(\beta)\) is positive-definite at every point of \(N\).
Proof.
The equivalence of (i) and (ii) is Corollary 3.3. The equivalence of (ii) and (iii) is immediate from the definition of \(\kappa_*\) as the infimum of the smallest \(h\)-eigenvalue of \(\mathcal{T}_h(\beta)\).
Assume now that \(N\) is compact. If (iv) holds, the continuous function \(y\mapsto \lambda_{\min}^h(\mathcal{T}_h(\beta)(y))\) attains a positive minimum, so \(\kappa_*>0\). Conversely, if \(\kappa_*>0\), then \(\mathcal{T}_h(\beta)\ge \kappa_* h\), hence \(\mathcal{T}_h(\beta)\) is positive-definite everywhere. \(\square\)
Example 3.5 (Three basic regimes).
The sign possibilities for \(\kappa_*\) are all realized.
If \(n\ge 3\), \(N=S^{n-1}\) with the round metric of sectional curvature \(1\), and \(\beta\equiv 1\), then \(\mathcal{T}_h(\beta)=(n-2)h\), so \(\kappa_*=n-2>0\).
If \(N=T^{n-1}\) is flat and \(\beta\equiv c>0\), then \(\mathcal{T}_h(\beta)=0\), so \(\kappa_*=0\).
If \(N=\mathbb{R}^m\) is Euclidean and \(\beta(y)=1+(y^1)^2\), then \[ \mathcal{T}_h(\beta) =\bigl(-2+2(y^1)^2\bigr)dy^1\otimes dy^1 +\bigl(2+6(y^1)^2\bigr)\sum_{\alpha=2}^m dy^\alpha\otimes dy^\alpha, \] and hence \(\kappa_*=-2<0\).
Qualitative Consequences and Model Families
The exact threshold criterion has a simple conceptual consequence: the separated theory is controlled by the sign of a single spectral quantity. This section spells out how the threshold interacts with the more local trace and tensor tests and explains why \(\kappa_*\) is the correct global invariant.
Proposition 4.1 (Hierarchy of sign conditions).
Assume the separated ansatz with \(A>0\) bounded and \(S\ge 0\).
If \(\kappa_*>0\), then \(\Theta_h(\beta)\) is positive-definite at every point of \(N\), and \(\mathcal{Q}_h(\beta)>0\) everywhere.
If there exist \(y_0\in N\) and \(0\neq X\in T_{y_0}N\) such that \(\Theta_h(\beta)(X,X)\le 0\), then \(\kappa_*\le 0\).
If \(N\) is compact, then \(\kappa_*>0\) if and only if \(\Theta_h(\beta)\) is positive-definite everywhere.
Proof.
If \(\kappa_*>0\), then every eigenvalue of \(\mathcal{T}_h(\beta)(y)\) is bounded below by \(\kappa_*\), so \(\mathcal{T}_h(\beta)(y)\) is positive-definite for every \(y\). Since \(\mathcal{T}_h(\beta)=\beta\Theta_h(\beta)\) and \(\beta>0\), the same is true of \(\Theta_h(\beta)\). Taking traces yields \[ \beta(y)\mathcal{Q}_h(\beta)(y)=\operatorname{tr}_h\mathcal{T}_h(\beta)(y)\ge m\kappa_*>0, \] so \(\mathcal{Q}_h(\beta)>0\).
If \(\Theta_h(\beta)(X,X)\le 0\) at \((y_0,X)\), then \[ \mathcal{T}_h(\beta)(X,X)=\beta(y_0)\Theta_h(\beta)(X,X)\le 0. \] Hence the smallest eigenvalue of \(\mathcal{T}_h(\beta)(y_0)\) is nonpositive, so \(\kappa_*\le 0\). The compact statement follows from Proposition 3.4. \(\square\)
Remark 4.2.
Proposition 4.1 clarifies the relation between the local and global criteria. A positive threshold forces positivity of every local obstruction, while a single nonpositive tensor direction excludes the positive-threshold regime. The threshold criterion is therefore best viewed as the global completion of the local obstruction theory.
Proposition 4.3 (Constant \(\beta\) and Einstein slices).
Suppose \(\beta\equiv c>0\) is constant. Then \[ \mathcal{T}_h(\beta)=c^2\operatorname{Ric}_h, \qquad \kappa_*=c^2\inf_{y\in N}\lambda_{\min}^h\bigl(\operatorname{Ric}_h(y)\bigr). \] In particular, if \(h\) is Einstein with \(\operatorname{Ric}_h=\lambda h\), then \(\kappa_*=c^2\lambda\).
Proof.
If \(\beta\) is constant, then \(\nabla_h\beta=0\) and \(\nabla^2_h\beta=0\), so \(\mathcal{T}_h(\beta)=\beta^2\operatorname{Ric}_h\). The eigenvalue formula is immediate. \(\square\)
Remark 4.4.
Proposition 4.3 shows that the threshold criterion extends the standard constant-curvature examples in the simplest possible way. Positive Einstein slices lie in the positive-threshold regime, Ricci-flat slices lie in the borderline regime, and negative Einstein slices lie in the obstructed regime.
Compact and noncompact model families
The compact case is especially transparent because positivity of \(\mathcal{T}_h(\beta)\) immediately produces a uniform positive lower bound. On noncompact slices, by contrast, the smallest eigenvalue may drift to zero or to \(-\infty\) at infinity, so the global invariant \(\kappa_*\) is genuinely stronger than pointwise positivity at individual points.
Example 4.5 (Compact positive regime).
If \(n\ge 3\), \(N=S^{n-1}\) is round, and \(\beta\equiv 1\), then \(\mathcal{T}_h(\beta)=(n-2)h\), so the positive regime is realized with a uniform gap. In particular, the periodic family \(A(t)=1+\varepsilon\sin(\omega t)\) with \(\omega^2<n-2\) gives explicit bounded positive nonconstant separated sub-static triples.
Example 4.6 (Borderline compact regime).
If \(N=T^{n-1}\) is flat and \(\beta\equiv c>0\), then \(\mathcal{T}_h(\beta)=0\) and the condition \(S\ge 0\) reduces to \(A''\ge 0\). The borderline regime is therefore genuinely rigid: every bounded positive line factor is constant.
Example 4.7 (One-dimensional negative regime).
Let \(N=\mathbb{R}\) with metric \(dy^2\) and \(\beta(y)=\cosh y\). Since the slice is one-dimensional, \(\operatorname{Ric}_h=0\) and \[ \mathcal{T}_h(\beta)=-2\beta\beta''+(\beta\beta''+(\beta')^2)dy^2. \] Using \(\beta''=\cosh y\) and \((\beta')^2=\sinh^2 y\), one obtains \[ \mathcal{T}_h(\beta)=\bigl(-\cosh^2 y+\sinh^2 y\bigr)dy^2=-dy^2. \] Thus \(\kappa_*=-1<0\).
Example 4.8 (Anisotropic noncompact negative regime).
The Euclidean example in Example 3.5(iii) shows that the negative regime may be caused by a single bad direction even when all transverse directions are uniformly positive. This anisotropy is precisely why the threshold is defined through the smallest eigenvalue rather than through a trace or averaged quantity.
Why the threshold is the correct invariant
The local quantities \(\mathcal{Q}_h(\beta)\) and \(\Theta_h(\beta)\) arise naturally from the slice decomposition, but the existence problem for bounded positive nonconstant line factors is not fundamentally a trace problem and not even a pointwise tensor problem. The relevant inequality is \[ A(t)\mathcal{T}_h(\beta)(y)+A''(t)h(y)\ge 0. \] For each fixed \(t\), the only slice quantity that controls this inequality in all unit directions simultaneously is the smallest eigenvalue of \(\mathcal{T}_h(\beta)(y)\). Passing from \(y\) to the whole slice then leads exactly to \(\kappa_*\).
From this point of view, the hierarchy of statements in the paper is natural. Tracing the slice block gives the trace obstruction, evaluating on one vector gives the tensor obstruction, and minimizing over all points and directions gives the exact threshold criterion. The strongest theorem is therefore not a separate idea from the local obstructions; it is the sharp global form of the same slice identity.
Scaling and One-Dimensional Slices
Two simple structural observations are worth recording separately. The first concerns the effect of rescaling the lapse factor \(\beta\). The second gives a complete description of the threshold tensor when the slice is one-dimensional.
Proposition 5.1 (Scaling in \(\beta\)).
For every constant \(c>0\), \[ \mathcal{T}_h(c\beta)=c^2\mathcal{T}_h(\beta), \qquad \kappa_*(h,c\beta)=c^2\kappa_*(h,\beta). \] Consequently, multiplying \(\beta\) by a positive constant preserves the sign of \(\kappa_*\).
Proof.
Using the definition, \[ \begin{aligned} \mathcal{T}_h(c\beta) &=(c\beta)^2\operatorname{Ric}_h-2(c\beta)\nabla^2_h(c\beta) \\ +\bigl((c\beta)\Delta_h(c\beta)+|\nabla_h(c\beta)|_h^2\bigr)h \\ &=c^2\beta^2\operatorname{Ric}_h-2c^2\beta\nabla^2_h\beta \\ +c^2\bigl(\beta\Delta_h\beta+|\nabla_h\beta|_h^2\bigr)h \\ &=c^2\mathcal{T}_h(\beta). \end{aligned} \] Taking the smallest eigenvalue and the infimum over \(N\) yields the formula for \(\kappa_*\). \(\square\)
Remark 5.2.
Proposition 5.1 is consistent with the separated metric \(g=\beta^2dt^2+h\). Multiplying \(\beta\) by a constant rescales the distinguished line direction and therefore changes the threshold by a quadratic factor. The sign of the threshold, which is the quantity relevant to rigidity and existence, is unaffected.
Proposition 5.3 (One-dimensional slice formula).
Assume \(m=1\), so \(N\) is one-dimensional with local coordinate \(y\) and metric \(h=dy^2\). Then \[ \mathcal{T}_h(\beta)=\bigl((\beta')^2-\beta\beta''\bigr)dy^2 =-\beta^2(\ln\beta)''dy^2. \] In particular, \[ \kappa_*=\inf_{y\in N}\bigl((\beta'(y))^2-\beta(y)\beta''(y)\bigr). \] Proof.
In one dimension \(\operatorname{Ric}_h=0\), \(\nabla^2_h\beta=\beta''dy^2\), and \(\Delta_h\beta=\beta''\). Substituting into the definition gives \[ \mathcal{T}_h(\beta)=-2\beta\beta''dy^2+\bigl(\beta\beta''+(\beta')^2\bigr)dy^2 =\bigl((\beta')^2-\beta\beta''\bigr)dy^2. \] Since \[ (\ln\beta)'=\frac{\beta'}{\beta}, \qquad (\ln\beta)''=\frac{\beta\beta''-(\beta')^2}{\beta^2}, \] the second identity follows. \(\square\)
Corollary 5.4 (Compact one-dimensional slices).
Assume \(m=1\) and \(N\) is compact. Then every bounded positive line factor is constant. Equivalently, there is no bounded positive nonconstant line factor in the separated sub-static ansatz over a compact one-dimensional slice.
Proof.
If \(\beta\) is constant, then \(\mathcal{T}_h(\beta)=0\) by Proposition 5.3, so \(\kappa_*=0\) and Corollary 3.3 gives constancy of \(A\).
Assume \(\beta\) is nonconstant. Since \(N\) is compact, \(\beta\) attains a minimum at some point \(y_0\). There one has \(\beta'(y_0)=0\) and \(\beta''(y_0)\ge 0\). Proposition 5.3 then gives \[ \mathcal{T}_h(\beta)(y_0)=-\beta(y_0)\beta''(y_0)dy^2\le 0, \] so \(\kappa_*\le 0\). Corollary 3.3 again forces \(A\) to be constant. \(\square\)
Remark 5.5.
Corollary 5.4 shows that the positive-threshold regime is inherently a higher-dimensional phenomenon, at least for compact slices. In one dimension the threshold can still be positive on noncompact examples, but compactness forces a minimum point of \(\beta\) and therefore a nonpositive value of \(\mathcal{T}_h(\beta)\).
Example 5.6 (Three one-dimensional models).
The one-dimensional formula makes several examples completely transparent.
If \(\beta(y)=e^{ay}\) on \(\mathbb{R}\), then \((\beta')^2-\beta\beta''=0\), so \(\kappa_*=0\).
If \(\beta(y)=\cosh y\), then \((\beta')^2-\beta\beta''=-1\), so \(\kappa_*=-1\); this recovers Example 4.7.
If \(\beta(y)=e^{-y^2}\) on \(\mathbb{R}\), then \[ (\beta')^2-\beta\beta''=2e^{-2y^2}, \] so \(\kappa_*=0\) even though the threshold is strictly positive at every finite point. This illustrates that on noncompact slices, pointwise positivity does not imply the uniform positive gap required by Corollary 3.3.
Three-Dimensional Consequences
Assume from now on that \(n=3\), so the slice \(N\) is a surface.
Proposition 7.1 (Critical-point eigenvalues).
Let \(y_0\) be a critical point of \(\beta\), and let \(\lambda_1,\lambda_2\) be the eigenvalues of \(\nabla^2_h\beta(y_0)\). Then the eigenvalues of \(\Theta_h(\beta)(y_0)\) are \[ \mu_1=\beta(y_0)K_h(y_0)+\lambda_2-\lambda_1, \qquad \mu_2=\beta(y_0)K_h(y_0)+\lambda_1-\lambda_2. \] Consequently, \[ \operatorname{tr}_h\Theta_h(\beta)(y_0)=2\beta(y_0)K_h(y_0)=\beta(y_0)R_h(y_0), \] and \[ \det_h\Theta_h(\beta)(y_0)=\beta(y_0)^2K_h(y_0)^2-(\lambda_1-\lambda_2)^2. \] Proof.
At a critical point, \(\nabla_h\beta(y_0)=0\), so \(\Delta_h\beta(y_0)=\lambda_1+\lambda_2\). Since \(N\) is a surface, \(\operatorname{Ric}_h=K_hh\). Therefore \[ \Theta_h(\beta)(y_0)=\beta(y_0)K_h(y_0)h-2\nabla^2_h\beta(y_0)+(\lambda_1+\lambda_2)h. \] Diagonalizing \(\nabla^2_h\beta(y_0)\) in an orthonormal basis gives the stated formulas. The trace and determinant formulas are immediate. \(\square\)
Theorem 7.2 (Three-dimensional curvature-anisotropy criterion).
Assume \(n=3\) and that \(A\) is bounded, positive, and nonconstant. At every critical point \(y_0\) of \(\beta\), \[ \beta(y_0)K_h(y_0)>|\lambda_1-\lambda_2|. \] Equivalently, \[ \det_h\Theta_h(\beta)(y_0)>0 \qquad\text{and}\qquad \operatorname{tr}_h\Theta_h(\beta)(y_0)>0. \] In particular, \(K_h(y_0)>0\) at every critical point of \(\beta\).
Proof.
By Proposition 6.2, nonconstant bounded line factors are possible only if \(\Theta_h(\beta)\) is positive-definite at every point. At a critical point \(y_0\), this means \(\mu_1>0\) and \(\mu_2>0\), where \(\mu_1,\mu_2\) are the eigenvalues from Proposition 7.1. Their positivity is exactly the curvature-anisotropy inequality. The determinant and trace conditions are equivalent on a surface. Since \(\beta>0\), the inequality implies \(K_h(y_0)>0\). \(\square\)
Corollary 7.3 (Critical-point obstruction).
Assume \(n=3\) and let \(y_0\) be a critical point of \(\beta\).
If \(K_h(y_0)<|\lambda_1-\lambda_2|/\beta(y_0)\), then no bounded positive line factor exists.
If \(K_h(y_0)=|\lambda_1-\lambda_2|/\beta(y_0)\), then every bounded positive line factor is constant.
If \(R_h(y_0)\le 0\), then every bounded positive line factor is constant.
Proof.
The first two statements follow from Proposition 7.1 together with Proposition 6.2: strict inequality makes one eigenvalue of \(\Theta_h(\beta)(y_0)\) negative, while equality makes one eigenvalue zero. For (iii), since \(R_h=2K_h\) on a surface and \(\nabla_h\beta(y_0)=0\), the trace formula gives \[ \mathcal{Q}_h(\beta)(y_0)=\beta(y_0)R_h(y_0). \] Thus \(R_h(y_0)\le 0\) implies \(\mathcal{Q}_h(\beta)(y_0)\le 0\), and Proposition 6.1 forces constancy. \(\square\)
Conformal Reformulation and Boundary Control
Set \(u=\beta^{-1}\) and \(\bar{h}=u^2h\). The threshold tensor admits the following conformal expression.
Proposition 8.1 (Conformal identity).
With \(m=\dim N\), \(u=\beta^{-1}\), and \(\bar{h}=u^2h\), \[ \mathcal{T}_h(\beta)=u^{-2}\operatorname{Ric}_{\bar{h}}+mu^{-3}\nabla^2_{\bar{h}} u. \] Consequently, \[ \kappa_*=\inf_{y\in N}\lambda_{\min}^{\bar{h}}\bigl(\operatorname{Ric}_{\bar{h}}+mu^{-1}\nabla^2_{\bar{h}} u\bigr). \tag{kstar-conf} \] Proof.
Set \(f=\ln u\), so that \(\bar{h}=e^{2f}h\). Under the conformal change \(\bar{h}=e^{2f}h\), \[ \operatorname{Ric}_{\bar{h}} =\operatorname{Ric}_h-(m-2)\bigl(\nabla^2_hf-df\otimes df\bigr) -\bigl(\Delta_hf+(m-2)|\nabla_hf|_h^2\bigr)h. \] Since \(f=\ln u\), \[ df=u^{-1}du, \qquad \nabla^2_hf=u^{-1}\nabla^2_hu-u^{-2}du\otimes du, \] \[ \Delta_hf=u^{-1}\Delta_hu-u^{-2}|\nabla_hu|_h^2, \qquad |\nabla_hf|_h^2=u^{-2}|\nabla_hu|_h^2. \] Substituting yields \[ \begin{aligned} \operatorname{Ric}_{\bar{h}} &=\operatorname{Ric}_h-(m-2)u^{-1}\nabla^2_hu+2(m-2)u^{-2}du\otimes du \\ &\quad -\bigl(u^{-1}\Delta_hu+(m-3)u^{-2}|\nabla_hu|_h^2\bigr)h. \end{aligned} \] The Hessian transforms by \[ \nabla^2_{\bar{h}}u =\nabla^2_hu-df\otimes du-du\otimes df+\langle \nabla_hf,\nabla_hu\rangle_hh, \] and therefore \[ \nabla^2_{\bar{h}}u =\nabla^2_hu-2u^{-1}du\otimes du+u^{-1}|\nabla_hu|_h^2h. \] Multiply the Ricci identity by \(u^{-2}\) and the Hessian identity by \(mu^{-3}\), then add: \[ \begin{aligned} u^{-2}\operatorname{Ric}_{\bar{h}}+mu^{-3}\nabla^2_{\bar{h}}u &=u^{-2}\operatorname{Ric}_h+2u^{-3}\nabla^2_hu-4u^{-4}du\otimes du \\ &\quad +\bigl(-u^{-3}\Delta_hu+3u^{-4}|\nabla_hu|_h^2\bigr)h. \end{aligned} \] Now substitute \(u=\beta^{-1}\). Then \[ du=-\beta^{-2}d\beta, \qquad \nabla^2_hu=-\beta^{-2}\nabla^2_h\beta+2\beta^{-3}d\beta\otimes d\beta, \] \[ \Delta_hu=-\beta^{-2}\Delta_h\beta+2\beta^{-3}|\nabla_h\beta|_h^2, \qquad |\nabla_hu|_h^2=\beta^{-4}|\nabla_h\beta|_h^2. \] Substituting these identities gives \[ u^{-2}\operatorname{Ric}_{\bar{h}}+mu^{-3}\nabla^2_{\bar{h}}u =\beta^2\operatorname{Ric}_h-2\beta\nabla^2_h\beta +(\beta\Delta_h\beta+|\nabla_h\beta|_h^2)h =\mathcal{T}_h(\beta). \] This proves the conformal identity.
For (kstar-conf), note that \(\bar{h}=u^2h\), so \[ h^{-1}\mathcal{T}_h(\beta) =h^{-1}\bigl(u^{-2}(\operatorname{Ric}_{\bar{h}}+mu^{-1}\nabla^2_{\bar{h}}u)\bigr) =\bar{h}^{-1}\bigl(\operatorname{Ric}_{\bar{h}}+mu^{-1}\nabla^2_{\bar{h}}u\bigr). \] Thus the corresponding endomorphisms have the same spectrum at each point, and taking the infimum of the smallest eigenvalue gives (kstar-conf). \(\square\)
Remark 8.2.
Proposition 8.1 rewrites the threshold condition as an eigenvalue bound for a Bakry–'Emery-type tensor on the conformally rescaled slice \((N,\bar{h})\). This provides a natural interface between the threshold invariant and the comparison geometry developed in [BF].
Remark 8.3 (Geometric meaning of the conformal formula).
The conformal identity separates two geometric effects. The Ricci term measures the intrinsic curvature of the compactified slice, while the Hessian term measures the deviation of the defining function \(u=\beta^{-1}\) from being affine in the compactified metric. When \(u\) is geodesic, the Hessian becomes a second-fundamental-form quantity for the level-set foliation, and the threshold can be read off from the curvature of those level sets. This is exactly the mechanism used in the collar argument below.
We now give a concrete geometric class in which the threshold is forced to be nonpositive.
Theorem 8.4 (Geodesic-slice boundary control).
Assume \(n=3\). With \(u=\beta^{-1}\) and \(\bar{h}=u^2h\), suppose in a collar \(\{0<u<\varepsilon\}\subset \bar N\) that:
(A1) \(|\nabla_{\bar{h}}u|\equiv 1\);
(A2) \(\partial\bar N\) is totally geodesic in \((\bar N,\bar{h})\);
(A3) the level sets \(\{u=c\}\) have geodesic curvature \(k_u\ge 0\).
Then \(\kappa_*\le 0\). More precisely:
- if \(k_u\not\equiv 0\) on the collar, then \(\kappa_*<0\), hence no bounded positive line factor exists;
- if \(k_u\equiv 0\) on the collar, then every bounded positive line factor is constant.
Proof.
Because \(n=3\), the slice is a surface, so \(\operatorname{Ric}_{\bar{h}}=K_{\bar{h}}\bar{h}\). Under (A1), if \(N_u\) denotes the unit normal to the level sets of \(u\) and \(T\) a unit tangent vector, then \[ \nabla^2_{\bar{h}}u(N_u,\cdot)=0, \qquad \nabla^2_{\bar{h}}u(T,T)=k_u. \] Therefore the two eigenvalues of \(\operatorname{Ric}_{\bar{h}}+2u^{-1}\nabla^2_{\bar{h}}u\) in the \((N_u,T)\)-frame are \[ \lambda_{N_u}=K_{\bar{h}}, \qquad \lambda_T=K_{\bar{h}}+2u^{-1}k_u. \] In geodesic coordinates \(\bar{h}=du^2+\gamma(u,\theta)^2d\theta^2\), the Riccati identity is \[ k_u'=-K_{\bar{h}}-k_u^2. \tag{riccati} \] By (A2), \(k_u(0,\theta)=0\) for all \(\theta\), and (A3) gives \(k_u\ge 0\) in the collar.
If \(k_u\equiv 0\), then (riccati) implies \(K_{\bar{h}}\equiv 0\), so \(\lambda_{N_u}=\lambda_T=0\) in the collar. By (kstar-conf), this gives \(\kappa_*\le 0\). Corollary 3.3 then implies that every bounded positive line factor is constant.
Assume now that \(k_u\not\equiv 0\). Then there exists \((u_0,\theta_0)\) with \(k_u(u_0,\theta_0)>0\). Since \(k_u(0,\theta_0)=0\) and \(k_u\ge 0\), there is \(u_1\in(0,u_0]\) such that \(k_u'(u_1,\theta_0)>0\). Evaluating (riccati) there gives \[ K_{\bar{h}}(u_1,\theta_0)=-k_u'(u_1,\theta_0)-k_u(u_1,\theta_0)^2<0. \] Hence \(\lambda_{N_u}(u_1,\theta_0)<0\), so (kstar-conf) gives \(\kappa_*<0\). Corollary 3.3 now rules out bounded positive nonconstant line factors. \(\square\)
Remark 8.5 (Role of the collar assumptions).
The assumptions in Theorem 8.4 are geometrically distinct. Assumption (A1) turns \(u\) into a distance-type function for the compactified metric. Assumption (A2) fixes the initial condition for the level-set curvature at the boundary. Assumption (A3) provides the sign that forces the Riccati equation to produce a point of nonpositive curvature unless the collar is flat in the relevant direction. The theorem is therefore a boundary-to-interior mechanism for forcing nonpositivity of the threshold.
A Formal Collar Expansion and the Wang Interface
The previous section produces a concrete collar condition implying \(\kappa_*\le 0\). To relate this more directly to Wang’s Neumann-data framework [WangArea], one needs pointwise control of the threshold eigenvalues near conformal infinity. The next lemma records the correct formal expansion in a geodesic collar.
Lemma 9.1 (Formal geodesic-collar expansion).
Assume in geodesic coordinates that \[ \bar{h}=du^2+\gamma(u,\theta)^2d\theta^2, \qquad \gamma(u,\theta)=1+a(\theta)u^2+b(\theta)u^3+c(\theta)u^4+O(u^5). \] Then \[ K_{\bar{h}}=-2a-6bu+\bigl(2a^2-12c\bigr)u^2+O(u^3), \] \[ k_u=2au+3bu^2+\bigl(4c-2a^2\bigr)u^3+O(u^4), \] and therefore the eigenvalues of \(\operatorname{Ric}_{\bar{h}}+2u^{-1}\nabla^2_{\bar{h}}u\) in the normal/tangential frame are \[ \lambda_N=-2a-6bu+\bigl(2a^2-12c\bigr)u^2+O(u^3), \] \[ \lambda_T=2a-\bigl(2a^2+4c\bigr)u^2+O(u^3). \tag{eigNT} \] In particular, if \(a\equiv 0\), then \[ \lambda_N=-6bu-12cu^2+O(u^3), \qquad \lambda_T=-4cu^2+O(u^3). \] Proof.
Differentiate the expansion of \(\gamma\): \[ \gamma_u=2au+3bu^2+4cu^3+O(u^4), \qquad \gamma_{uu}=2a+6bu+12cu^2+O(u^3). \] Also, \[ \gamma^{-1}=1-au^2-bu^3+(a^2-c)u^4+O(u^5). \] Since \[ K_{\bar{h}}=-\frac{\gamma_{uu}}{\gamma}, \qquad k_u=\frac{\gamma_u}{\gamma}, \] we obtain \[ K_{\bar{h}} =-\gamma_{uu}\gamma^{-1} =-2a-6bu+\bigl(2a^2-12c\bigr)u^2+O(u^3), \] and \[ k_u =\gamma_u\gamma^{-1} =2au+3bu^2+\bigl(4c-2a^2\bigr)u^3+O(u^4). \] Finally, substitute these expansions into (eigNT). \(\square\)
Remark 9.2.
Lemma 9.1 is intentionally formal. To turn it into a direct boundary criterion for \(\kappa_*\le 0\) in Wang’s setting, one still needs additional geometric input controlling the coefficient \(a(\theta)\) and the pointwise sign of \(b(\theta)\). Wang’s second-variation formula in [WangArea] gives integrated information on the boundary data, while the threshold condition is pointwise.
Remark 9.3.
This mismatch between integrated and pointwise information appears to be the basic obstruction to closing Wang’s remaining separated-ansatz question by the Neumann-data method alone. The threshold is governed by an extremal eigenvalue, hence by pointwise signs. A boundary integral can contribute to that problem only after an additional argument upgrades average information to a pointwise statement.
Discussion and Open Problems
Within the separated ansatz, Theorem 3.2 reduces the sub-static tensor inequality to a scalar ODE with a single geometric parameter \(\kappa_*\). This reformulation makes several consequences especially transparent:
- bounded positive nonconstant line factors exist exactly when \(\kappa_*>0\);
- local strict negativity of the trace or tensor obstruction gives nonexistence, not merely constancy;
- on compact slices, existence is equivalent to pointwise positive-definiteness of \(\mathcal{T}_h(\beta)\).
These statements seem to capture the cleanest reusable content of the separated theory.
The conformal and collar results point toward the broader splitting problem, but they do not by themselves settle the general conformally compact case considered in [WangRiccati, WangArea]. The remaining issue is the passage from boundary data to pointwise control of the threshold eigenvalues.
From a practical point of view, the paper suggests three different workflows for applying the theory. On compact slices one should check positive-definiteness of \(\mathcal{T}_h(\beta)\) directly, since Corollary 6.3 gives an if-and-only-if criterion. On noncompact slices, one should instead estimate the global lower spectral bound \(\kappa_*\), because pointwise positivity alone need not be enough. In dimension three, critical points of \(\beta\) supply a third route: Theorem 7.2 converts the problem to an explicit inequality involving curvature and Hessian anisotropy.
The boundary-control theorem suggests a fourth, more geometric route. Rather than computing \(\mathcal{T}_h(\beta)\) explicitly on the original slice, one may compactify conformally and study the geometry of the defining function \(u=\beta^{-1}\). This is useful when the slice comes with natural conformal boundary data. The obstacle, made precise in Section 9, is that the threshold is fundamentally pointwise while current Neumann-data formulas are primarily integral in nature.
A practical workflow
For actual applications, the threshold formalism suggests a concrete order of attack. First, one computes the tensor \(\mathcal{T}_h(\beta)\) from its definition. In many examples this is easier than working directly with the full sub-static tensor on \(\mathbb{R}\times N\), because the separated ansatz has already eliminated the mixed and \(tt\) components.
Second, one checks whether the problem is compact or noncompact. On compact slices, Corollary 6.3 reduces the existence question to a pointwise definiteness test for \(\Theta_h(\beta)\) or \(\mathcal{T}_h(\beta)\). No further global analysis is needed beyond continuity of the eigenvalues. On noncompact slices, however, one must genuinely estimate the infimum of the smallest eigenvalue. Examples such as the Gaussian-type one-dimensional model in Example 5.6(iii) show that pointwise positivity may coexist with \(\kappa_*=0\).
Third, one uses local obstructions whenever a full spectral estimate is unavailable. Propositions 6.1 and 6.2 are especially useful at critical points of \(\beta\) or at points where a preferred geometric frame is available. In dimension three this becomes particularly effective, because Theorem 7.2 translates positivity of \(\Theta_h(\beta)\) into a scalar inequality involving \(K_h\) and the Hessian anisotropy of \(\beta\).
Fourth, if the slice comes from a conformally compact problem, one can pass to the compactified metric and attempt to control the eigenvalues of \(\operatorname{Ric}_{\bar{h}}+mu^{-1}\nabla^2_{\bar{h}}u\). Theorem 8.4 is one instance of this strategy: it does not compute \(\mathcal{T}_h(\beta)\) directly, but instead infers the sign of \(\kappa_*\) from the geometry of the level-set foliation of \(u\). This separation between computational and geometric routes is one of the main practical advantages of the threshold viewpoint.
Open Problem 10.1.
Prove \(\kappa_*\le 0\) in Wang’s conformally compactifiable sub-static setting without the extra collar assumptions of Theorem 8.4. By Corollary 3.3, this would force the line factor to be constant and would complete the separated-ansatz step of the splitting program.
Open Problem 10.2.
Find geometric conditions on \((N,h,\beta)\) guaranteeing positivity or nonpositivity of \(\kappa_*\) without computing the full spectrum of \(\mathcal{T}_h(\beta)\). In dimension three, can one derive such conditions directly from topology or asymptotic data on the conformal boundary?
Open Problem 10.3.
Generalize Theorem 8.4 to higher-dimensional slices. In dimensions \(m\ge 3\), the Ricci tensor is no longer a scalar multiple of the metric, so one needs an analogue of the two-dimensional Riccati argument that directly controls the smallest eigenvalue of \(\operatorname{Ric}_{\bar{h}}+mu^{-1}\nabla^2_{\bar{h}}u\).
Appendices
Warped-Product Formulas
This appendix records the basic formulas behind Lemma 2.1. They are standard, but gathering them in one place keeps the main argument focused on the threshold reduction.
Let \[ g=\beta(y)^2dt^2+h_{ij}(y)dy^idy^j, \qquad V(t,y)=A(t)\beta(y). \] Then \[ g_{tt}=\beta^2, \quad g_{ij}=h_{ij}, \quad g^{tt}=\beta^{-2}, \quad g^{ij}=h^{ij}. \] The nonzero Christoffel symbols involving the \(t\)-direction are \[ \Gamma^t_{it}=\beta^{-1}\partial_i\beta, \qquad \Gamma^i_{tt}=-\beta\nabla_h^i\beta. \] All purely spatial Christoffel symbols are those of \(h\).
Hessian and Laplacian
Since \(\partial_tV=A'\beta\) and \(\partial_iV=A\partial_i\beta\), the Hessian components are \[ \begin{aligned} \nabla^2_gV(\partial_t,\partial_t) &=\partial_t\partial_tV-\Gamma^k_{tt}\partial_kV \\ =A''\beta+A\beta|\nabla_h\beta|_h^2, \\ \nabla^2_gV(\partial_i,\partial_t) &=\partial_i(A'\beta)-\Gamma^t_{it}A'\beta \\ =0, \\ \nabla^2_gV(\partial_i,\partial_j) &=A\nabla^2_h\beta_{ij}. \end{aligned} \] Tracing these identities gives \[ \Delta_gV=\beta^{-2}\nabla^2_gV(\partial_t,\partial_t)+h^{ij}\nabla^2_gV(\partial_i,\partial_j) =\beta^{-1}A''+A\Delta_h\beta+A\beta^{-1}|\nabla_h\beta|_h^2. \]
Ricci tensor
For the warped product \(g=\beta^2dt^2+h\) one has \[ \operatorname{Ric}_g(\partial_t,\partial_t)=-\beta\Delta_h\beta, \qquad \operatorname{Ric}_g(\partial_i,\partial_t)=0, \qquad \operatorname{Ric}_g(\partial_i,\partial_j)=\operatorname{Ric}_h{}_{ij}-\beta^{-1}\nabla^2_h\beta_{ij}. \] Substituting the Hessian, Laplacian, and Ricci formulas into \[ S(V,g)=V\operatorname{Ric}_g+(\Delta_gV)g-\nabla^2_gV \] gives the cancellation in the \(tt\)-component and yields the slice decomposition (Sij).
Auxiliary ODE Lemmas
The exact threshold criterion reduces the geometric problem to an elementary ODE inequality. The lemmas below isolate the one-dimensional facts used repeatedly in the main text.
Lemma A.1 (Bounded convex functions).
If \(A\in C^2(\mathbb{R})\) satisfies \(A''\ge 0\) and is bounded above on \(\mathbb{R}\), then \(A\) is constant.
Proof.
The derivative \(A'\) is monotone nondecreasing. If \(A'(t_0)>0\) at some point, then \(A'(t)\ge A'(t_0)>0\) for all \(t\ge t_0\), which forces \(A(t)\to+\infty\) as \(t\to+\infty\). If \(A'(t_0)<0\), then \(A'(t)\le A'(t_0)<0\) for all \(t\le t_0\), which forces \(A(t)\to+\infty\) as \(t\to-\infty\). Hence \(A'\equiv 0\). \(\square\)
Lemma A.2 (Positive forcing).
Let \(A\in C^2(\mathbb{R})\) be positive. If there exists \(c>0\) such that \[ A''(t)\ge cA(t) \qquad\text{for all } t\in\mathbb{R}, \] then \(A\) is unbounded on \(\mathbb{R}\).
Proof.
Since \(A''\ge cA>0\), the derivative \(A'\) is strictly increasing. Fix \(t_0\in\mathbb{R}\). If \(A'(t_0)\ge 0\), then \(A'(t)>0\) for all \(t>t_0\), so \(A(t)\to+\infty\) as \(t\to+\infty\). If \(A'(t_0)<0\), then \(A'(t)\le A'(t_0)<0\) for all \(t<t_0\), hence \[ A(t)\ge A(t_0)-A'(t_0)(t_0-t)\to+\infty \quad(t\to-\infty). \qquad\square \] Lemma A.3 (Periodic test family).
If \(\kappa>0\), then for every \(0<\omega^2<\kappa\) and \(0<\varepsilon<1\), \[ A(t)=1+\varepsilon\sin(\omega t) \] is bounded, positive, nonconstant, and satisfies \(A''+\kappa A>0\) on \(\mathbb{R}\).
Proof.
Direct computation gives \[ A''+\kappa A =\kappa+\varepsilon(\kappa-\omega^2)\sin(\omega t) \ge \kappa-\varepsilon(\kappa-\omega^2)>0. \qquad\square \] ## Expanded Model Calculations
This appendix records the detailed calculations for the model families used in the body of the paper.
Round sphere
Assume \(n\ge 3\), let \(N=S^{n-1}\) with the round metric of sectional curvature \(1\), and \(\beta\equiv 1\). Then \[ \operatorname{Ric}_h=(n-2)h, \qquad \nabla_h\beta=0, \qquad \nabla^2_h\beta=0, \] so \[ \Theta_h(\beta)=(n-2)h, \qquad \mathcal{T}_h(\beta)=(n-2)h, \qquad \kappa_*=n-2. \] This is the basic compact positive example.
Flat torus
If \(N=T^{n-1}\) is flat and \(\beta\equiv c>0\), then all derivative terms vanish and \(\operatorname{Ric}_h=0\). Hence \[ \Theta_h(\beta)=0, \qquad \mathcal{T}_h(\beta)=0, \qquad \kappa_*=0. \] This realizes the borderline rigid regime.
Euclidean anisotropic example
Let \(N=\mathbb{R}^m\) be Euclidean and \(\beta(y)=1+(y^1)^2\). Then \[ \nabla\beta=2y^1dy^1, \qquad \nabla^2\beta=2dy^1\otimes dy^1, \qquad \Delta\beta=2. \] Since \(\operatorname{Ric}_h=0\), \[ \begin{aligned} \mathcal{T}_h(\beta) &=-2\beta\nabla^2\beta+(\beta\Delta\beta+|\nabla\beta|^2)h \\ &=-4(1+(y^1)^2)dy^1\otimes dy^1 \\ +\bigl(2(1+(y^1)^2)+4(y^1)^2\bigr)\sum_{\alpha=1}^{m}dy^\alpha\otimes dy^\alpha \\ &=\bigl(-2+2(y^1)^2\bigr)dy^1\otimes dy^1 \\ +\bigl(2+6(y^1)^2\bigr)\sum_{\alpha=2}^{m}dy^\alpha\otimes dy^\alpha. \end{aligned} \] Thus the smallest eigenvalue is \(-2+2(y^1)^2\), and its infimum equals \(-2\).
One-dimensional negative example
Let \(N=\mathbb{R}\) with \(h=dy^2\) and \(\beta(y)=\cosh y\). Since \(\operatorname{Ric}_h=0\) in one dimension, \[ \mathcal{T}_h(\beta)=\bigl(-2\beta\beta''+(\beta\beta''+(\beta')^2)\bigr)dy^2. \] Using \(\beta''=\cosh y\) and \((\beta')^2=\sinh^2 y\), one finds \[ \mathcal{T}_h(\beta) =\bigl(-2\cosh^2 y+\cosh^2 y+\sinh^2 y\bigr)dy^2 =-dy^2. \] Hence \(\kappa_*=-1\).
Exponential example with positive trace
Consider the example from Section 6: \[ N=\mathbb{R}^2, \quad h=dx^2+dy^2, \quad \beta(x,y)=e^x+c, \quad c\ge 0. \] Because \(\operatorname{Ric}_h=0\), \[ \Theta_h(\beta)=-2\nabla^2_h\beta+\Bigl(\Delta_h\beta+\beta^{-1}|\nabla\beta|_h^2\Bigr)h. \] Now \[ \nabla^2_h\beta=e^xdx\otimes dx, \qquad \Delta_h\beta=e^x, \qquad |\nabla\beta|_h^2=e^{2x}. \] Therefore \[ \Delta_h\beta+\beta^{-1}|\nabla\beta|_h^2 =e^x+\frac{e^{2x}}{e^x+c}, \] and \[ \begin{aligned} \Theta_h(\beta) &=-2e^xdx\otimes dx \\ +\Bigl(e^x+\frac{e^{2x}}{e^x+c}\Bigr)(dx^2+dy^2) \\ &= \\ \Bigl(-e^x+\frac{e^{2x}}{e^x+c}\Bigr)dx\otimes dx +\Bigl(e^x+\frac{e^{2x}}{e^x+c}\Bigr)dy\otimes dy \\ &= \\ -\frac{ce^x}{e^x+c}\,dx\otimes dx +\Bigl(e^x+\frac{e^{2x}}{e^x+c}\Bigr)dy\otimes dy. \end{aligned} \] Taking the trace gives \[ \mathcal{Q}_h(\beta) =-\frac{ce^x}{e^x+c}+e^x+\frac{e^{2x}}{e^x+c} =2\frac{e^{2x}}{e^x+c}>0. \] Thus the trace obstruction is silent even though the tensor obstruction is active. When \(c=0\), the \(dx\)-eigenvalue vanishes identically. When \(c>0\), the \(dx\)-eigenvalue is strictly negative.
Surface critical-point algebra
At a critical point of \(\beta\) on a surface, the formula for \(\Theta_h(\beta)\) simplifies enough to be useful in explicit calculations. Let \(\lambda_1,\lambda_2\) be the eigenvalues of \(\nabla^2_h\beta\) in an orthonormal frame. Since \(\nabla_h\beta=0\) and \(\operatorname{Ric}_h=K_hh\), \[ \Theta_h(\beta)=\beta K_hh-2\nabla^2_h\beta+(\lambda_1+\lambda_2)h. \] Hence the two eigenvalues of \(\Theta_h(\beta)\) are \[ \mu_1=\beta K_h+\lambda_2-\lambda_1, \qquad \mu_2=\beta K_h+\lambda_1-\lambda_2. \] From these formulas one reads off immediately that \[ \operatorname{tr}_h\Theta_h(\beta)=2\beta K_h, \qquad \det_h\Theta_h(\beta)=\beta^2K_h^2-(\lambda_1-\lambda_2)^2. \] Thus positivity of \(\Theta_h(\beta)\) at a critical point is equivalent to the pair of inequalities \[ 2\beta K_h>0, \qquad \beta^2K_h^2-(\lambda_1-\lambda_2)^2>0, \] which in turn collapse to the single curvature-anisotropy inequality \[ \beta K_h>|\lambda_1-\lambda_2|. \] This is the algebraic core of Theorem 7.2.
A Checklist for Computing the Threshold
For readers who want to apply the theory to new examples, it is useful to summarize the workflow suggested by the paper in a compact form.
Step 1: compute the slice tensor
Start from \[ \mathcal{T}_h(\beta)=\beta^2\operatorname{Ric}_h-2\beta\nabla^2_h\beta+(\beta\Delta_h\beta+|\nabla_h\beta|_h^2)h. \] In symmetric examples this often collapses immediately. For constant \(\beta\), one reduces to the Ricci tensor. In one dimension, Proposition 5.3 gives the closed formula \(\mathcal{T}_h(\beta)=-\beta^2(\ln\beta)''dy^2\).
Step 2: decide whether compactness is available
If \(N\) is compact, the problem simplifies sharply: the existence of a bounded positive nonconstant line factor is equivalent to pointwise positive-definiteness of \(\mathcal{T}_h(\beta)\). In that case one does not need to estimate the infimum of the smallest eigenvalue globally; continuity and compactness do the work automatically.
If \(N\) is noncompact, by contrast, pointwise positivity is only preliminary information. One must determine whether the smallest eigenvalue stays uniformly away from zero. The examples in Sections 4 and 5 show that this distinction is essential.
Step 3: use local obstructions when full diagonalization is hard
If a direct spectral computation is cumbersome, the local obstructions can still be very effective. A single point with \(\mathcal{Q}_h(\beta)<0\) already rules out bounded positive line factors, and a single nonpositive tensor direction is enough to force rigidity or nonexistence. In dimension three, critical points of \(\beta\) provide especially convenient test points, because the condition reduces to the curvature-anisotropy inequality of Theorem 7.2.
Step 4: move to the conformal picture when boundary geometry is natural
When the slice is naturally described through a compactification, Proposition 8.1 replaces \(\mathcal{T}_h(\beta)\) by the tensor \[ \operatorname{Ric}_{\bar{h}}+mu^{-1}\nabla^2_{\bar{h}}u. \] This form is better adapted to geometric foliation arguments and to level-set methods. Theorem 8.4 is an example: it never computes \(\mathcal{T}_h(\beta)\) directly, but still controls the sign of \(\kappa_*\).
Step 5: conclude through the sign of \(\kappa_*\)
Once the sign is known, Corollary 3.3 finishes the argument. A positive threshold produces explicit oscillatory line factors. A zero threshold forces convexity and hence constancy. A negative threshold produces positive forcing and hence unboundedness. In this sense, all of the geometry in the separated ansatz feeds into a single scalar decision at the final step.
References
[BF] S. Borghini and M. Fogagnolo, Comparison geometry for sub-static manifolds, 2023, arXiv:2307.14618.
[CFMR] G. Colombo, A. Freitas, L. Mari, and M. Rigoli, Some splitting and rigidity results for sub-static spaces, 2024, arXiv:2412.05238.
[GSW] G. J. Galloway, S. Surya, and E. Woolgar, On the geometry and mass of static, asymptotically AdS spacetimes, and the uniqueness of the AdS soliton, Comm. Math. Phys. 241 (2003), 1–25; arXiv:hep-th/0108170.
[WangRiccati] Z. Wang, Riccati equation for static spaces and its applications, 2025, arXiv:2408.12180.
[WangArea] Z. Wang, Neumann data and second variation formula of renormalized area for conformally compact static spaces, 2025, arXiv:2504.15048.
Evaluation:
GPT:
My assessment is:
Correctness: the paper looks substantively plausible but not yet referee-safe. Novelty: moderate, specialized-note level, not breakthrough level. Journal fit: a specialized geometry/global analysis journal, most naturally Q2 at current strength.
The strongest part is the exact reduction in Sections 2–3: for the separated ansatz \[g=\beta(y)^2dt^2+h(y),\qquad V(t,y)=A(t)\beta(y),\] the paper isolates the slice tensor \[T_h(\beta)=\beta^2\mathrm{Ric}*h-2\beta\nabla_h^2\beta+(\beta\Delta_h\beta+|\nabla_h\beta|*h^2)h,\] defines \[\kappa**=\inf*{y\in N}\lambda_{\min}^h(T_h(\beta)(y)),\] and proves the equivalence \[S(V,g)\ge 0 \iff A''+\kappa_*A\ge 0,\] followed by the bounded positive nonconstant criterion “iff \(\kappa_*>0\).” That is the real mathematical contribution of the draft, and it is clearly articulated in the abstract and main theorem section.
On correctness, the core warped-product computation and the ODE consequences appear internally coherent. In particular, Lemma 2.1, Theorem 3.2, Corollary 3.3, and the compact-slice criterion fit together logically, and the examples are aligned with the sign trichotomy \(\kappa_*>0,\ \kappa_*=0,\ \kappa_*<0\). But I would still call the paper “correct modulo careful checking,” not “clearly correct.” The conformal reformulation and collar sections are coefficient-sensitive and therefore the part most likely to attract referee scrutiny; they may well be right, but they need especially careful line-by-line verification before submission. There are also some presentational/formal issues: for example, Theorem 3.2 really wants an explicit finiteness remark for \(\kappa_*\) in the statement, not only nearby; and the paper itself says the later sections are more of an “interface” to Wang’s program than a complete closure of the broader problem.
On novelty, I would rate it as moderate. The main theorem is a clean and useful extraction of the separated-ansatz problem into a single threshold invariant, together with sharp local corollaries, a compact criterion, and a 3D critical-point test. That is a legitimate contribution. But the paper also explicitly positions itself as addressing only the separated ansatz step rather than proving a full new splitting theorem, and Section 10 leaves the Wang-interface problem open. So this is not reading like a top-tier theorem paper; it reads more like a well-structured specialized note that clarifies one unresolved reduction step and packages it sharply.
Because of that novelty level, I would not aim first at journals like Journal of Differential Geometry or The Journal of Geometric Analysis. JDG is a clear mismatch at current strength, and JGA generally asks for stronger innovation at the geometry/analysis interface than this draft currently delivers. JGA’s scope emphasizes innovative work at the crossroads of geometry, analysis, and PDE, and it is currently listed as Q1; AGAG is aimed squarely at global analysis and geometry and is currently Q2; Differential Geometry and its Applications is also squarely on-topic and is currently Q2. (Springer)
So my recommendation is:
- Best fit: Annals of Global Analysis and Geometry — the scope matches very well: global analysis, differential geometry, variational/global geometric questions. Current standing is Q2. (Springer)
- Also reasonable: Differential Geometry and its Applications — also very natural in scope, currently Q2. (ScienceDirect)
- Broader fallback: Mediterranean Journal of Mathematics if you want a less specialized and somewhat safer venue; it is broad-scope and currently Q2. (Springer)
My bottom line:
\[\text{correctness status} \approx \text{promising but still needs hard checking},\] \[\text{novelty level} \approx \text{solid specialized note},\] \[\text{recommended tier} \approx \text{Q2 specialized geometry/global analysis}. \]
If you want a sharper label: I would call this “strong Q2 / maybe borderline low-Q1 only if significantly strengthened.” To move it upward, the paper would need one of the following: \[\text{either } \big(\text{close the Wang separated-ansatz problem}\big),\] \[\text{or } \big(\text{replace the collar/interface section by a genuinely stronger theorem}\big),\] \[\text{or } \big(\text{extract a broader geometric criterion beyond the separated model}\big).\]
At its current level, AGAG is the cleanest recommendation.
Claude:
Correctness
The paper is mathematically correct as far as I can verify. The arguments are clean and the logic is tight throughout.
The core computation in Lemma 2.1 — the slice decomposition of \(S(V,g)\) under the warped-product ansatz — is a straightforward Christoffel symbol / Hessian / Ricci calculation. I checked the key cancellation yielding \(S_{tt}=0\) and \(S_{it}=0\), and the surviving slice block formula (6) is consistent with standard warped-product identities.
Theorem 3.2 (the exact threshold criterion) follows almost immediately once you observe that the slice inequality \(A(t)\,T_h(\beta)(y) + A''(t)\,h(y) \geq 0\) is linear in the pair \((A\cdot\lambda, A'')\) and that minimizing over unit directions and points extracts exactly \(\kappa_*\). The logic is correct and essentially optimal — there’s no slack in the equivalence.
Corollary 3.3 (bounded-solution criterion) uses only elementary ODE facts: convex bounded functions on \(\mathbb{R}\) are constant, and \(A'' \geq cA > 0\) forces unboundedness. These are stated and proved cleanly in Appendix B.
The conformal identity (Proposition 8.1) involves a longer computation but follows a well-trodden path — conformal transformation of Ricci, substitution of \(u = \beta^{-1}\). I traced through the key steps and the algebra checks out.
The three-dimensional critical-point criterion (Theorem 7.2) is a direct algebraic consequence of diagonalizing \(\Theta_h(\beta)\) at a critical point of \(\beta\) on a surface. No issues there.
One minor organizational note: the paper references “Section 4” for local tests but the sharp local obstructions actually appear in Section 6. The introduction mentions this correctly, but Section 4’s discussion sometimes reads as if it’s stating the results that are only proved later. This is a presentational choice, not an error.
Verdict on correctness: Sound. No errors detected.
Novelty
This is where a more nuanced assessment is needed. The paper’s contribution should be evaluated at several levels:
What is genuinely new. The identification of \(\kappa_*\) as the exact threshold controlling the bounded-solution problem within the separated ansatz is, as far as I know, new and cleanly formulated. The equivalence in Theorem 3.2 is a precise structural result — it’s not just a sufficient condition but an if-and-only-if. The three-dimensional determinant criterion at critical points (Theorem 7.2) and the conformal reformulation (Proposition 8.1) are nice additions that give the threshold geometric content beyond the abstract spectral definition.
What limits the novelty. The paper is entirely confined to the separated ansatz (2). It does not solve the general sub-static splitting problem, and the authors are upfront about this. The separated ansatz is already a very special structural assumption — the metric is a warped product and the potential factors multiplicatively. Within this ansatz, the key observation (that the slice block of \(S\) factors into a geometric tensor times \(A\) plus \(A''\) times \(h\)) is natural once you write things out. The ODE analysis that follows is elementary.
The conformal and collar sections (8–9) gesture toward the broader problem but explicitly acknowledge they don’t resolve it. The formal collar expansion in Section 9 is essentially bookkeeping, and Remark 9.3 candidly states the fundamental obstacle (pointwise vs. integrated information).
The paper also doesn’t engage deeply with the competing approaches in the literature — the stable-minimal-hypersurface methods of Colombo–Freitas–Mari–Rigoli or the weighted comparison geometry of Borghini–Fogagnolo. The relationship is mentioned but not developed.
Assessment. The paper makes a correct, clean, and complete contribution to a specific technical subproblem (the separated-ansatz step) within the sub-static splitting program. The threshold criterion is sharp within its domain. However, the scope is narrow — the separated ansatz is a restricted setting, and the paper’s results don’t directly advance the general splitting conjecture. The techniques are competent but not deep; the main arguments use standard warped-product formulas and elementary ODE theory.
Journal Recommendation
Given the profile — correct and well-written, clean structural result, but narrow scope and limited technical depth — this paper fits naturally into a solid specialty journal in differential geometry or geometric analysis.
Appropriate journals would include:
- Annals of Global Analysis and Geometry — good fit for the scope and level
- Differential Geometry and its Applications — natural home for this type of structural result
- Journal of Geometry and Physics — if the GR motivation is emphasized
- Pacific Journal of Mathematics or Proceedings of the AMS — if shortened to focus on the main theorem and its immediate corollaries
These are roughly mid-tier journals (not top-tier like Annals of PDE, CPAM, Duke, or Inventiones, but reputable journals where a focused, correct contribution finds an appropriate audience). In terms of typical rankings, these sit in the range of solid Q1–Q2 journals depending on the classification system.
The paper would likely not be competitive for Annals of PDE (which the memory notes as a target), because Annals of PDE generally expects either substantially new techniques or resolution of significant open problems, and this paper does neither — it resolves a subproblem within an ansatz using relatively elementary methods. That said, if the threshold criterion turns out to be a key ingredient in eventually resolving the full splitting conjecture, the paper’s significance would be reassessed retrospectively.
Gemini:
Here is an evaluation of the article based on correctness, novelty, and subsequent journal recommendations.
1. Correctness
The article demonstrates a high degree of mathematical correctness, grounded in well-established principles of Riemannian geometry and geometric analysis.
Methodological Soundness: The fundamental computations rely on standard, verifiable techniques, such as warped product curvature formulas, slice decompositions, and conformal transformations.
Rigorous Proofs: The central result (Theorem 3.2), which reduces the multi-dimensional sub-static tensor inequality to the one-dimensional scalar ODE \(A''(t) + \kappa_* A(t) \ge 0\), is rigorously proven by systematically evaluating the bilinear forms.
Logical Deductions: The paper correctly utilizes elementary ODE maximum-principle arguments (detailed in Appendix B) to prove when line factors must be unbounded or constant based on the sign of the threshold invariant.
Self-Aware Constraints: The author is mathematically careful, acknowledging where the theory is incomplete, such as noting that the formal collar expansion (Lemma 9.1) requires additional geometric input to fully close Wang’s remaining questions.
2. Novelty
The novelty of the paper is moderate to high within its specific subfield, characterized by an elegant structural insight rather than a massive, field-altering breakthrough.
Specific Problem Resolution: The paper’s primary novel contribution is providing an exact, one-dimensional ODE criterion for the “separated ansatz,” a specific unresolved bottleneck in Wang’s recent line-based splitting approach.
Conceptual Shift: Instead of relying on local ad-hoc bounds, the author introduces the “threshold invariant” \(\kappa_*\), effectively showing that the existence problem is entirely governed by a single global spectral gap.
Scope Limitations: The author explicitly bounds the novelty, stating the paper should be read as a “contribution to the separated-ansatz step… rather than as a full new splitting theorem for arbitrary sub-static manifolds”.
Future Utility: The conformal reformulation of the threshold tensor (Proposition 8.1) and the explicit identification of a fundamental mismatch (pointwise vs. integrated information) in current Neumann-data boundary approaches offer valuable, original stepping stones for future research.
3. Journal Recommendations & Tier Assessment
Because the paper solves a very specific intermediate step (the separated ansatz) in an ongoing mathematical program rather than resolving the full sub-static splitting conjecture, it is best suited for high-quality, specialized journals rather than top-tier general mathematics journals (like Annals of Mathematics or Journal of the AMS).
Recommended Tier: Tier 2 / Strong Q1 in Specialized Mathematics The paper is crisp, exact, and structurally illuminating, making it an excellent candidate for respected journals focusing on geometric analysis, differential geometry, or mathematical physics.
Recommended Journals:
- Journal of Geometric Analysis: An ideal fit. This journal frequently publishes strong, targeted results in geometric PDEs, conformal geometry, and curvature inequalities.
- Communications in Analysis and Geometry: A highly respected journal that focuses heavily on the intersection of analysis, differential equations, and differential geometry.
- Classical and Quantum Gravity: If the author wishes to frame the work closer to its physical roots (static Einstein spaces and general relativity), this Q1 physics journal is highly appropriate.
- Proceedings of the American Mathematical Society (PAMS): Given the paper’s relatively short length, clear exact criteria, and self-contained proofs, PAMS is an excellent venue for concise, impactful results.