Complete Vanishing Theorem

• The complete vanishing theorem is a framework in algebraic geometry, differential geometry, and combinatorics that defines conditions ensuring that key invariants vanish identically.
• It is applied in local cohomology, L2 harmonic analyses, and minimal model programs, linking spectral connectedness, curvature bounds, and positivity conditions to vanishing results.
• Techniques involve Mayer–Vietoris sequences, Bochner–Weitzenböck formulas, and holographic transformations, offering practical insights into structural and computational dichotomies.

The term "complete vanishing theorem" encompasses several distinct contexts in contemporary mathematics, particularly in algebraic geometry, the theory of local cohomology, differential geometry, and computational complexity. Across these domains, a complete vanishing theorem typically refers to a rigorous criterion under which a significant class of algebraic, cohomological, or combinatorial invariants are forced to vanish identically, often under minimal or "complete" hypotheses. This article provides a comprehensive treatment of the notion as it appears in select settings, notably local cohomology, topological vanishing for surfaces, $L^2$-cohomology on Riemannian or Hessian manifolds, and in the theory of Holant complexity dichotomies.

1. Complete Vanishing Theorem in Local Cohomology

A central instance of a complete vanishing theorem arises in local cohomology, classically formulated as the second vanishing theorem (SVT). Let $(R, \mathfrak{m}, k)$ be a $d$-dimensional complete regular local ring and $I \subset R$ a proper ideal. The $i$-th local cohomology module with support in $I$ is denoted $H^i_I(R)$. The SVT establishes an equivalence between the vanishing of $H^{d-1}_I(R)$ and the connectedness of the punctured spectrum $\Spec^\circ(R/I) = \Spec(R/I) \setminus \{\mathfrak{m}\}$ under the following hypotheses:

• For all $\mathfrak{p} \in \operatorname{Min}(R/I)$, $\dim(R/\mathfrak{p}) \ge 3$,
• $H^j_I(R)$ has finite length for $j \le d-2$.

Precisely,

$H^{d-1}_I(R) = 0 \iff \Spec^\circ(R/I) \text{ is connected}$

in the Zariski topology, provided $R$ is unramified (for mixed characteristic, $p \notin \mathfrak{m}^2$). This result—initiated by Huneke, Lyubeznik, Ogus, Hartshorne, and Speiser—generalizes to mixed characteristic with additional finiteness conditions on certain "fibred" cohomology modules. The case of complete ramified regular local rings (i.e., $p \in \mathfrak{m}^2$ in mixed characteristic) remains partially unresolved; the presence of surjective elements in local cohomology and the finiteness of certain second local cohomology modules play a crucial role in extending the theorem to these settings (Asgharzadeh et al., 2021).

2. Vanishing Theorems for $L^2$ Harmonic Forms and Cohomology

Another significant family of complete vanishing theorems concerns $L^2$-cohomology on complete Riemannian or Hessian manifolds. The prototype is as follows:

On a complete Riemannian manifold $(M, g)$, the space of $L^2$ harmonic $k$-forms (with respect to the Hodge Laplacian) vanishes under sharp lower bounds on curvature and the existence of weighted Poincaré inequalities. Specifically, if $M$ satisfies a weighted Poincaré inequality

$\int_{M} q\, \phi^2 \leq \int_{M} |\nabla \phi|^2$

for every compactly supported smooth $\phi$, and the curvature operator $R$ on $k$-forms satisfies $(R \omega, \omega) \geq -a q |\omega|^2$ for all $k$-forms $\omega$ with $0 < a < C_{n,k}$, then all $L^2$-harmonic $k$-forms vanish unless $M$ is compact and $q \equiv 0$. These results extend with extra terms (e.g., an additional negative constant in the curvature bound) to more flexible scenarios, yielding rigidity and vanishing outcomes in the borderline case (Vieira, 2014).

In the context of manifolds equipped with parallel $1$-forms, the result is even more absolute: if $(M, g)$ is a complete Riemannian manifold carrying a nonzero parallel $1$-form $\theta$, then for all $0 \leq k \leq \dim M$,

$H^k_{(2)}(M, \theta) = 0,$

where the cohomology is taken with respect to the twisted differential $d_\theta = d + \theta \wedge \cdot$ (the Morse–Novikov cohomology). This encompasses, as a corollary, the vanishing of ordinary $L^2$-de Rham cohomology in all degrees (Huang et al., 2019).

On the side of Hessian manifolds, Akagawa proved that $L^2$-Dolbeault-type cohomology groups vanish for complete Hessian manifolds under positivity of the total second Koszul form; for regular convex cones with the Cheng–Yau metric, vanishing holds for bidegrees $p > q$, yielding a precise $L^2$-analogue of Kodaira–Nakano vanishing (Akagawa, 2017).

3. Complete Vanishing for Surfaces and the Minimal Model Program

Fujino and Moriyama provided a version of a vanishing theorem for surfaces that is both sufficient and minimal for the implementation of the minimal model program (MMP) for log surfaces. For a smooth complex analytic surface $X$, boundary $\mathbb{R}$-divisor $\Delta$ with simple normal crossings, line bundle $\mathcal{L}$, and proper morphisms $f: X \to Y$, $\pi: Y \to Z$ (with $\psi = \pi \circ f$ projective), the theorem asserts: If $\mathcal{L} - (\omega_X+\Delta)$ is $\psi$-nef and $\psi$-big over $Z$, and for every component $C \subset \lfloor\Delta\rfloor$ contracted to a point by $\psi$,

$(\mathcal{L} - (\omega_X + \Delta))\cdot C > 0,$

then all higher direct images vanish:

$$R^p\pi_*\,R^qf_*\,\mathcal L = 0 \quad \text{for all }p>0,\,q\ge0.$$

This formulation suffices for both the basepoint-free and abundance theorems and does not require the full strength of Hodge theoretic vanishing theorems. The proof relies on an inductive application of Reid–Fukuda type vanishing and strict support and torsion-freeness lemmas (Fujino et al., 24 Nov 2025).

4. Complete Vanishing Theorem in Holant Complexity: Vanishing Signatures

The notion of a complete vanishing theorem in computational complexity—specifically in the framework of Holant problems—takes a fundamentally distinct form. Here, a set of constraint functions (signatures) $\mathcal{F}$ over Boolean variables is said to be vanishing if every associated Holant sum

$\Holant_\Omega(\mathcal{F})$

is identically $0$ for every signature grid $\Omega$. Such sets are classified via the vanishing degree of symmetric signatures. Denoting by $\vd^+$ and $\vd^-$ the positive and negative vanishing degrees, and by $\V^+$ and $\V^-$ the corresponding classes of symmetric signatures with $2 \vd^\sigma(f) > \textrm{arity}(f)$ $(\sigma = +,\ -)$, the complete vanishing theorem (Theorem 3.14) asserts:

$\mathcal{F} \text{ is vanishing } \iff \mathcal{F} \subseteq \V^+ \ \text{or}\ \mathcal{F} \subseteq \V^-.$

This dichotomic criterion is essential in the full complexity classification of Holant problems over symmetric signatures, constituting the so-called "fourth" and "fifth" tractable cases in Yin, Cai, and Lu's complete dichotomy. The characterization can be recast in terms of recurrence relations or, after holographic transformations, in terms of the structure of vanishing tensors (Cai et al., 2012).

Context	Statement (summary)	Key Reference
Local cohomology (SVT)	$H^{d-1}_I(R)=0$ iff $\Spec^\circ(R/I)$ connected	(Asgharzadeh et al., 2021)
$L^2$ cohomology, parallel $1$-form	$H^k_{(2)}(M,\theta)=0$ for all $k$	(Huang et al., 2019)
$L^2$ cohomology, curvature lower bounds	$L^2$ harmonic $k$-forms vanish under Bochner–Poincaré hypotheses	(Vieira, 2014)
Surfaces for MMP	$R^p \pi_* R^qf_* \mathcal{L} = 0$ in stated positivity context	(Fujino et al., 24 Nov 2025)
Holant vanishing signatures	$\mathcal{F} \subseteq \V^{\pm} \implies$ all Holant sums vanish	(Cai et al., 2012)

5. Methodologies and Proof Techniques

In local cohomology, the proof of the second vanishing theorem leverages surjective elements in local cohomology modules, reduction to hypersurfaces, and Mayer–Vietoris sequences. Regular elements (often the characteristic $p$) act surjectively on artinian modules in a range of degrees, forcing vanishing at a critical index, while Mayer–Vietoris arguments translate connectedness of spectra into vanishing of cohomology.

For $L^2$-vanishing theorems, core arguments combine Bochner–Weitzenböck formulae, refined Kato inequalities, and weighted Poincaré inequalities. Integration by parts and heat kernel (spectral theory) techniques feature essentially. In the presence of parallel $1$-forms, the additional $|\theta|^2 \mathrm{Id}$ term in the Laplacian guarantees coercivity and forces vanishing.

In the context of surfaces, the methods reduce to inductive applications of Kawamata–Viehweg-like vanishing theorems and cohomological exact sequences, refined by strict support conditions that control the locus of potential nonvanishing higher direct images.

In Holant theory, the critical innovation consists in the classification of vanishing signatures via symmetrizations, analysis of recurrence relations of signature entries, combinatorial arguments for forced cancellation in the sum over edge assignments, and the decisive use of holographic transformations to bring signatures into forms where vanishing is manifest.

6. Scope, Significance, and Open Problems

Complete vanishing theorems serve as sharp tools in a wide range of contexts:

• In algebraic geometry, they underpin results on connectedness, abundance, and the foundational steps of the minimal model program for surfaces.
• In differential geometry, they imply rigidity or triviality of global invariants under strong geometric hypotheses, as in the absence of $L^2$ harmonic forms on manifolds with certain curvature properties or parallel forms.
• In computational complexity, complete vanishing criteria demarcate the tractable border for counting complexity in the Holant framework.

There remain unresolved problems, especially in the context of local cohomology for ramified regular local rings in mixed characteristic, where the existence of requisite surjective elements or desired finite-length properties of certain local cohomology modules is not generally assured (Asgharzadeh et al., 2021). In Holant theory, the spectrum of partial vanishing, when mixed vanishing types are present, leads to computational hardness, marking a sharp contrast with the dichotomous tractable cases (Cai et al., 2012).

7. Related Results and Generalizations

Vanishing theorems form an interwoven network with results such as the Kodaira vanishing theorem, the Kawamata–Viehweg vanishing theorem, the Bochner vanishing theorem, and the Kodaira–Nakano $L^2$-vanishing theorem for complete Hessian manifolds (Akagawa, 2017). The role of positivity—be it of line bundle curvature, Koszul forms, or Laplacian lower bounds—is pervasive throughout vanishing frameworks.

The structure of complete vanishing theorems, cutting across cohomological, geometric, and combinatorial invariants, continues to provide insights into the deep linkage between topological connectedness, algebraic positivity, geometric curvature, and computational tractability.