Nonlinear Stability and Instability

Nonlinear stability asks whether a spacetime arising from nearby initial data remains globally controlled and relaxes, after gauge and parameter adjustment, to a member of the same stationary family. The global nonlinear stability of Minkowski space supplies the asymptotically flat dispersive prototype Christodoulou and Klainerman 1993, Chapters 1–3. For black holes, the statement must additionally account for horizons, trapping, radiation, changing mass and spin, and pure coordinate modes. This chapter builds the claim from its definition through its linear estimates and nonlinear closure, then marks the exact boundary of current results.

Helpful background. Black-Hole Geometry and Mechanics supplies the Kerr family and horizon geometry; Black-Hole Perturbation Theory supplies linear decay and zero modes; Formulations of General Relativity supplies constraints and hyperbolic gauges.

The theorem-sized question

What Nonlinear Black-Hole Stability Means defines the data class, quotient by diffeomorphisms, orbital stability, endpoint modulation, decay norms, and completeness. The immediate lesson is that convergence to one fixed coordinate metric is neither necessary nor sufficient.

From Minkowski Stability to Kerr Stability identifies the genuinely new black-hole structures: event-horizon redshift, trapping, ergoregion energy, superradiance, and a stationary family of endpoints. Linear Estimates as Nonlinear Inputs then asks which boundedness, integrated decay, low-frequency, and source estimates are strong enough to control quadratic errors.

Gauge, Modulation, and Bootstrap Closure combines those inputs into a proof architecture. Nonlinear Instability Mechanisms distinguishes genuine nonlinear coupling from saturation of a linear instability. Criteria for Nonlinear Black-Hole Stability closes with the current evidence boundary and requirements for independent nonlinear evolutions.

The stability decomposition

The recurring decomposition is

$$g(t)=\Phi_t^*g_{b(t)}+h_{\rm rad}(t), \qquad b(t)\longrightarrow b_\infty, \qquad h_{\rm rad}(t)\longrightarrow0$$

in region- and norm-dependent senses. The diffeomorphism $\Phi_t$ removes pure gauge, $b(t)$ absorbs physical zero modes and asymptotic frame motion, and $h_{\rm rad}$ carries dispersive degrees of freedom.

A stability theorem must control all three terms. Without $\Phi_t$, gauge growth can resemble instability. Without $b(t)$, a harmless final mass shift prevents convergence. Without decay of $h_{\rm rad}$, bounded orbital motion need not represent relaxation.

Current result and its boundary

In the source notation used in this section, $G=c=1$: Kerr mass $M$ and spin parameter $a=J/M$ are lengths, and subextremality is $|a|/M\lt 1$. With dimensions restored, the same ratio is $\chi=cJ/(GM_{\rm phys}^2)$.

As of 2026-08-12, Hintz’s v2 preprint proves nonlinear exterior stability for small vacuum perturbations of every fixed subextremal Kerr background. The simple theorem assumes smooth constraint data agreeing with Kerr outside a compact set; the precise theorem permits a controlled partially polyhomogeneous asymptotic class. The solution approaches a nearby Kerr metric with quantified behavior near the horizon, null infinity, and timelike infinity Hintz 2026, Theorem 1.1, pp. 2–3, and Theorem 13.1, pp. 318–323, PDF.

“Every fixed subextremal background” does not include extremal Kerr or uniform estimates in the limit $|a|/M\to1$. The theorem is vacuum, small-data, high-regularity, asymptotically flat, and exterior. Matter instabilities, large-data final states, and complete interiors are different questions.

Earlier Schwarzschild-family and slowly rotating results remain conceptually important. They supply independent proof architectures and make the roles of geometric gauge, decay, and final-parameter identification transparent Dafermos et al. 2021, Theorem I.3.1 and §V, pp. 5–8 and 21–28, PDF, Klainerman and Szeftel 2022, §§1.3 and 4, pp. 4–6 and 23–32, PDF. Their asymptotic data classes also differ: wider spin coverage does not automatically mean inclusion of every earlier theorem’s data.

For a first pass, follow the definition, the flat-space comparison, and the solved exercises. Then return to the linear-estimate and gauge pages: they explain why a theorem needs quantitative decay rather than a list of damped frequencies. The research criteria page is most useful after those distinctions are familiar.

How to read an instability claim

Before comparing a claim with Kerr stability, identify:

• vacuum or matter equations;
• subextremal, extremal, or superextremal parameter regime;
• exterior, event horizon, or interior region;
• asymptotically flat, de Sitter, anti-de Sitter, or reflecting boundary;
• small open neighborhood or finite-amplitude data;
• linear growing mode, nonlinear resonance, backreaction, or basin-boundary crossing;
• gauge-dependent or invariant diagnostic.

This classification resolves many apparent contradictions. Superradiant growth of a massive field is not vacuum instability; Superradiant Evolution, Saturation, and Hair Formation develops that coupled matter problem and its distinct endpoints. Mass inflation concerns the interior. AdS turbulence uses a reflecting boundary. Critical collapse is a finite-amplitude threshold phenomenon.

Chapter diagnostic

An exact nearby Kerr solution $g_{M+\delta M,a+\delta a}$ is used as initial data in a stability test around $g_{M,a}$. Its difference from the reference metric does not decay. Has nonlinear stability failed?

Solution

No. The data lie exactly on the stationary Kerr family. The nondecaying difference is a physical parameter mode and should be absorbed by modulation to the endpoint $(M+\delta M,a+\delta a)$. After that adjustment the radiative remainder is zero. A definition demanding decay to the original parameters would incorrectly label the stationary family itself unstable.

What the chapter enables

You should now be able to parse a stability theorem by its data, gauge, region, endpoint, and norm; identify the linear estimates that support nonlinear closure; distinguish zero modes from radiation; and decide whether a proposed instability lies inside or outside a theorem’s hypotheses. Production algorithms and linear-response derivations remain in their dedicated volumes.

References

• Christodoulou, Demetrios, and Sergiu Klainerman, The Global Nonlinear Stability of the Minkowski Space, Princeton University Press, 1993.
• Dafermos, Mihalis, Gustav Holzegel, Igor Rodnianski, and Martin Taylor, “The Nonlinear Stability of the Schwarzschild Family of Black Holes”, arXiv:2104.08222v1, 2021. Open PDF.
• Hintz, Peter, “Nonlinear Stability of Subextremal Kerr Black Holes”, arXiv:2606.28253v2, 2026. Open PDF.
• Klainerman, Sergiu, and Jérémie Szeftel, “Brief Introduction to the Nonlinear Stability of Kerr”, arXiv:2210.14400v2, 2022. Open PDF.