$W^{1,2}_{\mathrm{loc}}$-formulation of strong cosmic censorship in the spherically symmetric Einstein-Maxwell-real-scalar-field model: inextendibility of all nontrivial $C^0$ extensions for generic two-ended data

For the two-ended, spherically symmetric Einstein-Maxwell-real-scalar-field system, the fate of the strong cosmic censorship conjecture is decided at low regularity. Dafermos (2014) proved that the maximal development of admissible data admits a $C^0$ extension through the Cauchy horizon (so the $C^0$-formulation of strong cosmic censorship is false) while exhibiting mass inflation and failure of $L^2_{\mathrm{loc}}$-regularity of the Christoffel symbols in that particular constructed extension; Luk and Oh (2019) proved $C^2$-inextendibility for generic data, closing the $C^2$-formulation. The surviving open core, stated explicitly as open in Dafermos's Section 13 remark and in Luk-Oh Remark 1.6, and confirmed as still open by Van de Moortel's 2025 survey, is Christodoulou's $H^1$-formulation (Statement II): whether for generic two-ended admissible data every nontrivial $C^0$ Lorentzian extension of the maximal future development fails $W^{1,2}_{\mathrm{loc}}$ regularity, i.e. whether the spacetime cannot be continued even as a weak solution of the equations.

Background

The Einstein-Maxwell-real-scalar-field (EMSF) system couples the Einstein equations of general relativity to an electromagnetic field and a real (uncharged) scalar field via the energy-momentum tensor. In spherical symmetry the metric can be written in double-null coordinates $(u,v)$ with angular part $r^2(u,v)\,d\omega^2$, and the equations reduce to a $(1+1)$-dimensional hyperbolic system for the scalar field $\phi$ together with constraint equations for the mass aspect and the charge. Two-ended asymptotically flat initial data are data on a complete spacelike hypersurface with two asymptotically flat ends, so that the maximal development describes both the exterior of a charged black hole and the mirror-image second exterior connected through an Einstein-Rosen bridge; by time reversal, the second exterior generates a black hole interior region. This is the model in which Dafermos (2014), following Christodoulou's earlier nonlinear scattering theory for the equations, carried out the definitive spherically symmetric analysis of the black hole interior.

The Reissner-Nordström spacetime (the charged, spherically symmetric electrovacuum solution) has a smooth interior in which the strong spacelike singularity of Schwarzschild is replaced by a Cauchy horizon: a regular null boundary beyond which the metric can be smoothly continued but the continuation is not determined by initial data, so determinism fails. The strong cosmic censorship conjecture asserts that for generic initial data such non-unique continuation is impossible once the metric is required to have enough regularity for the Einstein equations to be meaningful. The conjecture is stratified by regularity: Statement I requires inextendibility of the metric as a continuous ($C^0$) Lorentzian metric; Statement II, a formulation attributed to Christodoulou and discussed in Dafermos's introduction, requires inextendibility in the $W^{1,2}_{\mathrm{loc}}$ class, i.e. that no $C^0$ extension has locally square-integrable Christoffel symbols (the coordinate components of the Levi-Civita connection, whose square-integrability makes the field equations well defined in the weak, distributional sense); Statement III requires $C^2$-inextendibility, the classical form.

Perturbations of the two-ended data decay along the event horizon according to an analogue of Price's law (polynomial decay of the scalar field and its derivatives in the advanced and retarded times), and this decaying radiation is blue-shifted along the Cauchy horizon, where the renormalized Hawking mass (a mass functional that is constant on the horizons of Reissner-Nordström and blows up where the Cauchy horizon becomes singular) tends to infinity: this is the mass-inflation mechanism, conjectured by Israel and by Poisson-Israel and confirmed for this model by Dafermos under pointwise lower bounds on the horizon decay of the scalar field. Dafermos (2014) proved, for admissible data (the standard class of smooth two-ended asymptotically flat initial data of Christodoulou and Dafermos, fixed by regularity and asymptotic-decay hypotheses at the two ends under which the exterior of the black hole is well behaved; the precise definitions are given in Luk-Oh Part I, Section 3, and in the other papers cited below) satisfying the decay assumptions, that the Cauchy horizon of the maximal development is approached by a $C^0$-extendible metric with bounded area radius $r$ but infinite Hawking mass, so Statement I is false in this model; that under the pointwise horizon lower bounds of his Theorem 6 (bounds of the form $|\partial_v\phi|\gtrsim v^{-6+\tau}$ and $|\partial_u\phi|\gtrsim u^{-6+\tau}$ assumed in equation (22) of that paper) mass inflation holds and the metric cannot be extended with continuous metric derivatives; and that in the particular $C^0$ extension he constructs the Christoffel symbols fail to lie in $L^2_{\mathrm{loc}}$ near the Cauchy horizon.

Two later developments define the current frontier. First, Luk and Oh (2019) proved for generic two-ended data (open in a weighted $C^1$ topology, dense in weighted $C^\infty$) that the maximal development is $C^2$-future-inextendible, establishing Statement III; their Part I assumes an $L^2$-averaged polynomial lower bound on the horizon decay (weaker than the pointwise bounds) and their Part II proves that bound for the generic class, thereby resolving the genericity question (Dafermos's Conjecture 2) in stronger form than posed. Second, Sbierski developed gauge-invariant low-regularity inextendibility criteria: a holonomy criterion implying $C^{0,1}_{\mathrm{loc}}$ (Lipschitz) inextendibility for weak null singularities including the Luk-Oh spacetimes, and later a curvature blow-up criterion for Lipschitz inextendibility without symmetry. These establish the intermediate rung between $C^0$ and $C^2$, but Lipschitz regularity is strictly stronger than $W^{1,2}$, so they leave the $H^1$-formulation untouched. Luk and Oh moreover proved (their Proposition C.1) that in their constructed $C^0$ extension the Christoffel symbols fail $L^p_{\mathrm{loc}}$ for every $p>1$ and the scalar field fails $W^{1,2}_{\mathrm{loc}}$ near non-degenerate points of the Cauchy horizon, under the generic lower bound.

The remaining gap is that all known blow-up statements are proved in the particular $C^0$ extension constructed in double-null coordinates. Whether they hold for an arbitrary nontrivial $C^0$ Lorentzian extension — the quantifier that Statement II demands — is open, and it is known to be a genuinely delicate question: Cameron and Sbierski (2025) showed that $C^0$ extensions of spacetimes with weak null singularities are non-unique in general, so there is no canonical extension to reduce to; and in the neighbouring exactly-extremal charged-scalar model Gajic and Luk proved that the interior is $W^{1,2}_{\mathrm{loc}}$-extendible, so the statement is false in nearby models and any proof must use non-extremality in an essential way.

Problem Statement

Consider the Einstein-Maxwell-real-scalar-field system in spherical symmetry with two-ended asymptotically flat admissible initial data, for the generic class of Luk and Oh (data that are open in a suitable weighted $C^1$ topology and dense in weighted $C^\infty$; in particular the data satisfy the $L^2$-averaged polynomial lower bound on horizon decay proven in their Part II; the precise definitions of admissible data and of these weighted topologies are given in Luk-Oh Part I, Section 3). Let $(\mathcal{M},g,\phi,F)$ denote the maximal future development, whose future boundary contains the Cauchy horizon $\mathcal{CH}^+$.

Prove, or refute, the following statement of Christodoulou's $H^1$-formulation of strong cosmic censorship (Statement II) for this model: the spacetime $(\mathcal{M},g)$ is future-inextendible as a $W^{1,2}_{\mathrm{loc}}$ Lorentzian manifold, in the joint metric-and-scalar-field sense made precise in the next sentence. In precise form (the version posed in Luk-Oh Remark 1.6): there exists no nontrivial $C^0$ Lorentzian extension $(\tilde{\mathcal{M}},\tilde{g})$ of $(\mathcal{M},g)$ with $\tilde{\mathcal{M}}\setminus\mathcal{M}$ nonempty, such that in the extension the Christoffel symbols of $\tilde{g}$ are locally square-integrable and an extension $\tilde{\phi}$ of the scalar field lies in $W^{1,2}_{\mathrm{loc}}$ near $\mathcal{CH}^+$, so that the Einstein-Maxwell-scalar-field equations admit a weak continuation beyond the Cauchy horizon.

A proof must cover all nontrivial $C^0$ extensions, not only the extension constructed in double-null coordinates by Dafermos and by Luk-Oh, in which the $L^p_{\mathrm{loc}}$ Christoffel blow-up is already known. A refutation must exhibit, for generic data in the above class, a nontrivial $C^0$ extension with $W^{1,2}_{\mathrm{loc}}$ regularity in which the equations hold in the weak sense, and would refute Statement II of strong cosmic censorship in its most natural model.

References

1. Mihalis Dafermos, Black holes without spacelike singularities, Communications in Mathematical Physics 332 (2014) 729-757, DOI 10.1007/s00220-014-2063-4, https://doi.org/10.1007/s00220-014-2063-4 (preprint arXiv:1201.1797, https://arxiv.org/abs/1201.1797; source paper)
2. Jonathan Luk and Sung-Jin Oh, Strong cosmic censorship in spherical symmetry for two-ended asymptotically flat initial data I. The interior of the black hole region, Annals of Mathematics (2) 190 (2019), DOI 10.4007/annals.2019.190.1.1, https://doi.org/10.4007/annals.2019.190.1.1 (preprint arXiv:1702.05715, https://arxiv.org/abs/1702.05715)
3. Jonathan Luk and Sung-Jin Oh, Strong cosmic censorship in spherical symmetry for two-ended asymptotically flat initial data II. The exterior of the black hole region, Annals of PDE 5 (2019), article 6, DOI 10.1007/s40818-019-0062-7, https://doi.org/10.1007/s40818-019-0062-7 (preprint arXiv:1702.05716, https://arxiv.org/abs/1702.05716)
4. Jan Sbierski, On holonomy singularities in general relativity and the $C^{0,1}_{\mathrm{loc}}$-inextendibility of space-times, Duke Mathematical Journal 171 (2022), DOI 10.1215/00127094-2022-0040, https://doi.org/10.1215/00127094-2022-0040 (preprint arXiv:2007.12049, https://arxiv.org/abs/2007.12049)
5. Jan Sbierski, Lipschitz inextendibility of weak null singularities from curvature blow-up, Inventiones Mathematicae 243 (2025) 961-991, DOI 10.1007/s00222-025-01387-0, https://doi.org/10.1007/s00222-025-01387-0 (preprint arXiv:2409.18838, https://arxiv.org/abs/2409.18838)
6. Jonathan Luk, Sung-Jin Oh and Yakov Shlapentokh-Rothman, A scattering theory approach to Cauchy horizon instability and applications to mass inflation, Annales Henri Poincare 24 (2023) 363-411, DOI 10.1007/s00023-022-01216-7, https://doi.org/10.1007/s00023-022-01216-7 (preprint arXiv:2201.12294, https://arxiv.org/abs/2201.12294)
7. Onyx Gautam, Late-time tails and mass inflation for the spherically symmetric Einstein-Maxwell-scalar field system, preprint (2024), arXiv:2412.17927, https://arxiv.org/abs/2412.17927
8. Maxime Van de Moortel, The Strong Cosmic Censorship Conjecture, Comptes Rendus Mecanique 353 (2025), preprint arXiv:2501.13180, https://arxiv.org/abs/2501.13180
9. Peter Cameron and Jan Sbierski, On the uniqueness of continuous spacetime extensions in 1+1 dimensions with applications to weak null singularities, preprint (2025), arXiv:2511.13422, https://arxiv.org/abs/2511.13422
10. Dejan Gajic and Jonathan Luk, The interior of dynamical extremal black holes in spherical symmetry, Pure and Applied Analysis 1 (2019) 263-326, DOI 10.2140/paa.2019.1.263, https://doi.org/10.2140/paa.2019.1.263 (preprint arXiv:1709.09137, https://arxiv.org/abs/1709.09137)
11. Christoph Kehle and Maxime Van de Moortel, Strong cosmic censorship in the presence of matter: the decisive effect of horizon oscillations on the black hole interior geometry, Analysis and PDE 17 (2024) 1501-1592, DOI 10.2140/apde.2024.17.1501, https://doi.org/10.2140/apde.2024.17.1501 (preprint arXiv:2105.04604, https://arxiv.org/abs/2105.04604)
12. Maxime Van de Moortel, The coexistence of null and spacelike singularities inside spherically symmetric black holes, preprint (2025), arXiv:2504.12370, https://arxiv.org/abs/2504.12370
13. Mihalis Dafermos and Jonathan Luk, The interior of dynamical vacuum black holes I: The $C^0$-stability of the Kerr Cauchy horizon, Annals of Mathematics (2) 202 (2025), DOI 10.4007/annals.2025.202.2.1, https://doi.org/10.4007/annals.2025.202.2.1 (preprint arXiv:1710.01722, https://arxiv.org/abs/1710.01722)
14. Sebastian Gurriaran, Non-linear instability of the Kerr Cauchy horizon near $i_+$, preprint (2026), arXiv:2603.17911, https://arxiv.org/abs/2603.17911