- The paper establishes that generic smooth initial data for the circularly symmetric Einstein-scalar field avoids naked singularities via a rigorous formulation of weak cosmic censorship.
- It introduces a novel mass gap mechanism in 2+1 dimensions that ensures sub-critical initial mass prevents singularity formation and triggers a blueshift instability.
- The analysis employs precise energy and Hardy-type estimates under asymptotically AdS conditions, linking geometric properties with inevitable black hole formation.
Weak Cosmic Censorship for the Circularly Symmetric Einstein-Scalar Field System in $2+1$ Dimensions
Introduction and Background
The weak cosmic censorship conjecture (WCCC), proposed by Penrose, posits that for generic initial data within reasonable matter models, spacetime singularities arising from gravitational collapse are always hidden within event horizons, i.e., no "naked" singularities form that are visible from null infinity. While in $3+1$ dimensions only special model cases have been resolved, the status for lower dimensional gravity, particularly for the Einstein-scalar field system with negative cosmological constant in $2+1$ dimensions, has remained less developed with no previous rigorous mathematical results in this direction.
The present work rigorously formulates and proves a version of weak cosmic censorship for the circularly symmetric Einstein-scalar field system in $2+1$ dimensions on an asymptotically AdS background (Λ<0). The principal technical innovation is a mass gap mechanism (absent in $3+1$ dimensions), which both precludes generic formation of naked singularities and enables the proof to hold in arbitrarily smooth (Ck, k≥2) function spaces, not merely in low regularity classes.
The analysis focuses on solutions of the Einstein-scalar field equations
Ricμν−21Rgμν+Λgμν=∇μϕ∇νϕ−21gμν∣∇ϕ∣2,□gϕ=0
in $2+1$ dimensions, restricted to circular symmetry. The most relevant geometric backgrounds are the AdS solution ($3+1$0), and members of the BTZ family ($3+1$1), all with reflective boundary conditions at conformal infinity. The Penrose diagram analysis isolates possible global causal structures for black hole formation, non-collapse, local and global naked singularities.
The initial data is posed as a $3+1$2 asymptotically AdS characteristic problem with regular origin. The moduli space of maximal Cauchy developments decomposes, according to the singularity structure and horizon topology, into subsets labeled by collapse, non-collapse, black hole, and (locally) naked singularity configurations.
Main Results
Statement
Theorem (WCCC for Circularly Symmetric $3+1$3 Einstein-Scalar Field System):
For any $3+1$4, the maximal development of generic $3+1$5 asymptotically AdS characteristic data does not contain naked singularities. More precisely, the subset of initial data evolving to naked singularity spacetimes is non-generic (contained in the complement of an open dense set), and moreover is unstable to arbitrarily small perturbations in the data—such perturbations drive evolution towards black hole formation with only spacelike singularities.
This result is attained in smooth norm topologies, in contrast to the $3+1$6 case where the blueshift instability mechanism only works for low regularity data.
Mass Gap and its Consequences
Kinematically, a sharp “mass gap” property is established: if initial data has Hawking mass (Bondi mass at null infinity) less than 1, then no singularities can form in its future development. Thus, infinitesimal scalar perturbations of AdS spacetime remain entirely regular for all future times, a qualitative departure from higher-dimensional AdS. Numerically motivated conjectures for the $3+1$7 setup are thereby resolved.
As a corollary, the mass at the first singularity at the center must approach the critical value $3+1$8 from above, which forces the blueshift effect into a divergent regime along the ingoing cone emanating from the would-be singularity.
A precise criterion for trapped surface formation is formulated: if the scalar field energy and mass in an annular region become sufficiently concentrated (relative to length and cosmological scale), trapped surfaces inevitably develop, leading to black hole regions and hiding the singularity.
For data that is tuned to saturate the critical threshold and thus would produce a (locally) naked singularity, perturbations of arbitrarily high regularity supported away from the origin can be constructed such that they leave the past of the singularity unchanged but, due to divergent blueshift, become dynamically amplified arbitrarily close to the singular region. This amplification enables one to invoke the trapped surface criterion, converting the evolution from a locally naked singularity to one in which any such singularity is cloaked by a horizon.
The proof employs detailed a priori bounds for the scalar field and geometric variables in carefully constructed "blueshift-dominated" neighborhoods of the singularity, and propagates smoothness through energy and Hardy-type estimates adapted to asymptotically AdS settings.
Genericity and Structure of the Moduli Space
The WCCC is established in terms of the complement of an open and dense subset of the moduli space of initial data—the set of naked singularity data is nowhere dense and can be approached only by sequences of data in the black hole or regular classes. The decomposition of the moduli space is refined to classify all possible boundary behaviors in the Penrose diagrams, in correspondence with the recent systematic works on moduli space topology for spherically symmetric collapse (Angelopoulos et al., 11 Mar 2026).
Numerical and Theoretical Implications
The proof analytically confirms several phenomena previously observed in numerical studies, including the stability of the $3+1$9 AdS spacetime under arbitrarily small smooth perturbations, the formation of a mass gap, and the blueshift instability as a universal mechanism for removing naked singularities from the generic configuration space.
The presence of the mass gap in $2+1$0 dimensions, contrasted sharply with its absence in $2+1$1, reflects a fundamental difference in the critical behavior of scalar evolution in lower-dimensional gravity. It enables a direct activation of the blueshift-mediated instability across all smooth function spaces, and also provides strong uniform regularity controls that are unavailable in $2+1$2 collapse.
The techniques developed, particularly the precise energy estimates near infinity and the handling of global characteristic data in the presence of a timelike conformal boundary, open up new directions both for rigorous studies of critical phenomena in lower-dimensional numerical relativity and for exploring analogous mechanisms in higher-dimensional gravity, where mass gap behavior could play subtler roles.
Future developments include (i) explicit construction of critical data sets realizing the threshold between non-collapse and black hole formation [in preparation by the author and collaborators], (ii) extension of these arguments to non-symmetric perturbations or more exotic matter models, and (iii) systematic classification of the precise codimensions of the "naked" sets in moduli space.
Conclusion
This work provides the first rigorous proof of weak cosmic censorship in the setting of the circularly symmetric $2+1$3 Einstein-scalar field system with negative cosmological constant. The analysis shows that, generically, singularities formed in gravitational collapse are hidden within horizons, and that locally naked singularities cannot persist under $2+1$4 perturbations, due to a robust blueshift-amplification mechanism. The methods employed demonstrate the decisive role of the mass gap in lower-dimensional AdS gravity and chart a path for further mathematical study of gravitational critical phenomena and the global structure of solutions in both lower and higher dimensions.
Reference: "Weak cosmic censorship for the circularly symmetric Einstein-scalar field system in $2+1$5 dimensions" (2605.19143)