Explosive, Thunderbolt, and Bounce Scenarios

Endpoint scenarios share a disciplined split: semiclassical evolution may determine the incoming state, fluxes, conserved charges, apparent-horizon motion, and the first failed control parameter, but a final explosion, null thunderbolt, bounce, or topology change usually requires extra ultraviolet data. The decisive question is not whether a causal diagram can be drawn; it is which segment follows from controlled equations and which segment is imposed.

Helpful background. Quantum Fields in Curved Spacetime supplies the Hawking state; Einstein Equations supplies local conservation; Black-Hole Geometry and Mechanics supplies horizons and singularity distinctions; Semiclassical Gravity supplies backreaction control; Black-Hole Information treats radiation correlations; and Quantum Gravity studies microscopic singularity resolution and topology change.

A common endpoint decomposition

Let $\lambda$ measure evolution along a chosen foliation and let $\epsilon_i(\lambda)$ denote curvature, fluctuation, adiabatic, frequency, and cutoff diagnostics. Define the last controlled slice $\Sigma_*$ by

$$\max_i\epsilon_i(\Sigma_*)=\epsilon_{\rm max}\ll1,$$

with the next evolution step expected to violate the declared tolerance. Semiclassical data on $\Sigma_*$ include the induced geometry, extrinsic curvature, renormalized quantum state or its correlators, conserved charges, outgoing flux, and uncertainty from omitted terms. An endpoint proposal adds a continuation map

$$\mathcal C_{\rm UV}: \bigl(\Sigma_*,g,K,\rho,\text{charges}\bigr) \longrightarrow \text{future geometry and state}.$$

Here $K_{ij}=h_i{}^ah_j{}^b\nabla_an_b$ for the future unit normal to the spacelike matching slice. If $\mathcal C_{\rm UV}$ is not derived within the retained theory, its consequences are assumptions of a completion. Conservation and matching can constrain the map without selecting it uniquely.

Final bursts and null thunderbolts

A final burst extrapolates increasing Hawking temperature and luminosity into a short release. The controlled part is the flux history while curvature, state construction, and adiabaticity remain valid. The burst energy cannot exceed the remaining Bondi mass and must respect charge and angular-momentum balance. Its spectrum, duration, and correlations are not fixed by continuing a divergent luminosity formula past its cutoff.

A “thunderbolt” denotes a singular outgoing null feature, not merely a bright but finite burst. It should also be distinguished from the finite endpoint pulse often called a “thunderpop” in the counterterm-modified RST model. There the pulse accompanies a chosen matching to the linear-dilaton vacuum; its integrated energy and sign follow from that boundary prescription Russo, Susskind, and Thorlacius 1992, § 3, PDF.

The earlier CGHS large-$N$ equations supply a collapse and leading-backreaction framework Callan et al. 1992, Eqs. (1)–(23), PDF. An early one-loop analysis of those equations develops a curvature singularity and loses control before complete evaporation Russo, Susskind, and Thorlacius 1992, §§ II–IV, PDF. This is distinct from the later exactly solvable RST modification. None of these results is a four-dimensional endpoint prediction; the action and boundary data remain part of each result.

A null singularity in a model is not automatically a calculable pulse seen at infinity. One must distinguish divergence of curvature, failure of the renormalized stress tensor, a distributional shell with finite integrated energy, and a boundary beyond which the equations are undefined.

Bounces and topology-changing continuations

A bounce replaces a spacelike singular region by a transition to expansion, a white-hole-like region, or another asymptotic branch. A regular-core model can impose a minimum areal radius and match an outer evaporating geometry to an inner nonsingular metric. Bardeen develops such a scenario and states its nonsingular-core premises in Bardeen 2014, §§ II–V, PDF.

The semiclassical checks are stringent. The effective source must be conserved; its state and renormalization prescription must exist; curvature and stress fluctuations must remain below cutoff; junction conditions must not hide an unaccounted shell; and inner horizons must be tested for blueshift instability. If the bounce occurs precisely where these tests fail, the bounce is a proposed ultraviolet completion rather than an output of semiclassical evolution.

Topology change is stronger still. A single smooth globally hyperbolic mean metric generally cannot interpolate between arbitrary spatial topologies without singularity, causal pathology, or changed assumptions. Summing over geometries, superposing causal structures, or replacing manifold locality is a Quantum Gravity question. Semiclassical theory can pass forward boundary data and conservation constraints, not choose the microscopic transition.

Results and open questions through 2026-08-19

The evidence through 2026-08-19 demonstrates controlled partial evolutions and model-dependent breakdowns. No acquired calculation determines a universal four-dimensional endpoint.

Scenario or evidence	Controlled result	Uncontrolled boundary or limitation
Bardeen 2014, § I, Eqs. (1.2)–(1.4), pp. 2–3, and § III, pp. 13–14, PDF	Flux and Bondi-mass decrease while the Unruh-like state and adiabatic hierarchy hold	Final spectrum and duration once curvature or fluctuations reach cutoff
Callan et al. 1992, Eqs. (1)–(23), PDF	Two-dimensional collapse, Hawking flux, and leading large-$N$ backreaction	Model-specific endpoint and missing four-dimensional transverse dynamics
Russo, Susskind, and Thorlacius 1992, §§ II–IV, PDF	The one-loop equations exhibit a curvature singularity; their large-mass backreaction regime loses validity as the receding apparent horizon approaches it	It is not the later exactly solvable RST model, and it does not determine complete evaporation
Bardeen 2014, §§ II–V, PDF	Consistent kinematic regular-core scenario and matching questions	Core stress, state, and quantum-gravity singularity avoidance are assumed

The conservative position stops the spacetime prediction at $\Sigma_*$ and reports only boundary data. The competing scenario-building position chooses a constrained continuation and studies its consequences. Both are useful if labels remain explicit; they conflict only when an imposed map is advertised as a derived endpoint. The relevant comparison axes are dimension, state, counterterm, inner-boundary condition, topology, and control parameter.

Discriminating checks and open questions

Energy check. Integrate the proposed outgoing pulse at future null infinity. Its energy plus the final remnant or interior energy must equal the mass remaining on $\Sigma_*$ within flux uncertainty. A formally infinite luminosity with finite remaining energy requires a distributional interpretation and a regulator.

Matching check. Across a non-null transition surface, compare the induced metric and extrinsic curvature using a common normal convention. A jump in the latter requires a surface stress tensor in Einstein gravity. A null shell instead requires the null junction conditions and transverse-curvature data, because its induced metric is degenerate. Treating a null pulse with the ordinary spacelike matching formula would miss this distinction.

Model-translation counterexample. CGHS and RST share two-dimensional ancestry yet differ in action and counterterm structure and reach different late behavior. Their contrast is direct evidence that endpoint diagrams are sensitive to formulation, not universal consequences of Hawking flux.

The frontier remains unresolved for controlled four-dimensional evolution through the first cutoff. The conclusions could change after an independent four-dimensional backreaction solution reaches $\Sigma_*$ with quantified errors, an ultraviolet-complete calculation derives a continuation map, or a theorem excludes a class of bounces under stated energy and causality assumptions; new evidence resolving the inner-horizon instability or final-pulse spectrum could also change the conclusion.

Exercise

A model evolves reliably to mass $M_*$, then postulates a null pulse of energy $0.8M_*$ and a remnant of mass $0.3M_*$. What fails, and how could the proposal be repaired?

Solution

The proposed future carries $1.1M_*$ before accounting for any other flux or binding convention, violating the remaining energy balance by $0.1M_*$. The model must reduce one contribution, identify a negative-energy channel with a controlled source, or revise the matching definition of mass. Conservation can reject the continuation even though it cannot select a unique valid one.

References

• Bardeen, J. M. (2014). “Black Hole Evaporation without an Event Horizon.” arXiv:1406.4098. Open PDF.
• Callan, C. G., Jr., Giddings, S. B., Harvey, J. A., and Strominger, A. (1992). “Evanescent Black Holes.” Physical Review D 45, R1005–R1009. Open PDF.
• Russo, J. G., Susskind, L., and Thorlacius, L. (1992). “Black Hole Evaporation in 1+1 Dimensions.” Physics Letters B 292, 13–18. DOI. Open PDF.
• Russo, J. G., Susskind, L., and Thorlacius, L. (1992). “End Point of Hawking Radiation.” Physical Review D 46, 3444–3449. DOI. Open PDF.