Evidence for Event Horizons Versus Material Surfaces
Scrollable formula. Use Left and Right arrow keys to reveal hidden content.
Scrollable table. Use Left and Right arrow keys to reveal additional columns.
Scrollable figure. Use Left and Right arrow keys to inspect hidden content.
Observations strongly support astrophysical objects with extremely compact, Kerr-compatible exteriors and highly absorbing inner boundaries. They also exclude concrete alternatives: for Sgr A*, a steady thermalizing surface supplied by the inferred accretion flow would be too bright, and a fully reflecting surface is difficult to reconcile with the horizon-scale data. At the evidence cutoff of 2026-08-20, this is substantial support for the black-hole interpretation. It is not a local observation of the globally defined event horizon, nor a model-independent exclusion of every delayed, weakly reflecting, nonthermal, or high-heat-capacity surface.
Helpful background. Horizon and Surface Evidence explains gravitational redshift, surface-emission bounds, image depressions, and the logical distinction between horizons and material surfaces. This record weighs the current combined evidence and exact EHT products.
The claim an observation can actually test
Section titled “The claim an observation can actually test”Four questions are often compressed into “does it have a horizon?” Geometry may mean a Kerr-like exterior, a photon-ring-scale lensing structure, a local trapped surface, or the global event horizon. A material boundary may thermalize promptly, reflect, absorb and store energy, or reprocess it nonthermally. The observable may be a millimeter visibility, infrared spectrum, accretion efficiency, quiescent luminosity, tidal response, or ringdown echo. Each also carries nuisance assumptions about accretion rate, source age, inclination, plasma, extinction, heat capacity, and response time.
Finite-time electromagnetic data probe exterior propagation and a response function. An event horizon is defined using the entire future causal structure. The strongest empirically meaningful proposition is therefore that the data favor an extremely compact, absorbing Kerr-compatible exterior and disfavor named surface responses. Stronger language requires additional causal and astrophysical assumptions.
| Assessment at 2026-08-20 | Meaning |
|---|---|
| Question status | Partially resolved |
| Evidence status | Supported |
| Confidence | Moderate because the tested surface classes are constrained strongly but the universal/global claim is not observable directly |
| Domain | Sgr A* horizon-scale imaging and broadband surface-emission tests, interpreted with declared accretion and response models; supporting exterior evidence from black-hole phenomenology |
| What could change the conclusion | A new surface-sensitive observation, viable countermodel, or systematic reanalysis |
Entry project: reconstruct a thermal-surface exclusion
Section titled “Entry project: reconstruct a thermal-surface exclusion”Choose this topic if you like radiative physics and want an analytic project tied to observations. You need blackbody spectra, gravitational redshift, luminosity-distance conversions, and uncertainty propagation. The first calculation needs a laptop and published spectral limits; a new accretion-flow or interferometric analysis requires more specialized support.
Start with Broderick and Narayan 2006, Eqs. (1)–(4) and Fig. 1. Derive the luminosity seen at infinity and the apparent emitting area under the paper’s steady, thermalizing-surface assumptions. Translate radius and redshift conventions explicitly. Next read EHT Collaboration 2022, Paper VI, §4.1 and Fig. 14 to identify the updated size and spectral inputs. The official EHT multiwavelength release 2022-D02-03 provides an input anchor; record which measurements and extinction treatment actually enter your calculation.
The finite deliverable is a reproducible predicted spectrum and an exclusion plot for the same thermal response, with one controlled variation of accretion luminosity and one of apparent radius. Reproduce a baseline curve or bound before broadening the surface model. If only values extracted from a figure are available, retain their extraction uncertainty and describe the exercise accordingly. It is a training reproduction of a model-dependent argument, not an independent demonstration of an event horizon.
A candidate extension, with novelty unchecked, is to introduce a single finite thermal-relaxation time and determine where the steady-state bound ceases to apply for a declared accretion history. Seek supervision on energy conservation and source-age assumptions. Stop if the proposed storage or reradiation prescription has no closed energy budget, or if uncertain extinction dominates the claimed improvement. A useful pivot is a map of which response assumptions the observation actually constrains.
What surface-emission and imaging tests establish
Section titled “What surface-emission and imaging tests establish”The cleanest thermodynamic argument predates the image. If gas reaches a material surface, its binding energy must eventually emerge unless it is stored or otherwise removed. Broderick and Narayan combine an accretion estimate, surface redshift, and infrared limits to show that a compact thermalizing surface for Sgr A* would overproduce emission Broderick and Narayan 2006, Eqs. (1)–(4), Fig. 1, and conclusion. The result is powerful precisely because its assumptions are explicit: approximate steady state, thermal reradiation, and enough time to equilibrate.
The Event Horizon Telescope adds a direct size scale and calibrated millimeter data. Paper VI combines the observed source size and broadband spectral constraints to reject the tested thermal-surface models and to make a fully reflective surface unlikely EHT Collaboration 2022, Paper VI, §§ 4.1–4.2 and Fig. 14. Paper I establishes the ring-scale morphology and central brightness depression through several imaging pipelines EHT Collaboration 2022, Paper I, §§ 5–7 and Figs. 3–6. Together they favor a compact absorbing object; neither converts the image center into a photograph of a causal boundary.
Results and open questions at 2026-08-20
Section titled “Results and open questions at 2026-08-20”| Claim or evidence | What it supports | What it does not establish |
|---|---|---|
| Infrared thermal-surface bound | Exclusion of steady, promptly thermalizing surfaces under the inferred accretion supply | Arbitrary long-term storage or nonthermal response |
| Sgr A* visibility size and spectrum | Stronger joint constraint on tested thermal and reflecting models | Global event-horizon existence |
| Multi-pipeline ring morphology | Compact Kerr-compatible exterior and lensing-scale structure | Unique spacetime without plasma assumptions |
| Horizonless-object model survey | A finite taxonomy of alternative response signatures | Evidence that a particular alternative forms or is stable |
| 2025 imaging review | Imaging-alone scope limit and discriminants for future arrays | New observational exclusion by itself |
| Exact public EHT products | Inspectable visibility and multiwavelength inputs | Complete public packaging of every GRMHD library and likelihood choice |
The thermal and reflecting constraints are direct tests of specified alternatives and receive the strongest weight. The image morphology corroborates compactness and exterior geometry, while the reviews control how far those results can be generalized.
Strongest competing interpretations
Section titled “Strongest competing interpretations”The black-hole interpretation is economical: an absorbing inner boundary, Kerr-compatible exterior, and accretion plasma account for the absence of surface light and the observed ring-scale structure without adding a response law. Independent stellar dynamics fixes the mass and distance well enough that the image scale becomes a meaningful strong-field test.
The strongest alternative is not a generic “hard surface.” It is an ultracompact object whose response is sufficiently delayed, absorptive, frequency dependent, or nonthermal that prompt infrared and millimeter bounds do not apply. Cardoso and Pani organize such horizonless compact objects and their potential electromagnetic and gravitational signatures, while emphasizing formation, stability, and timescale constraints Cardoso and Pani 2019, §§ 2–4 and 6–7. Such models preserve a logical loophole, but an unspecified response function is not positive evidence.
Counterexamples, null results, and failed routes
Section titled “Counterexamples, null results, and failed routes”A dark image center is not a counterexample to surfaces: optically thin emission, lensing, absorption, scattering, variability, and reconstruction can all create or alter a central depression. Therefore the failed route is to infer the global event horizon from morphology alone. Ayzenberg and collaborators explicitly conclude that certainty about horizon existence is virtually impossible from imaging alone and explain how horizonless spacetimes can retain central depressions Ayzenberg et al. 2025, § 5.2.1.
Conversely, a hypothetical surface with arbitrary storage time can evade a thermal bound by construction. That is not a refutation of the bound; it exits its steady-reradiation domain. A viable counterexample must specify microphysics, source age, heat capacity, reflectivity, and escape time, then fit the visibility, spectrum, and variability without tuning a new response for each dataset.
Combining observables without double counting
Section titled “Combining observables without double counting”Surface luminosity and EHT morphology share astrophysical inputs, especially accretion flow and inclination. They should not be multiplied as independent likelihoods unless their correlations are represented. The robust combination is qualitative and conditional: dynamics fixes the compact mass; broadband limits disfavor prompt reradiation; interferometry fixes an emitting/lensing scale; polarization and variability constrain plasma models. Each narrows the allowed exterior-plus-response space.
The global-versus-local distinction remains even after combination. Under cosmic censorship, long-lived stationarity, and ordinary matter assumptions, an absorbing Kerr exterior is naturally interpreted as a black hole. The observations strongly support that package. The data alone do not reconstruct the entire future spacetime required by the formal event-horizon definition.
What would change the assessment
Section titled “What would change the assessment”The most informative reanalysis would compare Kerr absorption with a preregistered surface family spanning thermal, reflecting, partially absorbing, delayed, and nonthermal response. It should use the public Sgr A* visibilities and spectral compilation, marginalize plasma and extinction models, and report constraints as functions of surface redshift and response time rather than only for two extreme cases.
A surface model that predicts the ring-scale visibilities, broadband spectrum, and variability without post-hoc tuning—and has a credible formation and stability route—would lower support for the absorbing-black-hole interpretation. Stronger multi-epoch or multi-frequency bounds that exclude broad response families would raise confidence, while still not converting a finite-time measurement into a theorem about the global boundary.
Research checklist
Section titled “Research checklist”- State whether the target claim concerns exterior geometry, absorption, a local trapped region, or the global event horizon.
- Parameterize surface radius/redshift, emissivity, reflectivity, heat capacity, delay, and nonthermal response; do not use “surface” as one model.
- Pin EHT visibility and multiwavelength products, calibration tables, imaging/forward-model code, GRMHD library, extinction, and priors.
- Vary accretion rate, inclination, electron thermodynamics, source age, and equilibration time.
- Perform prior-predictive and injection checks for Kerr and surface models; report likelihood correlations across datasets.
- Separate exclusion of a named response from a universal statement about all horizonless objects.
- Publish posterior constraints in physical response coordinates, not only a model Bayes factor.
- No independent end-to-end reproduction was performed for this assessment; the public inputs below support the stated bounded reanalysis.
Source and data map
Section titled “Source and data map”Broderick and Narayan supply the foundational thermal-emission argument. EHT Paper VI is the primary surface-sensitive observation, while Paper I supplies imaging context. Cardoso and Pani provide the structured alternative-model taxonomy; Ayzenberg et al. provide the current imaging scope and future discriminants.
The exact public Sgr A* products are EHT 2022-D02-01 DOI 10.25739/m140-ct59, 2022-D02-03 DOI 10.25739/26fq-k306, and the complete 2017 level-1 release 2022-D01-01 DOI 10.25739/kat4-na03. They permit visibility and declared spectral-bound checks, but no complete Paper VI likelihood with every GRMHD and extinction choice was verified. Official inputs are not an independent scientific analysis.
References
Section titled “References”- Ayzenberg, D., et al. “Fundamental physics opportunities with the next-generation Event Horizon Telescope.” Living Reviews in Relativity 28, 1 (2025). DOI.
- Broderick, A. E. and Narayan, R. “On the nature of the compact dark mass at the Galactic Center.” Astrophysical Journal Letters 638, L21–L24 (2006). DOI.
- Cardoso, V. and Pani, P. “Testing the nature of dark compact objects: a status report.” Living Reviews in Relativity 22, 4 (2019). DOI.
- EHT Multiwavelength Science Working Group. “Multi-wavelength Observations of Sgr A* During the 2017 EHT Campaign.” Data product 2022-D02-03 (2022). DOI.
- Event Horizon Telescope Collaboration. “First Sagittarius A* Event Horizon Telescope results. I. The shadow of the supermassive black hole in the center of the Milky Way.” Astrophysical Journal Letters 930, L12 (2022). DOI.
- Event Horizon Telescope Collaboration. “First Sagittarius A* Event Horizon Telescope results. VI. Testing the black hole metric.” Astrophysical Journal Letters 930, L17 (2022). DOI.