Star Camouflage: How Boson Stars Masquerade as Black Holes ✦
The paper considers the hypothesis that the compact source at the Galactic Center, Sagittarius A*, is a massive boson star. To test this, astrometric data from infrared flares were fitted: 12 discrete boson star configurations were considered, with each flare modeled as a hot spot on a circular equatorial orbit. The analysis was done using a Bayesian approach with nested sampling, yielding marginalized posterior parameter distributions and Bayesian evidence for each model. The same procedure was applied to a Schwarzschild black hole. The Bayesian evidence for the boson star and black hole differ insignificantly, and the mass of Sgr A* (~4.296×10⁶ M⊙) falls within the 68% highest-density interval for all configurations. The conclusion is that under current near-infrared astrometric constraints and within the parameter ranges considered, a massive boson star and a Schwarzschild black hole are statistically indistinguishable.
Context
Testing the nature of supermassive compact objects in galactic nuclei is a key test of general relativity in the strong-field regime. Black holes predict an event horizon and singularity, but alternative models such as boson stars (giant gravitationally bound condensates of scalar particles) lack these features. Their compactness can rival that of neutron stars, making them realistic black hole mimics. Distinguishing them observationally is extremely difficult, especially for distant objects. The center of our Galaxy, 27,000 light-years away, offers a unique opportunity: the object Sagittarius A*, with a mass of about 4.3 million solar masses, is close enough to measure the motion of individual stars and hot spots using infrared interferometry with the GRAVITY instrument. If it turns out not to be a black hole, this would point to new physics and could link it to dark matter, as boson stars are natural candidates for dark matter made of light scalar fields such as axions.
Methods
The study considered 12 boson star configurations obtained by numerically solving the Einstein–Klein–Gordon equations. Each model was defined by two parameters: the self-interaction constant Λ and the central field amplitude ψ_0. The Schwarzschild solution was used for comparison. Flare emission was modeled as a Gaussian hot spot with radius 0.3 mass units moving on a circular equatorial orbit. Using ray tracing in curved spacetime, images were computed for various orbital radii and inclination angles. Then the luminosity centroid—the analog of the interferometrically measured position—was extracted from each image. For each (orbital radius, inclination) pair, the centroid trajectory was precomputed, allowing efficient likelihood construction based on photometric data. The analysis employed nested sampling with the Dynesty package. The fitting parameters were orbital radius, inclination, position angle, initial spot longitude, and object mass.
Results
The Bayesian evidence (log Z) for all 12 boson star models turned out virtually identical to that of the Schwarzschild black hole: the maximum difference was only 0.16, which on the Jeffreys scale indicates weak distinction, failing to favor any model. The object's mass fell within the 68% highest density intervals (HDI) for all configurations, and its peak values matched independent measurements from S-star orbits: (4.297 ± 0.012) × 10^6 M_⊙. However, mass distributions for boson stars were systematically shifted toward larger values and more asymmetric than for the black hole. This is explained by 'shooting-through' photons—rays that pass through the star without meeting an event horizon. They shift the centroid toward the center, and to compensate the model slightly overestimates the mass. The orbital orientation parameters (inclination and position angle) remained weakly constrained, but their credible intervals overlap for all models.
Implications
The results show that, within existing infrared astrometric data, boson stars remain just as probable candidates for the central object of the Galaxy as classical black holes. This underscores the need for stricter tests of the event horizon and paves the way for searching for alternative models. If future gravitational-wave or polarimetric observations reveal anomalies, this could be the first direct hint of exotic compact objects and point to a connection with dark matter, first suspected by Fritz Zwicky. The work also illustrates the power of the Bayesian approach in astrophysics: even without formally singling out one model, one can quantitatively assess their degree of equivalence.
Future development
Future plans include extending the analysis to the full continuous parameter range (Λ, ψ_0) and incorporating rotating boson stars. This may break some degeneracies and reveal distinctive features in images—for example, additional asymmetry. It is particularly important to include polarimetric data, which promise to significantly improve the determination of inclination and position angles. More detailed modeling of accretion flows and the use of machine learning for rapid comparison of many configurations are also promising.
Impact
The work addresses fundamental issues of gravity in the strong-field regime, dark matter physics, stellar dynamics in galactic nuclei, and very-long-baseline interferometry methods.
Next steps
The next step is a systematic scan of the entire two-dimensional parameter space of boson stars and a joint analysis of astrometry and polarimetry, which will narrow down the allowed models.
Key open problems
The study is directly connected to unsolved problems in modern physics: the black hole information paradox, resolution of singularities, and the nature of dark matter. Horizonless objects like boson stars could resolve these difficulties by providing an alternative to classical black holes, a concept refined by scientists such as John Wheeler.
🎯 Boson stars are so dense that their compactness (mass-to-radius ratio) can approach that of black holes: the effective radius of such a star can be just 2.81 mass units, only slightly larger than the gravitational radius of a black hole. Interestingly, photons that pass through the star retain information about its internal structure, unlike in a black hole where everything is hidden behind the horizon.
Key numbers
- Sgr A* mass: ~4.3×10^6 M⊙
- Distance to Galactic center: ~27,000 light-years
- Flare orbital period: ~1 hour
- Number of boson star configurations: 12
- Max log Bayes factor difference: 0.16