First Sagittarius A* Event Horizon Telescope Results. I. The Shadow of the Supermassive Black Hole in the Center of the Milky Way

This paper presents the Event Horizon Telescope (EHT) Collaboration’s first horizon-scale images of Sagittarius A* (Sgr A*), the compact radio source at the center of the Milky Way. It investigates whether the source’s appearance matches the shadow and surrounding emission expected from a supermassive black hole, whether its size agrees with independent stellar-orbit measurements, and what the observations reveal about accretion and general relativity. Previous measurements indicated a mass of approximately four million solar masses and a distance of roughly 8 kiloparsecs, predicting a shadow about 50 microarcseconds across. The observations were conducted in April 2017 using eight observatories at six locations, operating together as an Earth-sized very-long-baseline interferometer at a wavelength of 1.3 millimeters. The analysis focuses primarily on April 7, which provided the best baseline coverage, while April 6 data support validation and selected joint analyses. The array achieved an angular resolution of approximately 24 microarcseconds. Two independent processing pipelines calibrated the data, and several complementary imaging and geometric-modeling methods reconstructed the source. Interstellar scattering and Sgr A*’s variability on timescales of minutes required specialized analysis. The researchers incorporated statistical descriptions of structural variability, compared reconstructions with and without scattering corrections, and tested their methods on synthetic observations of both ring-like and non-ring sources. The principal result is a bright, thick emission ring surrounding a central brightness depression. The ring has a measured diameter of 51.8 ± 2.3 microarcseconds and a fractional width of approximately 30–50%. Quoted measurement uncertainties represent 68% credible intervals. More than 95% of the selected imaging reconstructions show a ring, and independent geometric modeling also favors ring morphologies. However, the ring’s orientation and detailed brightness distribution remain poorly constrained. Total-flux variations within individual nights were approximately 4–13%, and the interferometric data also reveal intrinsic structural variability. Despite detecting this variability, the observations cannot reliably establish how the image evolves over time. To interpret the observations, the collaboration compared them with an extensive library of general relativistic magnetohydrodynamic simulations, including 200 fiducial models. These models explored black hole spin, viewing inclination, magnetic-field strength, and electron-temperature prescriptions. They were evaluated against EHT measurements, emission at other wavelengths, and variability constraints. The comparisons generally favor prograde rotation, relatively low viewing inclinations, dynamically strong magnetic fields, and substantially different proton and electron temperatures. Two models satisfy all constraints except variability. Both feature magnetically arrested flows viewed at an inclination of 30 degrees, with dimensionless spins of 0.5 and 0.94. They suggest accretion rates of approximately one billionth to one hundred-millionth of a solar mass per year. Crucially, none of the fiducial models satisfies all 11 observational constraints, largely because the simulations typically predict greater variability than is observed. Calibrating the emission ring against simulated observations yields an inferred shadow diameter of 48.7 ± 7.0 microarcseconds and an EHT-based mass estimate of 4.0 million solar masses, with uncertainties of −0.6 and +1.1 million solar masses. These results agree with stellar-orbit measurements and the expected appearance of a Kerr black hole. Together with observations of M87*, they support general relativity across more than three orders of magnitude in black hole mass. They also disfavor certain horizonless alternatives, although they do not exclude all black hole mimickers. The authors emphasize that sparse baseline coverage, rapid variability, uncertain electron physics, and model-dependent calibration limit stronger conclusions about spin, orientation, and gravity. Improved observations, polarization measurements, and more realistic plasma modeling are needed to address these limitations.

Imaging the Black Hole at the Heart of the Milky Way

At the center of our galaxy lies Sagittarius A*, or Sgr A*, a compact radio source associated with a black hole about four million times the mass of the Sun. The Event Horizon Telescope Collaboration has now revealed its appearance on the scale of its event horizon: a bright, thick ring surrounding a darker center.

The paper, “First Sagittarius A* Event Horizon Telescope Results. I. The Shadow of the Supermassive Black Hole in the Center of the Milky Way,” explains how that image was reconstructed and what it reveals. Its central finding is striking: the observed structure has the size and appearance expected for a supermassive black hole. Yet the results also expose important gaps in our understanding of the turbulent plasma surrounding it.

Building an Earth-Sized Telescope

Sgr A* is roughly 8 kiloparsecs from Earth. Earlier measurements tracking stars around the galactic center established its approximate mass and distance, predicting a black hole shadow about 50 microarcseconds across—an extraordinarily small angle on the sky.

To resolve that scale, the EHT linked eight observatories at six locations in April 2017. Operating at a wavelength of 1.3 millimeters, they formed an Earth-sized very-long-baseline interferometer with an angular resolution of approximately 24 microarcseconds.

The study focuses primarily on April 7, when the array provided its best coverage of spatial scales and directions. April 6 observations supported validation and selected joint analyses. Two independent processing pipelines calibrated the measurements, while several imaging and geometric-modeling methods reconstructed the source.

A Target That Changes During the Observation

Sgr A* presented two major obstacles. Interstellar material scatters its radio emission, blurring the view. Meanwhile, the emitting material close to the black hole changes on timescales of minutes. The telescope therefore had to assemble information about a source that would not sit still.

The researchers incorporated statistical descriptions of structural variability, compared reconstructions with and without scattering corrections, and tested their methods using synthetic observations of both ring-like and non-ring sources. These checks helped establish whether the ring was supported by the data rather than imposed by the reconstruction process.

What the Ring Reveals—and What It Does Not

The principal result is a bright emission ring with a diameter of 51.8 ± 2.3 microarcseconds. Its fractional width is approximately 30–50%, making it a thick structure rather than a narrow outline. The quoted uncertainties represent 68% credible intervals.

More than 95% of the selected imaging reconstructions display a ring, and independent geometric modeling also favors ring-like shapes. That agreement makes the overall morphology a robust result. The ring’s orientation and detailed brightness pattern, however, remain poorly constrained.

The bright ring should not be confused with the event horizon itself. It is radiation from hot material near the black hole, shaped by strong gravity. The central brightness depression is associated with the expected black hole shadow, but translating the measured emission ring into a shadow diameter requires modeling.

The observations also detected changes in the source. Total flux varied by approximately 4–13% within individual nights, and the interferometric measurements revealed intrinsic structural variability. Nevertheless, the data cannot reliably show how the image evolves over time. This is evidence of a changing source, not a dependable movie of its motion.

Testing Models of the Surrounding Plasma

To interpret the image, the collaboration compared the observations with an extensive library of general relativistic magnetohydrodynamic simulations, including 200 fiducial models. These explored black hole spin, viewing inclination, magnetic-field strength, and prescriptions for electron temperatures.

The comparisons generally favored:

• Prograde rotation and relatively low viewing inclinations.
• Magnetic fields strong enough to influence the accretion flow dynamically.
• Substantially different temperatures for protons and electrons.

Two models met every constraint except variability. Both involved magnetically arrested flows viewed at an inclination of 30 degrees, with dimensionless spins of 0.5 and 0.94. They suggested accretion rates between approximately one billionth and one hundred-millionth of a solar mass per year.

Crucially, none of the fiducial models satisfied all 11 observational constraints. The simulations generally predicted more variability than the observations showed. These preferences therefore offer clues about the plasma, not definitive measurements of the black hole’s spin or orientation.

A Powerful Consistency Check for General Relativity

Calibrating the emission ring against simulated observations yielded an inferred shadow diameter of 48.7 ± 7.0 microarcseconds. The corresponding EHT-based mass estimate was 4.0 million solar masses, with uncertainties of −0.6 and +1.1 million solar masses—consistent with independent stellar-orbit measurements.

The results match the expected appearance of a Kerr black hole. Together with EHT observations of M87*, they support general relativity across more than three orders of magnitude in black hole mass. They also disfavor certain horizonless alternatives, although they do not eliminate every possible black hole mimicker.

The next steps are clear: improved observations, polarization measurements, and more realistic plasma models. Sparse telescope coverage, rapid variability, uncertain electron physics, and model-dependent calibration still limit stronger conclusions. The ring is compelling evidence, but understanding everything happening around it remains an open scientific challenge.