In videos of our Galaxy’s center, stars trace noticeable arcs over a few decades and turn near a place where no ordinary star is visible. What can change their motion so strongly? The answer emerged by comparing several different kinds of observations.
The Milky Way’s center contains a supermassive black hole associated with the compact source Sagittarius A, or Sgr A. Its mass is approximately four million solar masses.** A dense star cluster, gas and dust surround it; the entire complex region should not be called a single black hole. NASA: the nuclear star cluster, Chandra: the central environment, 2026.
Dust hides the center from an ordinary view
The Galactic center lies in the direction of Sagittarius, approximately 27,000 light-years away. Interstellar dust between us and it strongly attenuates visible light. A dark patch in an ordinary photograph of that direction is therefore not a picture of the black hole: dust obscuration may cause it.
Longer wavelengths, including infrared and millimeter waves, let us investigate what dust hides in visible light. The discovery therefore combined radio observations of the compact source with infrared measurements of individual stellar motions. The name Sagittarius A* refers to an especially compact source within a much more complex central radio region. ESO: observing Sagittarius A*.
Stars let us weigh an unseen center
Imagine seeing only a satellite orbiting a hidden body. The orbit’s size and period help estimate the mass governing its motion. This does not infer mass from brightness: a massive body affects its neighbor even if it emits no light itself.
Near Sgr A*, astronomers spent years measuring stellar positions on the sky and line-of-sight velocities, determined from spectra. Together, these data allow three-dimensional orbits to be recovered. The star S2 completes an orbit in about 16 years and passes approximately 120 astronomical units from the central object at its closest approach. An astronomical unit is the mean Earth–Sun distance. ESO: S2’s orbit.
The orbits indicate more than a large mass: a very large mass lies within a very small region. Measurements also constrain how much matter can be distributed around the central mass inside S2’s orbit. An ordinary extended cluster does not produce the same gravitational field across all these distances. GRAVITY, 2020: measurements, orbital model and limits on extended mass.
Branch · With formulasOptional · you can skip this
How orbital size and period give a mass
For a teaching estimate, take a Newtonian orbit of a small star around a much more massive compact body. Ignore other bodies and relativistic corrections. If a is the semimajor axis, half the orbit’s long axis, and P its period, then:
M ≈ (4(π)^(2)(a)^(3))/(G(P)^(2)).
Here G is the gravitational constant. In solar masses, astronomical units and years, the same estimate has a convenient form:
(M)/(M_(⊙)) ≈ ((a/au())^(3))/((P/yr())^(2)).
OpenStax: orbital motion and mass.
Choose a deliberately rounded teaching example: a = 1000 astronomical units and P = 16 years. This gives (1000)^(3)/(16)^(2) = 3 906 250, approximately 3.9 million solar masses. It demonstrates the scale; these numbers are not a new precise measurement of Sgr A*.
For the real S2, researchers fit a more complex model, including orbital orientation, distance to the center and relativistic effects. GRAVITY’s 2020 paper reports a period near 16.05 years and mass near 4.26 million solar masses; these results must not be attributed to our rounded substitution. GRAVITY: Table 1 and model parameters.
A black hole has a boundary beyond which light cannot return
The event horizon bounds a region of spacetime from which a signal can no longer reach a distant outside observer. It is not a solid black surface like a planet’s crust. The light we detect near a black hole comes from matter outside the horizon.
Nearby gas can heat up and emit radiation. Its variable emission provides another way to study the surroundings, but a bright spot alone does not locate the horizon. We must know what an instrument measures and which model connects its data to the object. NASA: horizon, accretion and shadow.
Compare the three evidence chains below. Each has its own measurement, inference and limit on what the method can reveal.
Stellar motion
- Measurement
- Stellar positions over years and velocities from spectra.
- Inference
- Mass is concentrated within a small region; motion tests a model of gravity.
- Method limit
- Orbits are not an image of the horizon.
Variable emission
- Measurement
- Brightness and spectrum of matter near the compact source.
- Inference
- Surrounding matter emits; the gas changes state.
- Method limit
- A flare alone is not a picture of the black hole.
Millimeter structure
- Measurement
- EHT interferometric data at a wavelength of about 1.3 mm.
- Inference
- Reconstructions show ring-like emission with a dim center, consistent with a black hole.
- Method limit
- Ring and shadow are not identical to the horizon; this is not visible-light footage.
Schematic icons represent methods, not data. Orbits, emitting region and ring use different arbitrary scales; sizes and colors are not measurements. Read each row from measurement to limitation.
The EHT image reconstructs the structure of emission
In May 2022, the Event Horizon Telescope collaboration released the first image of Sgr A*, derived from radio interferometric observations in April 2017, at a wavelength around 1.3 mm. Telescopes at different locations on Earth jointly measured the signal; reconstruction methods recovered its brightness distribution.
The result shows a bright ring and a dimmer central region, consistent with the expected appearance of emission near a black hole. The shadow is not identical to the horizon itself: its size and appearance also depend on bent light paths and the surrounding gas distribution. The image’s orange color represents radio brightness, not the visible color of matter to a human eye. EHT: original paper, Sections III, V and VII.
Gas changes quickly, and the measurements have limitations. The published image combines reconstructions and captures a robust ring-like feature; it is not an ordinary camera’s instantaneous frame. Particularly compelling is that the ring’s scale agrees with the mass inferred from stellar orbits by a very different method. EHT: what the observations produced.
Branch · DeeperOptional · you can skip this
Even S2’s orbit does not close forever
In an isolated Newtonian two-body problem, a bound elliptical orbit repeats. General relativity predicts a slight shift in the direction of closest approach. Successive loops form a rosette rather than one stationary ellipse.
In 2020, GRAVITY reported measuring this precession of S2, consistent with the relativistic description. A closed ellipse in a teaching diagram is therefore an approximation. This tests stellar motion and a model of gravity; it is not a photograph of the horizon or an observation of the black hole’s interior. GRAVITY: Section 3 and conclusions.
The central black hole is not a drain for the entire Galaxy
With such a mass there, why do all stars not fall into the center? Gravity alone does not remove a body’s sideways motion. On a suitable orbit, a star passes the center rather than moving straight toward it. Far from a black hole, its gravity acts like that of another compact body of the same mass: there is no special suction. NASA: black holes are not cosmic vacuum cleaners.
Test the explanation: does a video of S2’s motion or the EHT ring directly show the black hole’s surface? Neither does. The first shows a star’s motion, from which central mass properties are inferred. The second reconstructs surrounding matter’s emission and a dim central region. Their strength lies in the agreement between different data.
The Milky Way’s center is thus a rich environment of stars, gas and dust, with the supermassive black hole Sgr A* at its heart. We know about it through its effects on neighbors and the light of external matter. To explore how an orbit yields a prediction, revisit predicting motion mathematically.