A three-dimensional reconstruction of the eruption around the black hole at the center of our galaxy has been made

Now, for the first time, we can see what an eruption around the Milky Way’s supermassive black hole looks like in three dimensions. The recent research, created with the cooperation of astronomers and computer scientists, has reconstructed with artificial intelligence (AI) the flash of matter circulating around the object at a dizzying speed, which contributes to the understanding of the extreme processes around black holes.

Sagittarius A* (Sgr A*), located roughly 27,000 light-years from Earth at the center of the Milky Way, is not considered a giant among supermassive black holes, but its mass is still about four million times greater than that of the Sun. The heated gas orbiting Sgr A* is moving at nearly the speed of light, forming a huge ring called an accretion disk—which can be seen in orange in the Event Horizon Telescope collaboration’s 2022 image around the shadow of the black hole:

The central black hole of the Milky Way, Sagittarius A*

Photo: Event Horizon Telescope collaboration

“This special environment produces high-energy bursts called flares that can be observed in the X-ray, infrared and radio regions,” Aviad Levis, a science imaging researcher at the California Institute of Technology (Caltech) and the study published Monday in the journal Nature Astronomy, told Qubit one of its authors. According to the expert, the most important result of their research is that they were able to show what the three-dimensional structure of the brightening around Sgr A* looks like in a really interesting period, right after a flare.

The radio brightening was observed back in April 2017 by Chile’s ALMA (Atacama Large Millimeter/submillimeter Array), a network of 66 radio telescopes located in the Atacama Desert at an altitude of 5,000 meters. At that time, ALMA participated in the joint observation campaign of the Event Horizon Telescope collaboration, which resulted in the recording of two years ago. However, the resolution of the radio telescope system alone is not nearly enough to reveal the surroundings of the supermassive black hole, so Levis and his colleagues had to reconstruct the three-dimensional structure from a light curve – essentially a video of a single flickering pixel – which was extremely challenging.

The method used in the research is similar to how a computer tomography (CT) scans a person lying in it, with the important difference that, unlike the body, the flirt can only be observed from a single angle. Their approach is therefore based on the measurement of the temporal change of the structure orbiting Sgr A*, which, according to them, triggers the multiple viewpoints required for conventional tomography.

In order for this time-series-based reconstruction to be possible with ALMA’s measurements, an advanced neural network was needed, which the researchers had to limit with the expected physical regularities around the black hole. Levis gave two examples of this: one is how the gas circulates around the black hole, and the other is how the particles emit electromagnetic radiation called synchrotron radiation during the process.

According to the expert, the research was carried out by combining the fields of gravitational physics and artificial intelligence, with which they were able to create high-tech algorithms. The result, as he wrote, is not a normal photograph, but a three-dimensional image created by a computer, which was produced by a neural network based on ALMA’s time series observations and the expected physics around the black hole.

“We place the three-dimensional using a computer method [fény]emission in the orbit around Sgr A*, starting from an arbitrary shape,” explained Levis, how the reconstruction was carried out. They then modeled how the radio telescope system would see the structure in 10-20-30 minutes, or even later, using ray tracing, familiar from modern video cards (used to simulate the realistic lighting of scenes). “Neural radiation fields and ray-tracing technologies based on general relativity provide a way to change the three-dimensional structure until the model matches the observations,” he wrote in his response.

But what exactly did they find? In response to our question, Levis wrote that theoreticians have recently developed several different mechanisms for how such a flare can arise – including one that links the formation of flares to extremely bright, compact regions that suddenly appear in the accretion disk.

Based on what we currently think about the physical processes taking place around Sgr A*, according to the researcher, “the brightness is concentrated in a few smaller regions within the accretion disk, which is an exciting measurement result that matches the theoretical models”. These bright spots orbit between 66 and 84.5 million kilometers from the black hole, roughly half the average distance between Earth and the Sun of 150 million kilometers.

“We modeled the flirt as coming from the accretion disk, so the validity of the results is based on this assumption,” wrote the researcher when we asked about the main assumptions behind the models. However, he says this is based on previous evidence, such as an analysis by the GRAVITY scientific collaboration, which tracked the brightening of a flare around the center of mass of Sgr A*.

The GRAVITY instrument combines the images of the four telescopes that make up the Very Large Telescope (VLT) of the European Southern Observatory (ESO), with which it can measure the motion of celestial bodies at the center of our galaxy with unprecedented precision. “We hope to extend our investigation in the future by relaxing the physical assumptions made in this study, such as the requirement that the brightening be inside the disc,” Levis wrote.

We last covered the Milky Way’s supermassive black hole in March, when the Event Horizon Telescope collaboration published a stunning image of the magnetic fields around Sgr A*. The measurements allow us to conclude that the object may be hiding a high-energy beam (jet) that has not been detected so far.