Scientists took a panoramic picture of the black hole that humans "saw" for the first time!

Produced by: Science Popularization China

Author: Lu Rusen ( Shanghai Astronomical Observatory, Chinese Academy of Sciences )

Producer: China Science Expo

Original title: "Change the "channel" to watch the black hole: M87 black hole is no longer alone! "

On the evening of April 26, 2023 Beijing time, an international research team led by Chinese scientists announced a "panoramic photo" of a black hole taken in a "new channel", that is, the shooting results of the M87 black hole, the "star black hole" that was captured by humans for the first time, in a new observation frequency band, as shown in Figure 1.

This image shows for the first time the shadow of the M87 black hole and the accretion flow and jet formation region around it in the same image. This new image can help astronomers decompose and analyze the different physical processes in the environment around the M87 black hole and thus understand its overall picture. This work was published in the current issue of the journal Nature.

Figure 1: The jet structure of M87 observed by the Global Millimeter VLBI Array (GMVA) in conjunction with the Atacama Large Millimeter/submillimeter Array (ALMA) and the Greenland Telescope (GLT) at 3.5 mm (colors are marked on a logarithmic scale). For the first time, its dense core was resolved in this band and presented as a ring structure under high-resolution conditions (inset, colors are marked on a linear scale). The angular diameter of this ring structure is nearly 50% larger than the ring structure observed by the Event Horizon Telescope (EHT) at 1.3 mm.

Image credit: R.-S.Lu (SHAO), E.Ros (MPIfR), and S.Dagnello (NRAO/AUI/NSF)

Photographing the "tail" of a black hole on the "new channel"

On April 10, 2019, the Event Horizon Telescope (EHT) collaboration announced that it had captured humanity's first-ever photo of a black hole at the center of the M87 galaxy, making the M87 black hole "popular" around the world overnight and a well-known "celebrity" among black holes.

However, those who are familiar with the observation history of M87 know that more than 100 years ago, people discovered that there was a peculiar collimated beam of light, a "curious straight ray" (i.e. a jet) in the M87 galaxy, emanating from its center and extending to 5,000 light-years away. This is the jet of M87.

In fact, so far, the jets in M87 have been fully imaged in all electromagnetic radiation bands, from low-energy radio bands to high-energy gamma-ray bands, as shown in Figure 2. But strangely, no jets were seen in the black hole images taken by the EHT in the early stage. In addition, the theory believes that there is an accretion flow around the black hole, which is the energy source for "lighting up" the jets, but there has been no direct imaging detection of the accretion flow before.

Figure 2: Images of the center of the M87 galaxy at different scales

Image source: Reference 4

So the question is, why didn't the EHT capture the jet? There are two possible reasons. One is the distribution of the telescopes: the EHT is an array of eight radio telescopes around the world, and the distances between the telescopes are too far. The "field of view" of the virtual telescope formed by the array is limited to a very small area around the black hole, so it is not able to capture the jet outside the black hole.

Another reason is the jet itself. The jet appears darker at shorter observation wavelengths, making it difficult to detect. Especially in the 1.3 mm band where the EHT works, due to the strong gravitational lens effect of the black hole, the light from the accretion flow and the jet will be bent into a ring structure of similar size. So even if the EHT captures the jet, it is likely to be hidden in the bright ring around the shadow. Therefore, this also means that simply looking at the 1.3 mm image of the bright ring around the black hole cannot distinguish whether it is produced by the accretion flow or the jet.

In order to capture the initial jet formation area that the EHT failed to capture, to image the accretion flow around the M87 black hole, and to explore the connection between the black hole and the accretion flow and jet, we conducted imaging observations of M87 in the 3.5 mm band for the first time in April 2018 using 16 telescopes to form a radio telescope equivalent to the diameter of the Earth. The schematic diagram of the observation array is shown in Figure 3.

These 16 telescopes are combined into an array using very long baseline interferometry (VLBI), including 14 telescopes in the Global Millimeter VLBI Array (GMVA) and the Atacama Large Millimeter/submillimeter Array (ALMA) and the Greenland Telescope (GLT). The addition of ALMA greatly improves the observation capabilities of the array compared to previous observations with GMVA alone. It improves our resolution in the north-south direction (i.e., the direction perpendicular to the jet) by a factor of four, and "anchors" the entire array with its ultra-high sensitivity. ALMA is like a real game changer in the current millimeter-wave VLBI observation array, so from the moment we got the ALMA observation project, the whole team was excited because everyone knew that the real "king bomb" was coming!

Figure 3: Schematic diagram of the 2018 observation array consisting of the Global Millimeter-wave VLBI Array (GMVA), the Atacama Large Millimeter/submillimeter Array (ALMA) and the Greenland Telescope (GLT)

Image credit: MPIfR/Helge Rottmann

New features captured by the new array

It is not easy to win with a good hand. When a newly formed telescope array works together, all kinds of unexpected things happen from time to time. For example, the Greenland Telescope that joined the array this time is a new telescope, and it is still in the debugging stage when it participates in the work. During the observation process, its waveguide-based phase rotator was configured incorrectly, so that the subsequent data processing and analysis required the development of special algorithms to handle it. There are many unexpected situations in such observations.

To this end, we performed four VLBI analyses of “cross-correlation” and corresponding “post-correlation processing” before finally obtaining data that could be used for reliable imaging.

Although this back-and-forth process was grueling, I was struck by interesting features when I first processed some of the data and checked them. We found that the "closed phase" measured on some of the baseline triangles formed by the three telescopes in the array that were far apart was almost 180 degrees, which meant that the "radio core" of M87 had been resolved!

At the same time, we found that the amplitude of the measured “visibility” first decreases and then increases as the baseline length increases, forming the so-called “visibility null” [see below and Figure 4]. These are data features that have not been seen before in any similar 3.5 mm observations.

This unexpected feature is exciting and gives us great motivation to keep going. All team members know that this time we should definitely have the product. To ensure the reliability of the product, we verify the results by having different team members perform independent data calibration and confirm the results by using different data calibration methods.

In fact, these data features have already told us a lot about the structure of the M87 black hole before imaging. For example, both the "closed phase" and "visibility" amplitude features are consistent with the expected ring structure, and the position of the "zero point" in the amplitude information can also tell us the size of the ring. Figure 4 shows the change in the visibility amplitude corresponding to an ideal ring as the baseline length between telescopes changes. As the size of the ring changes, the position of the "zero point" will also change.

Interestingly, the position of the "zero point" we measured at 3.5 mm is significantly different from the position of the first "zero point" observed by EHT at 1.3 mm, indicating that the sizes of the two ring structures are different. Since the baseline length corresponding to the position of the "zero point" we measured is shorter, it means that the ring observed at 3.5 mm is larger! At the same time, we also have sufficient and reliable evidence that the different sizes of the two rings are not caused by different observation times (the black hole did not get fat!).

So the question is, does it matter if the rings are of different sizes?

Figure 4: The amplitude of the visibility data corresponding to the idealized ring model varies with the (projected) inter-telescope baseline length (in units of observation wavelengths). As the diameter of the ring increases, the position of the first "zero point" of its visibility will move toward the short baseline. The black vertical line in the figure marks the longest baseline length that can be achieved by the global 3.5 mm array on the ground (about 3×109 times the wavelength). For an idealized ring of 42 microarcseconds, which is consistent with previous EHT observations, the ground-based array has difficulty detecting this characteristic feature at 3.5 mm, and therefore it is difficult to resolve it and determine its specific shape. However, for an idealized ring of 64 microarcseconds, its corresponding first "zero point" can be detected by the ground-based 3.5 mm array.

Image source: Shanghai Astronomical Observatory, Chinese Academy of Sciences

A big donut that overturns cognitive assumptions

In fact, at first no one thought that the ground-based interferometer array could observe a ring-shaped structure at 3.5 mm, or detect this "zero point" in visibility!

This is because if the "donut" in the EHT black hole shadow image corresponds to the colorless (achromatic) photon ring around the black hole, then its size (angular diameter) at 3.5 mm should be 42 microarcseconds, the same as at 1.3 mm. The position of the first "zero point" of its visibility amplitude will be far beyond the range that can be covered by the 3.5 mm array on Earth (shown by the black line in Figure 4), so it will be undetectable.

Under this entrenched assumption, people do not believe that ground-based arrays can detect a ring structure at 3.5 mm. An interesting thing is that in order to maintain this assumption, some collaborators even do not allow the use of "ring" or "ring-like" when proposing new observations, which are "specific" words used to describe the donut in the EHT image.

Figure 5: Actual observed images of M87* (i.e., the radio core of M87) at 3.5 mm and 1.3 mm

Image source: Shanghai Astronomical Observatory, Chinese Academy of Sciences

We actually saw larger rings than we had hypothesized, overturning previous assumptions.

We further confirmed this result in the process of reconstructing the image from the observational data. We found a ring-like structure in the image using different VLBI imaging methods, which is consistent with the features found in the observational data. Similar to the previous EHT work, we also searched a large number of imaging parameters to determine the final image.

In addition to the imaging parameters, we also considered the potential impact of the jet component in the data on the imaging of the ring structure in the compact core region. By analyzing a large number of images and the possible impact of the layout of the telescopes in the array and various imaging parameters on the image details, we were able to finally determine that there is indeed a ring structure in the core of M87.

Ultimately, by measuring the size of the ring structure in a large number of images and by direct model fitting of the observational data, we were able to determine that the size of the ring structure observed at 3.5 mm is 64 microarcseconds, which is nearly 50% larger than the ring structure measured by EHT at 1.3 mm (42 microarcseconds)!

So what does this new giant donut mean?

New images reveal physics

We used computers to simulate the accretion flow and jet of a black hole, and calculated how the light generated by these physical processes formed the observed image. We wanted to understand whether the light that formed the ring structure mainly came from the accretion flow or the jet, so we compared the two cases where all the light came from the accretion flow and all the light came from the jet (considering the case where the light mainly came from the accretion flow and the light mainly came from the jet does not change the final conclusion).

We found that at 1.3 mm, light from both the accretion flow and the jet can form a ring consistent with the EHT observation under the gravitational lens effect. But at 3.5 mm, only the accretion flow model can produce a larger ring consistent with our observation. This is because the accretion flow is not completely "transparent". The light generated in the inner region of the accretion flow will be partially absorbed when passing through the outer region, while the light generated in the outer region will not be absorbed. In this way, more light will come from the outer region of the accretion flow, forming a larger ring structure image.

Theoretically, if the "donut" around the black hole is further decomposed, it is actually composed of many different sub-rings. Affected by gravitational lensing, some light will circle the black hole several times before reaching the observer, forming very thin sub-rings. If n is used to represent the sequence of the sub-rings, the light will have circled the black hole n/2 times before reaching the observer. The n=0 ring is the image formed by the light directly reaching the telescope after it is emitted, and only the size of this ring changes with the observation wavelength. The longer the wavelength, the more "opaque" the accretion flow, and the larger the n=0 ring [Figure 6].

Figure 6: Schematic diagram of the composition of a black hole image. Due to its strong gravitational force, a black hole casts a "shadow" on the bright matter around it. This shadow is bounded by a bright ring of light, corresponding to the photons that passed near the black hole before escaping. The ring is composed of increasingly sharp sub-rings, and the photons corresponding to the nth sub-ring rotated around the black hole n/2 times before reaching the observer, where the n=0 sub-ring is the "direct" image of the radiation zone around the black hole. When the observation wavelength increases, the diameter of the n=0 ring will increase due to the opacity of the radiation.

Image source: Shanghai Astronomical Observatory, Chinese Academy of Sciences

The successful explanation of the observations by the accretion flow model also means that this observation is the first direct imaging detection of the accretion flow.

On the other hand, the jet structure in the new image also allows us to further understand its origin.

Since our observation array has more stations (16 in total) than the EHT array (the EHT's first black hole imaging only had 7 stations distributed in 5 locations), it has high sensitivity, and because the jet is brighter in this band, we can achieve detailed imaging of the area within ~100 Schwarzschild radii (Rs) from the black hole [see Figures 1 and 5].

We found that the jet is indeed generated near the event horizon of the black hole and there is a three-toothed structure with a "spine/sheath" and brightened edges. This structure is likely caused by velocity stratification in the jet. By measuring the width of the jet at different locations, we found that the width profile of the M87 jet (i.e., the change in jet width with the distance from the black hole) is exactly consistent with the jet produced by the Blandford–Znajek mechanism, that is, the jet is produced by extracting the rotational energy of the black hole. However, at the edge of the black hole (within ~20Rs), the observed jet is significantly wider than the jet predicted by this mechanism. This may be caused by the influence of the "wind" in the accretion flow. Future observations for longer periods of time will hopefully reveal the dynamic process of how the "wind" affects the jet.

Figure 7: Jet width as a function of distance from the black hole

Image source: Reference [1]

Starting from the "new": Taking "color" photos and movies of black holes

Currently, both the previous EHT black hole images at 1.3 mm and the 3.5 mm black hole images taken this time are static "black and white" photos taken with a single color of "radio light". In the near future, we will be able to take "color" photos or even "color" movies of black holes through multi-frequency simultaneous observations. In this way, we can distinguish between the "colorless" "eternal" structure caused by gravity and the "colored" "time-varying" structure caused by astrophysical processes in the black hole image, so as to explore the space-time around the black hole more deeply and understand the related astrophysical processes around the black hole.

After nearly five years, we used the global millimeter-wave VLBI array and combined with the ALMA and GLT telescopes to image the M87 black hole and its surrounding accretion flows and jets at different observation frequency bands with the EHT, taking a "panoramic photo" of the black hole. The main completion period of the entire work coincided with the global COVID-19 pandemic, but this did not prevent close communication and collaboration among team members. Roughly speaking, the author organized nearly a hundred telephone conference discussions and communicated through nearly a thousand emails before finally completing this work successfully. Looking to the future, through long-term simultaneous monitoring of nearby supermassive black holes including M87 at multiple frequencies, we hope to shoot a "color" movie of a black hole in the near future.

References:

[1] Lu, R.-S., Asada, K., Krichbaum, TP et al., “A ring-like accretion structure in M87 connecting its black hole and jet”, 2023, Natrue, in press.

[2] Lu Rusen and Zuo Wenwen, “The world’s first black hole photo is out. What contribution have Chinese scientists made?”, 2019, Science Daily

[3] Curtis, HD “Descriptions of 762 Nebulae and Clusters Photographed with the Crossley Reflector”, 1918, Publications of Lick Observatory, 13, 9.

[4] EHT MWL Science Working Group, Algaba, JC, Anczarski, J., et al., “Broadband Multi-wavelength Properties of M87 during the 2017 Event Horizon Telescope Campaign”, 2021, ApJL, 911, L11.

[5] Event Horizon Telescope Collaboration, Akiyama, K., Alberdi, A., et al., "First M87 Event Horizon Telescope Results I: The Shadow of the Supermassive Black Hole", 2019, ApJL, 875, L1.

[6] Kim, J.-Y., Lu, R.-S., Krichbaum, TP, et al., “Resolving the Base of the Relativistic Jet in M87 at 6Rsch Resolution with Global mm-VLBI”, 2016, Galaxies, 4, 39.

[7] Kim, J.-Y., Krichbaum, TP, Lu, R.-S., et al., “The limb-brightened jet of M87 down to the 7 Schwarzschild radii scale”, 2018, A&A, 616, A188.

[8] Event Horizon Telescope Collaboration, Akiyama, K., Alberdi, A., et al., “First M87 Event Horizon Telescope Results. IV. Imaging the Central Supermassive Black Hole”, 2019, ApJL, 875, L4.

[9] Blandford, RD & Znajek, RL, “Electromagnetic extraction of energy from Kerr black holes”, 1977, MNRAS, 179, 433.