First image of a black hole on April 10, 2019

[Hornblower]

It was not about observing infalling objects. I was demonstrating the big difference in gravitational gradients between BH masses, thus a difference might be discernible for the velocity profile comparison. The question is whether or not, even with perhaps nano-arcsec resolution, what is observed for the BH diameter, along with a velocity profile, will improve distance determinations?

I think I can do some quantitative thought-exercise analysis here. Suppose we have systems A and B, each consisting of a black hole with stars orbiting in close where it is gravitationally dominant. Let B be 10 times as far away as A, but the observed mean orbital velocity is the same at a given angular separation from the central body. Since this is 10 times as much linear orbital radius for B, the centripetal acceleration is 1/10 as great. This is consistent with a central mass 10 times that of A. The event horizon diameters will be in this same proportion, so their angular diameters will be equal. So far, the two systems look alike.

If by velocity profile you mean the orbital velocity as a function of orbital radius, that will be inversely as the square root of the radius for both systems. They still look alike, so with just the data given so far we have no means of determining the distance.

Now let us consider the tidal stress on a star at a given angular separation from the center. That varies in proportion to the central mass and inversely as the cube of the linear separation. So yes, A would be rougher than B on a star at a given angular separation. The challenge would be to use the telescope of our dreams to observe stars close to the centers and look for signs of disruption. If we can do that, we can get more direct distance indicators from statistical analysis of the apparent magnitudes of the stars, as we have been doing for many decades.

If I am missing something in your line of thought, please let me know.

[George]

If by velocity profile you mean the orbital velocity as a function of orbital radius, that will be inversely as the square root of the radius for both systems. They still look alike, so with just the data given so far we have no means of determining the distance.

Yep, darn it. Since the math isn't that hard, I crunched some numbers....

Comparing a 1 million sol mass BH to a 10 million sol mass BH gives an EH radius (or Shadow radius) of 10x for the more massive BH. At, say, 100 AU, the 1M BH will have velocities of about 30,000 kps and the 10M BH at 100 AU will have velocities of about 95,000 kps. But when we match for angular scaling, we get the same velocities. The 100M BH extended to 10x the distance (for our angular match) will have the 30,000 kps velocity at 1000 AU (10x the 100AU of the 1M mass BH).

Only if the mass of the BH is predetermined can the shadow size be useful, though it would be nice to have this as perhaps the only direct form of distance measurement in the tool box. My guess that the velocity profile could be useful seems highly unlikely after all.

[Fiery Phoenix]

I had thought that they were going to do Sgr A* as their first target. M87's is 1/3 the angular diameter of our central black hole. I am guessing this means that Sgr A* is too inactive to produce an interesting telltale image.
Hurray for the EHT team for their huge effort to get this image!

It sounds like Sgr A* is next on the list nonetheless, using essentially the same techniques.

[Shaula]

It sounds like Sgr A* is next on the list nonetheless, using essentially the same techniques.

I think someone mentioned earlier but the big issue with SgrA* is mainly its variability. Over the long data collection times and repeat observations this effectively adds in noise, making the reconstruction poorer. Mainly because your assumption that you are looking at the same 'object' every time breaks down if it varies too much.

[Don Alexander]

That would have been me.

Yes, one of the basic assumptions of VLBI is that your source is constant over the time span the Earth's rotation causes you to observe curves on the so-called "u-v plane." I think most of these observations are several hours in length, limited by the time the source is visible to the telescopes along the longest baselines. That is obviously much longer than the variability timescale of Sgr A*.

[George]

That would have been me.

Yes, one of the basic assumptions of VLBI is that your source is constant over the time span the Earth's rotation causes you to observe curves on the so-called "u-v plane." I think most of these observations are several hours in length, limited by the time the source is visible to the telescopes along the longest baselines. That is obviously much longer than the variability timescale of Sgr A*.

Is this a declination issue (Sgr A: -29 deg.)?

[Shaula]

Is this a declination issue (Sgr A: -29 deg.)?

How would declination affect source variability or the Earth's rotation period?

[George]

How would declination affect source variability or the Earth's rotation period?

If the scopes are land-based, then access might be limited perhaps, such as if it has less time above the horizon.

I would have guessed, however, that having to look through the disk might have been the bigger disadvantage.

I'm trying to understand what constitutes variability in observing Sgr A. What is the variability?

[Shaula]

I'm trying to understand what constitutes variability in observing Sgr A. What is the variability?

The source itself appears to be pretty variable for reasons unknown. It is thought to be a combination or some or all of accretion variation, reconnection events, disk hotspots, temporary/collapsing jets, shock effects and hydrodynamic effects like plasma expulsion and expansion.

There have been loads of surveys of this since (I think) the 80s when it was noticed - a quick google gives a number of relevant papers depending on the level of detail you want. This one is about IR variations but is useful because it lays out some of the processes (I shamelessly cribbed the above list from it) and also sets out the observed properties of the variation. The variability does seem to change timescale with wavelength range. The bad news is that it seems to be random.

The disk might have some effect but at the wavelengths used most of it acts as structured clutter that can be accounted for. The problems that they have are related to the source, unfortunately.

[lpetrich]

Is this a declination issue (Sgr A: -29 deg.)?

Yes. Let us consider what is visible for some place and for how much of the day it is visible. For an object at declination (delta) and latitude b, the visibility fraction f is



Modern science and industrialism were developed in mid northern latitudes, and that's where most observatories have been. But some observatories have recently been built in the Southern Hemisphere.

For northern latitudes, the farther north that an object is in the sky, the longer that it will be visible. Likewise, the farther south, the shorter it will be visible. For southern latitudes, this effect will be reversed.

[lpetrich]

I will consider Sgr A*, at declination -29.0 degrees, M87*, at +12.4 d, and M31*, at +41.2 d (the Andromeda Galaxy BH).

The Event Horizon Telescope includes telescopes from Summit Station, in Greenland, at latitude 72.6d N, to the South Pole, at latitude 90d S.

Most of its telescopes are at latitudes 19d N (Hawaii, US and Sierra Negra, Mexico) to 45d N (Plateau de Bure, France), with one mid-latitude southern one at 23d S (ALMA, Atacama Desert, Chile).

Visibility table:

Latitude	Sgr A*	M87*	M31*
73 N	0	0.76	1
45 N	0.31	0.57	0.84
19 N	0.44	0.52	0.60
23 S	0.58	0.47	0.38
90 S	1	0	0

[Don Alexander]

On the other hand, ALMA is by far the largest and most sensitive telescope in this wavelength range, so it is a driver and it is perfectly situated to observe Sgr A*.

This will change somewhat with the addition of the full NOEMA array in the Pyrenees.

[George]

Yes. Let us consider what is visible for some place and for how much of the day it is visible. For an object at declination (delta) and latitude b, the visibility fraction f is...

What is f? Is this a percent of time out of 24 hours above the horizon?

Thanks. This looks handy, once I understand f.

[George]

On the other hand, ALMA is by far the largest and most sensitive telescope in this wavelength range, so it is a driver and it is perfectly situated to observe Sgr A*.

This will change somewhat with the addition of the full NOEMA array in the Pyrenees.

Yes, it had to be a big player in that imaging.

Though the atmospheric window is wide open at these wavelengths (vs. optical), I'm still surprised atmospheric turbulence isn't a big problem for microarcsecond resolution imaging. They don't have adaptive mirrors (disks), right? Do they take thousands of fast exposures and combine them later to counter scintillation, assuming this is an issue?

[George]

The source itself appears to be pretty variable for reasons unknown. It is thought to be a combination or some or all of accretion variation, reconnection events, disk hotspots, temporary/collapsing jets, shock effects and hydrodynamic effects like plasma expulsion and expansion.

There have been loads of surveys of this since (I think) the 80s when it was noticed - a quick google gives a number of relevant papers depending on the level of detail you want. This one is about IR variations but is useful because it lays out some of the processes (I shamelessly cribbed the above list from it) and also sets out the observed properties of the variation. The variability does seem to change timescale with wavelength range. The bad news is that it seems to be random.

The disk might have some effect but at the wavelengths used most of it acts as structured clutter that can be accounted for. The problems that they have are related to the source, unfortunately.

Thanks, Shaula. If I had a list of astronomical ironies, this would make the list; in effect, M87 is more serene than Sgr A. Who would have guessed that?

[Hornblower]

Thanks, Shaula. If I had a list of astronomical ironies, this would make the list; in effect, M87 is more serene than Sgr A. Who would have guessed that?

Maybe not more serene so much as more ponderous, as in comparing an elephant with a mouse.

[lpetrich]

What is f? Is this a percent of time out of 24 hours above the horizon?

Indeed it is.

[ngc3314]

Yes, it had to be a big player in that imaging.

Though the atmospheric window is wide open at these wavelengths (vs. optical), I'm still surprised atmospheric turbulence isn't a big problem for microarcsecond resolution imaging. They don't have adaptive mirrors (disks), right? Do they take thousands of fast exposures and combine them later to counter scintillation, assuming this is an issue?

At short centimeter wavelengths (where I have little bit of experience), the equivalent of optical "seeing" is ionospheric phase disturbances, which have an analogous effect on the reconstructed image. However, by the miracles possible when your detector actually measures phase and does so at a rapid cadence, one can use the source itself to track these changes (known as self-calibration) as long as it gives enough signal during the time needed to measure a phase fluctuation. This became common practice in, for example, VLA data reduction. (Not that I've dug in to see exactly how they did it for the EHT data flow...)

[Roger E. Moore]

More news for everyone following the M87 marvels.

https://arxiv.org/abs/1905.02143

The Spin of M87*

Rodrigo Nemmen (Submitted on 6 May 2019)

Now that the mass of central black hole in the galaxy M87 has been measured by imaging the shadow of the event horizon, the remaining parameter of the Kerr metric that needs to be estimated is the spin a*. We have modeled measurements of the power of the relativistic jet and the mass accretion rate onto the black hole with general relativistic magnetohydrodynamic models of jet formation. This allows us to derive constraints on a* and the black hole magnetic flux phi. We find a lower limit on M87*'s spin and magnetic flux of |a*| > 0.4 and phi > 6 in the prograde case, and |a*| > 0.5 and phi > 10 in the retrograde case, otherwise the black hole is not able to provide enough energy to power the observed jet. These results indicate that M87* has a moderate spin at minimum and disfavour a variety of models typified by low values of phi known as "SANE". M87* seems to prefer the magnetically arrested disk state.

[George]

At short centimeter wavelengths (where I have little bit of experience), the equivalent of optical "seeing" is ionospheric phase disturbances, which have an analogous effect on the reconstructed image. However, by the miracles possible when your detector actually measures phase and does so at a rapid cadence, one can use the source itself to track these changes (known as self-calibration) as long as it gives enough signal during the time needed to measure a phase fluctuation. This became common practice in, for example, VLA data reduction. (Not that I've dug in to see exactly how they did it for the EHT data flow...)

That's pretty incredible. They must be extremely close to reaching their diffraction-limited resolution, and at a very low cost vs. optical space scopes.

[Hornblower]

That's pretty incredible. They must be extremely close to reaching their diffraction-limited resolution, and at a very low cost vs. optical space scopes.

In principle that resolution could be matched in visible light with an optical array with a baseline of some 5 kilometers. If I am not mistaken, the optical and mechanical challenges in a combiner are much more formidable than in the case of their radio counterparts.

[publiusr]

Now, in the movie interstellar, we saw a "horizontal" line of light. So where is the lap band.

I'm supermassive, and I have a lap band.

Or do galactic core BHs not have this due to great size, less severe gradiant, etc.

[CJSF]

Because 1) Interstellar is a movie. 2) The black hole in the movie is being looked at more sideways/equatorially. 3) The M87 BH is facing us nearly pole-on, tilted only 10-15 degrees.

CJSF

[George]

In principle that resolution could be matched in visible light with an optical array with a baseline of some 5 kilometers. If I am not mistaken, the optical and mechanical challenges in a combiner are much more formidable than in the case of their radio counterparts.

Yeah, the shorter wavelength improves resolution for a given aperture, but would adaptive optics allow microarcsecond resolution for any ground-based optical system?

[publiusr]

Because 1) Interstellar is a movie.

Ah--bought into the hype
https://io9.gizmodo.com/the-truth-be...ate-1686120318

The secret no one wants you to know
http://up-ship.com/blog/?p=40894