If black holes are spherical, so shouldn't it be absorbing light from all the directions? How would one explain the light ring. which seems more like a 2D ring than a sphere?
1 Answer
One can never actually see the event horizon, one sees the gas that is about to enter the event horizon and gets hot along the way. But this gas has a history, it doesn't just fall straight in. The history imparts some angular momentum to it, like the water in a draining bathtub. So it spirals in, not falling straight in, and this creates a disk shape called an accretion disk. The disk does not have to be as thin as a pancake, but it's far from spherical, and that gets further distorted by the strong gravity (creating the brightness variations and the central dark sphere).
The disk does have a hole in the center because at some point the orbits go unstable and fall quickly below the event horizon. It could be argued we are not actually seeing the event horizon where it goes black, we are seeing the drop in gas density where the orbits are no longer stable and the slow spiral turns into a very rapid fall, but the way the strong gravity bends the light rays is always very important as well.
The distortion by the strong gravity is what provides the test of general relativity that everyone is so excited about. The strong gravity produces a "shadow" effect, so you don't have to see the accretion disk from above to get a spherical blackness at the center.
-
1$\begingroup$ You don't see (the shape) of the accretion disk at all. $\endgroup$– ProfRobApr 12, 2019 at 8:44
-
$\begingroup$ Yet the shape of the accretion disk is an input to the gravitational distortion, so it leaves a mark. But yes, it's not like you are looking at the physical shape of the disk, that would not leave a circular black spot in the center. It's gravitationally distorted to look like a round "shadow." The disk is probably more relevant to the reason it shows bright spots than to the black circle at the center, and to the reason it doesn't quite look like a circle (though the resolution is pretty poor so it's not clear how much can be known for sure). $\endgroup$– Ken GApr 13, 2019 at 5:19