The trajectory of a light ray passing by any distribution of mass will be deflected, an effect referred to as gravitational lensing. These deflections are slight in most astrophysical scenarios, but they can be dramatic when caused by compact massive objects like black holes.
In these panoramas we simulate the perspective of an observer that is looking out at a distant, luminous, celestial sphere. In the absence of a black hole, or other source of gravity, the light rays emitted from the celestial sphere will travel along straight lines. As a result the observer will see whatever portion of the celestial sphere is directly ahead of them.
If we now introduce a black hole, the image of the celestial sphere seen by the observer will be modified in a couple of ways. Some light rays that would not otherwise reach the observer are now lensed towards them. And other light rays that would have reached the observer will follow trajectories that fall into the black hole. The net result is the appearance of a region of darkness - as though the black hole cast a shadow - and a curiously distorted image of the background celestial sphere around the black hole.
Black holes are predicted to be the endpoint of the evolution of sufficiently massive stars. However until recently there has been only indirect experimental evidence supporting their existence; the exception being the remarkable direct detection of gravitational waves from a merger of two black holes.
An experiment exploring another direct detection approach, associated with optically (at radio wavelengths) resolving the near-horizon region of the supermassive black hole at the center of our galaxy, is now online called the Event Horizon Telescope.
The objective with this approach is to resolve the 'shadow' that the event horizon of a black hole - should there be one - casts due to the strong bending of light by its gravitational field. An observation of the galactic core which reveals a shadow will signal the existence of a black hole, and moreover the precise shape of the shadow can be matched against templates to discriminate between different candidate black hole solutions.
Our goal with this project is to generate such black hole shadow templates for a variety of interesting black hole solutions.
For our simulation we treat the observer as a camera. Each pixel of the image captured by the camera is associated to a light ray, whose motion through space is governed by a set of equations. We numerically integrate these equations backwards in time to trace the ray to its source, either on the celestial sphere, or the horizon of the black hole. The process is illustrated in the animation below.
Illustration of backward raytracing from an observer in various directions around a black hole.
If a ray originated at a point on the celestial sphere, we set the color of the corresponding camera pixel to match that of the point. If the ray 'originated' at the black hole we simply color the camera pixel black - no light could have come from that direction.
The reason we trace rays backward from the camera, rather than forward from their source as would be more intuitive, is that the vast majority of rays emitted by distant sources never make it to the camera; hence following them is wasteful if our aim is just to determine what the observer/camera sees.