Binary black hole merger — gravitational lensing visualization
Binary black hole merger as seen by a nearby observer — gravitational lensing wraps the background starfield around each black hole

PhD Research

What Do Colliding Black Holes Look Like?

Our team's doctoral work in numerical relativity addressed a question nobody had answered before: if you were near two black holes as they spiraled together and merged, what would you actually see? The answer required building new simulation tools, a ray-tracing renderer, and a lot of supercomputer time — and a great team to make it happen.

Nobel Prize in Physics · 2017

Featured at the Nobel Prize Ceremony

The gravitational wave detection by LIGO in 2015 was one of the most significant experimental results of the century — confirming a prediction of general relativity that Einstein himself doubted could ever be verified. The Nobel Prize in Physics was awarded in 2017 to the LIGO team for this discovery.

Images and video from our team were featured at that Nobel Prize ceremony. The same work is on permanent display in a 20-year Smithsonian exhibit, informed the visual effects of the film Interstellar, and appeared in documentaries from HBO and PBS NOVA. The underlying science was published in Physical Review Letters and featured as a NASA Astronomy Picture of the Day.


Gravitational Lensing

Tracing Light Through a Merger

General relativity predicts that massive objects curve spacetime, and light follows those curves. Near the sun, the deflection is a small but measurable fraction of a degree. Near a black hole, the effect is extreme — light can orbit the black hole multiple times, creating stacked rings, complex shadow structures, and inverted ghost images of the entire sky.

For a binary system, nobody had computed what a nearby observer would see. We built a ray-tracing code that propagated light rays backward through numerically-generated spacetimes from binary black hole merger simulations, mapping where each ray originated in the distant background star catalog.

High-resolution binary black hole lensing simulation
High-resolution lensing — Einstein rings, layered shadows ("eyebrows"), and secondary images form around both black holes
Cornell clock tower lensed by a black hole
The Cornell clock tower as seen through the lens of a black hole — demonstrating Einstein rings and inverted ghost images

The binary images reveal features that don't exist around isolated black holes: asymmetric brightness due to relativistic Doppler effects as each black hole orbits at a significant fraction of the speed of light, dynamically shifting Einstein rings, and the merger flash itself. We also modeled a maximally spinning (Kerr) black hole with an accretion disk, reproducing effects similar to those depicted in the film Interstellar.

Maximally spinning black hole with accretion disk
Maximally spinning black hole with accretion disk — this style of simulation informed the visual effects of the film Interstellar

Toroidal Event Horizons

The Topology of a Merger

An event horizon — the surface of a black hole from which nothing, not even light, can escape — is topologically a sphere around an isolated black hole. During a merger, as two horizons approach and join, there is a brief but real moment where the combined horizon has a fundamentally different topology: a torus, with a handle connecting the two surfaces before it pinches off into a single sphere.

My thesis developed adaptive algorithms to locate and track event horizons throughout the full inspiral, merger, and ringdown of binary black hole simulations. The resulting visualizations showed for the first time the complete topological evolution of event horizons through a merger — including the toroidal phase that lasts only a fraction of the final orbital period.

Event horizons of merging black holes — wireframe visualization
Event horizons of two merging black holes — the surface of no return briefly forms a toroidal topology before settling into a single sphere

The Team

None of This Was Done Alone

The lensing project started with three of us — Will Throwe, François Hébert, and me — asking what merging black holes would actually look like. It grew into a team of seven, with each person bringing essential pieces: simulation expertise, rendering code, astrophysics intuition, and a lot of late nights debugging ray-tracers.

Will Throwe, Andy Bohn, and François Hébert — the lensing team
Will Throwe, Andy Bohn, and François Hébert — the founding trio of the gravitational lensing project at Cornell · Photo: Jason Koski / Cornell University Photography

All of this ran on supercomputer simulations from the Simulating eXtreme Spacetimes (SXS) collaboration — the same group whose gravitational waveform templates were used to identify the first LIGO detections. None of the visualizations would exist without that underlying infrastructure, the collaboration's decades of work solving Einstein's equations numerically, and the mentorship of Saul Teukolsky.