EHT 2017
The stations of the Event Horizon Telescope

The EHT is globe-spanning VLBI array that observes the nearest supermassive black holes in Sgr A* and M87 at 1.3 mm wavelength. After years of preparation the EHT observed Sgr A* and M87 with telescopes around the world in the spring of 2017 and 2018 (some pictures from the observations)

By correlating the recorded electric field measured simultaneously at telescopes around the world, the EHT can effectively synthesize the resolving power of an Earth-sized telescope with an angular resolution of about 10 microarcseconds (1 / 360000000th of a degree!)

Both because EHT measurements are sparse and because absolute phase calibration at millimeter wavelengths is impossible, recovering an image from EHT observations is a difficult and ill-posed problem. Images must be reconstructed using algorithms that find the best-fit images to data under a (hopefully minimal) set of additional assumptions about the source structure.

Simulated EHT image of M87 and reconstruction with eht-imaging.
(Left) an image of M87 at 230 GHz from one of my simulations. (Right) the image reconstructed with ehtim from realistic simulated data similar to the EHT’s observations in 2017. The circle at the lower right represents the EHT’s effective resolution

As a leading member of the EHT’s Imaging Working Group, I develop new Bayesian imaging methods that push the EHT’s imaging capabilities to higher fidelity and resolution. I have developed algorithms that bypass traditional self-calibration both for total intensity (Chael+ 2018) and polarization (Chael+ 2016).

My software library ehtim has become a standard tool across the collaboration for imaging and analyzing EHT data (click here for more information). The techniques and software I’ve developed for the EHT have wide application across radio and optical interferomerty, and we are currently exploring their application on LOFAR and ALMA datasets.

Reconstruction of an ALMA image of HL Tau using traditional CLEAN vs eht-imaging.
(Left) an image of the protoplanetary disk in HL Tau from Band 7, 0.87 mm ALMA observations using the traditional CLEAN algorithm. (Right) the image reconstructed from the same data with ehtim using an imaging algorithm robust to errors in amplitude and phase calibration (Chael+ 2018).

I work closely on developing imaging algorithms and applying them to EHT observations with Katie Bouman, Kazu Akiyama, and Michael Johnson. Below you can watch Katie’s amazing TEDx talk where she explains the basics of EHT imaging.