VLBI Data Series 3, Imaging Techniques
Facilitator: Andrew Chael, Princeton University
Air date: Tuesday, March 18, 2020
The Event Horizon Telescope (EHT) has imaged the emission just outside the event horizon of the supermassive black hole in M87 with an effective resolution of ~20 microarcseconds, approximately 5 billionths of a degree. The EHT obtains this extremely high resolution with Very Long Baseline Interferometry (VLBI), correlating the signals received by millimeter telescopes separated by thousands of kilometers to measure individual spatial frequencies of the black hole image on the sky. Because the EHT array is sparse, these measurements do not uniquely constrain the source structure and algorithms must be used to infer the underlying image.
In this webinar, we will learn the basics of image reconstruction with the EHT and other interferometers. We will focus on a relatively new class of methods called “regularized maximum likelihood” (RML) techniques. RML algorithms find the most likely image fit to the measured data under constraints or ‘regularizers’ that enforce characteristics of the image like smoothness or sparsity. We will learn about the unique challenges of imaging using EHT data, which cannot be absolutely phase-calibrated, and how these challenges can be addressed in the RML framework. We’ll also provide a tutorial on using the EHT-imaging library, one of the software packages developed for RML image reconstruction by the EHT. While we will focus on applications of RML imaging to the EHT, we will also show how the same techniques can be used for reconstructing images from other arrays like the VLBA, VLA, and ALMA.
This session is part of a monthly series.
Subscribe and watch past webinars here: