Past Projects

Image Reconstruction for ngEHT

The ngEHT is a significant upgrade to the original EHT project; it aims to construct images and movies of black holes using VLBI with an expanded array of telescopes. I pursued several projects for the ngEHT Analysis Challenges, an effort to encourage the development of novel imaging algorithms trained on synthetic datasets. These datasets are intended to simulate realistic ngEHT observing conditions, including systematic errors such as thermal noise, antenna pointing offsets, and atmospheric turbulence. For the challenges, I provided image reconstructions of total intensity and polarized images for multiple ngEHT observing frequencies using the standard eht-imaging package. You can see one of my total intensity image reconstructions in this ngEHT special issue of Galaxies.

On a personal level, it was a truly awesome experience to learn from the ngEHT team at the Black Hole Initiative. Their science and algorithms are truly cutting-edge, and it was incredibly inspiring to witness a culture where theorists, observers, engineers, and philosophers are working hand-in-hand to capture the dynamics of black holes.

Twenty different image reconstructions of M87 stitched together to make a movie (gif). On the left is the ground truth GRMHD model, and the right is the ngEHT synthetic reconstruction.

Testing a Flexible Disk Emission Model

disk The morphology and kinematics of a protoplanetary disk are fundamental properties to constrain in characterizing the dynamics of a disk, and thus its formation. Current spectral line emission models are not optimized for quickly obtaining these physical parameters from an observation. An emission model in preparation by Andrews et al. aims to quantify the morphology and kinematics of an observation using statistical inference and Bayesian techniques.

Using simulated ALMA observations, I ran MCMC retrieval tests on datasets that varied in stellar mass, noise, and spectral and spatial resolution to fit for parameters like the inclination and brightness temperature of a disk. In collaboration with Dr. Feng Long, I modeled the physical properties of the circumbinary disk around the binary star V892 Tau, and determined values for the stellar mass, inclination, and position angle of the disk. You can see our results of our analysis of V892 Tau here.

CERN and Fermilab

CMS My past research in experimental high-energy physics fall under the realm of software development and testing.

At Northeastern, I worked on the track-trigger system of the CMS experiment at CERN. My work investigated a known issue in their particle trajectory-finding software when enabling a setting that approximated all mathematical calculations in the software. This setting simulated how the tracking software would perform on hardware components, which is how the tracking will be implemented in the future upgrade to the High-Luminosity LHC. What was the cause of the issue? About one line of code out of thousands. How long did it take to find it? About three months. That one fix (along with some others) improved the efficiency of the track-finding software in reconstructing particle trajectories in simulated collisions.

As a part of the DOE SULI Program, I developed a data reduction and analysis software pipeline for the REDTOP experiment at Fermilab. The REDTOP experiment is a proposed detector experiment at Fermilab that aims to study fundamental symmetry violations in physics by capturing and analyzing the decay products of the η meson. My work involved reducing and analyzing data collected from test beams to understand the performance of the photomultipliers in the novel dual-readout calorimeter.