Project:	Fantastic Beasts, but Where to Find Them? Machine Learning–Driven Discovery and Characterisation of Giant Black Hole Jets
Disciplines:	Astronomy, Computer Science; Astrophysics
Mentor:	Gregg Hallinan, Professor of Astronomy; Director, OVRO, (PMA), gh@astro.caltech.edu
Mentor URL:	https://pma.caltech.edu/people/gregg-w-hallinan  (opens in new window)
AO Contact:	Martijn Oei, oei@caltech.edu
Background:	Modern, large-area sky surveys at decimetre to decametre wavelengths, such as LOFAR’s at λ = 2 m, are revealing that the generation of giant jets by supermassive black holes is far more common than previously thought. Upcoming instruments, such as Caltech's DSA-2000, will open a new window to study this exciting phenomenon. Giant jets transport relativistic electrons and positrons, atomic nuclei from stellar winds, magnetic fields, and heat from the centres of galaxies to the depths of the intergalactic medium. Once these jets finally break down as a result of magnetohydrodynamical instabilities, they have formed the largest galaxy-made structures in the Universe; some have carried as much energy with them as is released during galaxy cluster mergers. At Caltech, the radio astronomy group leads efforts to better understand how many giant black hole jets have existed from the present day up to the first gigayears, how they form, and what effects they have on the magnetisation, thermodynamics, and the elemental enrichment of the cosmic web.
Description:	There is one SURF project on offer, which will be jointly mentored by Martijn Oei and Gregg Hallinan (DSA-2000 PI). The goal of the project will be to build on current, or devise new, machine learning methods to accelerate the discovery of giant black hole jets in real-world radio survey data. This process also involves finding the galaxies from which the jets were launched. To show that our algorithms work, and to roughly tune them for application to DSA-2000 imagery, we will test them on the yet unexplored LOFAR Two-metre Sky Survey DR3, a premier data set for metre-wavelength radio astronomy. While we prepare for the future, the project may thus reveal thousands of previously unknown giant jet systems.
References:	https://www.nature.com/articles/s41586-024-07879-y https://ui.adsabs.harvard.edu/abs/2024A&;A...691A.185M/abstract https://www.nytimes.com/2024/09/25/science/space/black-hole-m87-energy.html https://www.quantamagazine.org/physicists-identify-the-engine-powering-black-hole- energy-beams-20210520/
Student Requirements:	Python programming experience and some familiarity with machine learning methods (e.g. convolutional neural networks, attention, transformers)
Programs:	This AO can be done under the following programs:   Program    Available To        SURF    Caltech students only  Click on a program name for program info and application requirements.