| 27th of September 2023 |
|---|
 |
| ATNF Colloquium |
| Enabling rapid discovery of gravitational waves using machine learning |
| Chayan Chatterjee (UWA) |
|
Abstract:
Gravitational waves are ripples in spacetime curvature created by some
of the most energetic events in the universe like the collisions of
compact astrophysical objects like black holes and neutron
stars. These ripples propagate through space at the speed of light and
are detected by several km long laser interferometers called LIGO and
Virgo, located in the US and Italy. The simultaneous observation of
gravitational waves and prompt electromagnetic emissions from the
compact object mergers can help reveal properties of extreme matter
and gravity during and immediately after the coalescence. However,
such simultaneous observations rely on rapid detection and sky
localization of gravitational waves, often requiring alerts to be sent
out before merger.
In this seminar, I will describe different machine learning models that I have developed to solve challenging problems in rapid gravitational wave discovery – pre- and post-merger sky localization, and waveform extraction from real detector data. I will demonstrate the accuracy and feasibility of these methods on simulated data as well as on real gravitational wave events detected during the first three observation runs of LIGO and Virgo. Finally, I will talk about the future scope of this research work and highlight some of the other areas in gravitational wave astronomy where these deep learning techniques can be applied. |