Simulated inspiral, merger and ringdown, shifted into audible range.
Gravitational wave (GW) lensing arises from the dynamic interplay between gravitational waves and the spacetime curvature induced by massive celestial bodies. This spacetime curvature, a consequence of the presence of mass and energy, effectively acts as a refractive medium—akin to a lens—reshaping the paths of gravitational waves.
Much of my work has been centered around exploring the parameter space of the lensed GWs, the lens being fuzzy dark matter (FDM).
I have also worked on extending a Python library, lensGW to produce lensed GW signals using GW waveform templates available on PyCBC. Recently, I have been contributing to developing the AEI lensing library under the supervision of Dr. Miguel Zumalacarregui.
Gravitational waves (GWs) can be represented as a combimation of dominant fundamental mode and a linear combination of higher modes. While the primary or fundamental mode usually has the highest power, higher harmonics add complexity and information to the signal. These harmonics are essentially multiples of the fundamental frequency and are generated by asymmetric or non-central-symmetric sources of gravitational waves, such as binary black holes or neutron star mergers.
Studying higher harmonics provides valuable insights into the source's dynamics, shape, and the underlying physics of extreme astrophysical events. They can reveal details about the system's pre-merger evolution, the merger process itself, and the properties of the resulting remnant, all of which enrich our understanding of the universe's most energetic and cataclysmic phenomena.
With inclusion of higher harmonics its is possible to refine measurements and improve our ability to infer crucial parameters like the masses and spins of the merging objects, the inclination of the orbital plane, and the distance to the source.
Inclusion of higher harmonics can also allow detecting GWs much ealier before the merger so in pricinple this property can be used to study the pre-merger detection - a study I conducted under supervision of Prof. Alex Nitz.
I worked on compressing waveforms of extreme mass-ratio inspirals (EMRIs) under the supervision of Dr. A. Hees and Dr. N. Korsakova. EMRIs involve a compact stellar-mass object, such as a black hole or neutron star, inspiraling into a supermassive black hole. The resulting gravitational wave signals contain an abundance of information about both the EMRI source and the central supermassive black hole, making them invaluable for astrophysical studies.
However, the sheer volume of data generated by EMRI waveforms can be overwhelming, especially for space-based observatories like LISA (Laser Interferometer Space Antenna). If compressed successfully, it would enable extracting the rich astrophysical information in EMRI signals.
Machine learning has become an indispensable tool in modern astronomy. By leveraging algorithms and computational techniques, astronomers can analyze vast datasets, identify patterns, and make sense of complex celestial phenomena. From classifying galaxies and discovering exoplanets to predicting space weather and improving data analysis, machine learning is propelling astronomical research to new frontiers, helping us unravel the mysteries of the cosmos more efficiently and effectively than ever before.
I have used the capability of ML techniques towards parameter estimation of Gravitational Waves, parameter estimation of merging galaxies and the ability of ML to predict future astronomical events. Recently, I have been exploring further Reinforcement learning and Physics Informed learning and how well we can solve some of the most time-consuming and intricate problems.