The dawn of low frequency Gravitational Wave astronomy
Lead Research Organisation:
University of Surrey
Department Name: Physics
Abstract
Einstein's general theory of relativity predicts that binaries of massive objects like neutron stars and black holes lose energy by emission of gravitational waves, ripples in the fabric of spacetime that propagate through the cosmos at the speed of light. Gravitational waves have been the last untested prediction of general relativity, until in 2015 the LIGO collaboration announced the first ever detection of the signal emitted by the coalescence of two stellar mass black holes. Since then, many more mergers have been detected, opening a new window on the Universe and allowing us to study black holes, the most mysterious objects in the Universe. LIGO can only detect the smallest black holes, formed by the evolution of massive stars. But much more massive black holes can be found at the centres of galaxies, and binaries of such black holes represent the loudest sources of gravitational waves in the Universe. These waves have much lower frequencies, and require a different detector. Currently, radio telescopes around the globe, linked in so called Pulsar Timing Arrays (PTAs), process radio signals from known pulsars to detect the tiny perturbations caused by travelling gravitation waves. The superposition of the signals emitted by a large population of supermassive black hole binaries (BHBs) produces a background (GWB) that can be used to answer fundamental questions about black hole and galaxy formation and evolution: What is the efficiency of galactic mergers and black hole mergers? What sets the amplitude of the GWB?
Recently, a low frequency signal consistent with a large population of BHBs has been identified by PTAs. While a confirmation of a BHB origin will require the detection of spatial correlations in the signal that have not yet been found, it is expected this will occur in the next few years, allowing us to study massive black holes in a new way. The main aim of this proposal is to drive changes in current theoretical models of supermassive black hole binary evolution so that we can use the observed amplitude of the background to answer fundamental questions on the connection between galaxies and black holes.
We will combine high accuracy and high resolution simulations of galactic mergers and BHBs with machine learning techniques to compute merger timescales as a function of both black hole and galaxy properties. Because of the computational cost of state-of-the-art simulations, we will select a sample of representative initial conditions and then develop a new, tailored, machine learning tool to learn the outcomes of simulations and probe the wider parameter space required to compute the GWB. We will constrain the role of black hole and galaxy properties in setting the amplitude of the background, breaking known degeneracies. As a result, this work will provide evidence for either the need to revise scaling relations between black holes and galaxies or the need to include environmental effects (galaxy density profiles) into existing models.
Recently, a low frequency signal consistent with a large population of BHBs has been identified by PTAs. While a confirmation of a BHB origin will require the detection of spatial correlations in the signal that have not yet been found, it is expected this will occur in the next few years, allowing us to study massive black holes in a new way. The main aim of this proposal is to drive changes in current theoretical models of supermassive black hole binary evolution so that we can use the observed amplitude of the background to answer fundamental questions on the connection between galaxies and black holes.
We will combine high accuracy and high resolution simulations of galactic mergers and BHBs with machine learning techniques to compute merger timescales as a function of both black hole and galaxy properties. Because of the computational cost of state-of-the-art simulations, we will select a sample of representative initial conditions and then develop a new, tailored, machine learning tool to learn the outcomes of simulations and probe the wider parameter space required to compute the GWB. We will constrain the role of black hole and galaxy properties in setting the amplitude of the background, breaking known degeneracies. As a result, this work will provide evidence for either the need to revise scaling relations between black holes and galaxies or the need to include environmental effects (galaxy density profiles) into existing models.
