Probing planetary-mass primordial black holes with continuous gravitational waves

Gravitational waves can probe the existence of planetary-mass primordial black holes. Considering a mass range of $[10^{-7}-10^{-2}]M_odot$, inspiraling primordial black holes could emit either continuous gravitational waves, quasi-monochromatic signals that last for many years, or transient continuous waves, signals whose frequency evolution follows a power law and last for $mathcal{O}$(hours-months). We show that primordial black hole binaries in our galaxy may produce detectable gravitational waves for different mass functions and formation mechanisms. In order to detect these inspirals, we adapt methods originally designed to search for gravitational waves from asymmetrically rotating neutron stars. The first method, the Frequency-Hough, exploits the continuous, quasi-monochromatic nature of inspiraling black holes that are sufficiently light and far apart such that their orbital frequencies can be approximated as linear with a small spin-up. The second method, the Generalized Frequency-Hough, drops the assumption of linearity and allows the signal frequency to follow a power-law evolution. We explore the parameter space to which each method is sensitive, derive a theoretical sensitivity estimate, determine optimal search parameters and calculate the computational cost of all-sky and directed searches. We forecast limits on the abundance of primordial black holes within our galaxy, showing that we can constrain the fraction of dark matter that primordial black holes compose, $f_{rm PBH}$, to be $f_{rm PBH}lesssim 1$ for chirp masses between $[4times 10^{-5}-10^{-3}]M_odot$ for current detectors. For the Einstein Telescope, we expect the constraints to improve to $f_{rm PBH}lesssim 10^{-2}$ for chirp masses between [$10^{-4}-10^{-3}]M_odot$.

Probing new light gauge bosons with gravitational-wave interferometers using an adapted semi-coherent method

We adapt a method, originally developed for searches for quasi-monochromatic, quasi-infinite gravitational-wave signals, to directly detect new light gauge bosons with laser interferometers, which could be candidates for dark matter. To search for these particles, we optimally choose the analysis coherence time as a function of boson mass, such that all of the signal power will be confined to one frequency bin. We focus on the dark photon, a gauge boson that could couple to baryon or baryon-lepton number, and explain that its interactions with gravitational-wave interferometers result in a narrow-band, stochastic signal. We provide an end-to-end analysis scheme, estimate its computational cost, and investigate follow-up techniques to confirm or rule out dark matter candidates. Furthermore, we derive a theoretical estimate of the sensitivity, and show that it is consistent with both the empirical sensitivity determined through simulations, and results from a cross-correlation search. Finally, we place Feldman-Cousins upper limits using data from LIGO Livingstons second observing run, which give a new and strong constraint on the coupling of gauge bosons to the interferometer.

How effective is machine learning to detect long transient gravitational waves from neutron stars in a real search?

We present a comprehensive study of the effectiveness of Convolution Neural Networks (CNNs) to detect long duration transient gravitational-wave signals lasting $O(hours-days)$ from isolated neutron stars. We determine that CNNs are robust towards signal morphologies that differ from the training set, and they do not require many training injections/data to guarantee good detection efficiency and low false alarm probability. In fact, we only need to train one CNN on signal/noise maps in a single 150 Hz band; afterwards, the CNN can distinguish signals/noise well in any band, though with different efficiencies and false alarm probabilities due to the non-stationary noise in LIGO/Virgo. We demonstrate that we can control the false alarm probability for the CNNs by selecting the optimal threshold on the outputs of the CNN, which appears to be frequency dependent. Finally we compare the detection efficiencies of the networks to a well-established algorithm, the Generalized FrequencyHough (GFH), which maps curves in the time/frequency plane to lines in a plane that relates to the initial frequency/spindown of the source. The networks have similar sensitivities to the GFH but are orders of magnitude faster to run and can detect signals to which the GFH is blind. Using the results of our analysis, we propose strategies to apply CNNs to a real search using LIGO/Virgo data to overcome the obstacles that we would encounter, such as a finite amount of training data. We then use our networks and strategies to run a real search for a remnant of GW170817, making this the first time ever that a machine learning method has been applied to search for a gravitational wave signal from an isolated neutron star.