AG-2024.09-367·astro-ph.HE·cross-listed: gr-qc
Kalman tracking and parameter estimation of continuous gravitational waves with a pulsar timing array
Authors
- Tom Kimpson
- Andrew Melatos
- Joseph O'Leary
- Julian B. Carlin
- Robin J. Evans
- William Moran
- Tong Cheunchitra
- Wenhao Dong
- Liam Dunn
- Julian Greentree
- Nicholas J. O'Neill
- Sofia Suvorova
- Kok Hong Thong
- Andrés F. Vargas
Abstract
Continuous nanohertz gravitational waves from individual supermassive black hole binaries may be detectable with pulsar timing arrays. A novel search strategy is developed, wherein intrinsic achromatic spin wandering is tracked simultaneously with the modulation induced by a single gravitational wave source in the pulse times of arrival. A two-step inference procedure is applied within a state-space framework, such that the modulation is tracked with a Kalman filter, which then provides a likelihood for nested sampling. The procedure estimates the static parameters in the problem, such as the sky position of the source, without fitting for ensemble-averaged statistics such as the power spectral density of the timing noise, and therefore complements traditional parameter estimation methods. It also returns the Bayes factor relating a model with a single gravitational wave source to one without, complementing traditional detection methods. It is shown via astrophysically representative software injections in Gaussian measurement noise that the procedure distinguishes a gravitational wave from pure noise down to a characteristic wave strain of $h_0 \approx 2 \times 10^{-15}$. Full posterior distributions of model parameters are recovered and tested for accuracy. There is a bias of $\approx 0.3$ rad in the marginalised one-dimensional posterior for the orbital inclination $ι$, introduced by dropping the so-called `pulsar terms'. Smaller biases $\lesssim 10 \%$ are also observed in other static parameters.
Submitted
22 September 20241 year ago
Version
v1
License
CC-BY-4.0
DOI
10.48550/arXiv.2409.14613
Chat with this PDF
Ask questions, probe assumptions, request a plain-English summary. Answers cite sections from the preprint itself.
Community
Questions and answers about this paper from other readers. No formal peer review — just a place to think out loud.