Searching for continuous gravitational waves in the Parkes Pulsar Timing Array Data Release 3

TL;DR

This study conducts an all-sky search for continuous gravitational waves from individual supermassive binary black holes using the Parkes PPTA DR3 dataset, applying both Bayesian model selection and frequentist F-statistics while rigorously modeling noise components including a common red noise. No CGW detection is found; Bayes factors and FAP analyses show CGW signals are not favored once the CRN is accounted for, though the analysis achieves a factor of ~4 improvement in CGW sensitivity over the prior PPTA DR1 across 1–200 nHz. The authors report sky-averaged and directional upper limits on the CGW strain amplitude $h_0$, with a 95% upper limit around $7\times10^{-15}$ near 6 nHz and structured skies showing up to a factor of ~5 variation, translating into distance and SMBH population constraints. The results strengthen the nanohertz GW landscape, provide concrete SMBBH density and merger-rate limits, and demonstrate the ongoing potential of PPTA and the IPTA to eventually identify individual CGWs as nanohertz gravitational-wave astronomy matures.

Abstract

We present results from an all-sky search for continuous gravitational waves from individual supermassive binary black holes using the third data release (DR3) of the Parkes Pulsar Timing Array (PPTA). Even though we recover a common-spectrum stochastic process, potentially induced by a nanohertz gravitational wave background, we find no evidence of continuous waves. Therefore, we place upper limits on the gravitational-wave strain amplitude: in the most sensitive frequency range around 10 nHz, we obtain a sky-averaged 95\% credibility upper limit of $\approx 7 \times 10^{-15}$. Our search is sensitive to supermassive binary black holes with a chirp mass of $\geq 10^9M_{\odot}$ up to a luminosity distance of 50 Mpc for our least sensitive sky direction and 200 Mpc for the most sensitive direction. This work provides at least 4 times better sensitivity in the 1-200 nHz frequency band than our last search based on the PPTA's first data release. We expect that PPTA will continue to play a key role in detecting continuous gravitational waves in the exciting era of nanohertz gravitational wave astronomy.

Figures (11)

• Figure 1: Savage-Dickey Bayes factors in support for a CGW at each GW frequency. The orange curve indicates results for the case where common red noise (CRN) is not included in the analysis, while the black curve is for the case where it is included. We found that no CGW are detected in the PPTA DR3 if a CRN is accounted for. The black and gray vertical dotted lines mark, respectively, the frequency of $1/\mathrm{yr}$ and of $1/T_{\mathrm{obs}}$. These two vertical lines retain the same meaning in all subsequent figures of this paper.
• Figure 2: Bayes factor for comparison between different models, where numerical values represent the extent to which the data favors the model to which the arrow points. CGW here refers to a continuous wave signal with frequencies between 1 and $\unit[2.5]{nHz}$, the frequency range where the BF peaks in Figure \ref{['fig:S-D_BF']}.
• Figure 3: The values of the all-sky CGW frequentist detection statistic $\mathcal{F}_\text{p}$ and the respective FAP are shown against GW frequency.
• Figure 4: The values of the CGW detection statistic $\mathcal{F}_\text{e}$, maximized over a range of $192$ uniformly-distributed sky positions for every GW frequency.
• Figure 5: All-sky 95% credibility upper limits on the CGW strain amplitude $h_0$. The PPTA DR3 (black) is more sensitive than the DR1 (blue) by at least a fractor of four over a wide frequency range. For comparison, the most recent strain upper limits from the EPTA, NANOGrav, and IPTA are also included in the figure.