Evaluating data analysis challenges in the 3G detector era and discussing the solutions
The second data release from the Parkes Pulsar Timing Array (PPTA) was not sufficient to confidently detect the nanohertz-frequency gravitational-wave background. Evidence for quadrupolar spatial correlations of the background was statistically marginal. Constraints on spatial correlation coefficients roughly followed the expected Hellings-Downs prediction, but the amplitude of the signal remained unconstrained. However, the search showed strong evidence for the hypothesis that pulsars show a common stochastic signal with the same spectrum of temporal correlations. It is expected that this may appear as a precursor to the detection of spatial correlations. This is because the amplitude of the spatially-uncorrelated component of the signal is twice that of the spatially-correlated component required for the detection. However, with the PPTA, we also showed that the same signal may appear due to mismatch of the assumed (prior) distribution of pulsar noise with the one found in nature. In the follow-up publication, we accounted for this effect and concluded that pulsar spectra really share the common component. Below are the details of the both publications.
Search for the gravitational-wave background with the second data release of the PPTA
Boris Goncharov, Ryan Shannon, Daniel Reardon, George Hobbs, Andrew Zic, and the PPTA
- Search for the gravitations-wave background;
- Constraints on the amplitude of the background;
- Constraints on inter-pulsar spatial correlations;
- Constraints on the amplitude and spectral index of the common-spectrum process;
- Dropout analysis of the common-spectrum process;
- Limitation of the common-spectrum process model: simulations with false positives.
Consistency of the common-spectrum signal with the spatially-uncorrelated "pulsar" term of the background
Boris Goncharov, Eric Thrane, Ryan Shannon, and the PPTA
- Introduction of the importance sampling for pulsar timing arrays;
- Generalisation of a common spectrum and distinct pulsar spectra in one model;
- Simulation validation of the general "quasi-common-spectrum" model;
- Measurement of how distinct are pulsar spectra and the conclusion that they are consistent with being the same spectra;
- The quasi-common-spectrum model can also be applied to modelling the distribution of pulsar intrinsic noise parameters.
In the news
Amaldi-14 conference talk
Is the common-spectrum process observed with pulsar timing arrays a precursor to the detection?