Harmonic analysis for pulsar timing arrays

Abstract

We investigate the use of harmonic analysis techniques to perform measurements of the angular power spectrum on mock pulsar timing data for an isotropic stochastic gravitational-wave background (SGWB) with a dimensionless strain amplitude $A_{\text{gw}}=2\times 10^{-15}$ and spectral index ${\hat{I}}_{\text{gw}}^3=13/3$. We examine the sensitivity of our harmonic analysis to the number of pulsars (50, 100, and 150) and length of pulsar observation time (10, 20, and 30 years) for an isotropic distribution of pulsars. We account for intrinsic pulsar red noise and use an average value of white noise of ~100 ns. We are able to detect the quadrupole for all our mock harmonic analyses, and for the analysis with 150 pulsars observed for 30 years, we are able to detect up to the $\ell =5$ multipole. We provide scaling laws for the SGWB amplitude, the quadrupole, and $\ell =3$ as a function of pulsar observation time and as a function of number of pulsars. We estimate the sensitivity of our harmonic approach to deviations of general relativity that produce subluminal gravitational wave propagation speeds.

18 pages, 12 figures; v2: matches published version