- Accepted Paper
Addressing leakage and mode suppression in angular power spectrum estimation for gravitational-wave backgrounds using pulsar timing arrays
Phys. Rev. D - Accepted 19 August, 2026
DOI: https://doi.org/10.1103/cm48-4k18
Phys. Rev. D - Accepted 19 August, 2026
DOI: https://doi.org/10.1103/cm48-4k18
Mapping gravitational-wave background (GWB) anisotropy with pulsar timing arrays (PTAs) is affected by harmonic space mode suppression and mode coupling arising from an array’s nonuniform sky response. Due to computational limitations, one must truncate spherical harmonic expansions at a finite multipole for recovery, l_max^rec, often chosen to equal l_max^Npair=int[sqrt(Npair}-1], where Npair= Npsr(Npsr-1)/2 is the number of distinct pulsar pairs in an array comprised of Npsr pulsars. This choice is motivated by the counting argument that cross-correlations provide at most Npair independent pieces of information for a single frequency bin. Here, we explicitly obtain the multipole value (denoted l_max^res) that approximates the maximum informative angular scale of an array of pulsars. This value is defined by the requirement that a spherical harmonic expansions out to l_max^res approximately span the space of "observable skies" extracted by the array, which is encoded in the Npair eigenmaps of the Fisher information matrix. The value of l_max^res depends on specifics of the PTA configuration. We also explicitly show that GWB power contained in multipoles l >= l_max^res do not significantly affect analyses that use expansions out to l_max^res, due to the mode suppression (i.e., low-pass filtering) induced by the PTA response to the GWB. However, truncating spherical harmonic expansions at l_max^rec< l_max^res leads to leakage of small-scale angular power from multipoles l_max^rec < l < l_max^res. Nonetheless, even if we use l_max^res for recovery, the standard frequentist estimator of the angular power spectrum components C_l is still biased due to modes that are not observable by the array. Although we can partially debias the standard estimator—improving its agreement with an injected angular power spectrum—this reduction in bias comes at the expense of an increase in variance, with especially large variances arising for poorly constrained modes, l>> l_eff. In summary, the results of our analyses strongly advocate for: (i) using l_max^res for PTA analyses involving spherical harmonic expansions, and (ii) using the debiased standard estimator for C_l recovery, but only out to multipoles l<=l_eff (<< l_max^res) corresponding to sufficiently constrained modes.
If the author has provided any supplemental materials with this article they will be available upon publication of the version of record.