Null-stream pointing with pulsar timing arrays

Abstract

Locating sources on the sky is one of the largest challenges in gravitational wave astronomy, owing to the omni-directional nature of gravitational wave detection techniques, and the often intrinsically weak signals being observed. Ground-based detectors can address the pointing problem by observing with a network of detectors, effectively triangulating signal locations by observing the arrival times across the network. Space-based detectors will observe long-lived sources that persist while the detector moves relative to their location on the sky, using Doppler shifts of the signal to locate the sky position. While these methods improve the pointing capability of a detector or network, the angular resolution is still coarse compared to the standards one expects from electromagnetic astronomy. Another technique that can be used for sky localization is null-stream pointing. In the case where multiple independent data streams exist, a single astrophysical source of gravitational waves will appear in each of the data streams. Taking the signals from multiple detectors in linear combination with each other, one finds there is a two parameter family of coefficients that effectively null the gravitational wave signal; those two parameters are the angles that define the sky location of the source. This technique has been demonstrated for a network of ground-based interferometric observatories, and for 6-link space interferometers. This paper derives and extends the null-stream pointing method to the unique case of pulsar timing residuals. The basic method is derived and demonstrated, and the necessity of using the method with multiple sub-arrays of pulsars in the pulsar timing array network is considered.