First results from the UTMOST-NS pulsar timing programme

TL;DR

This paper presents initial results from the UTMOST-NS pulsar timing program, including hardware, software, timing noise analysis, and glitch detection in 173 pulsars over two years.

Contribution

It provides the first comprehensive overview of the UTMOST-NS pulsar timing program and introduces new methods for analyzing timing noise and detecting glitches.

Findings

Timing noise residuals scale with pulsar parameters and timespan as in previous studies.

Uncertainty in second frequency derivatives depends on the low-frequency cutoff in noise models.

Detected 17 glitches, including one previously unreported.

Abstract

The UTMOST-NS pulsar timing programme operated at the Molonglo Observatory Synthesis Telescope from April 2021 to June 2023, observing 173 pulsars with an average cadence of 50 pulsars per day. An overview of the programme is presented, detailing the hardware, software, and observing strategy. Pulsar timing results are discussed, focusing on timing noise and glitches. It is shown that the scaling of residuals due to timing noise with pulsar parameters and observing timespan is consistent with earlier studies, and that the recovered timing noise parameters remain consistent as the observing timespan is increased. Second frequency derivatives are investigated, and it is shown that the uncertainty on $\ddot{\nu}$ is sensitive to the frequency cutoff in the timing noise model, varying by three-fold approximately depending on whether Fourier modes with frequency lower than the reciprocal of the observing timespan are included. We measure 39 non-zero values of $\ddot{\nu}$ when considering both models with and without low-frequency modes. An analytic scaling relating anomalous braking indices to timing noise amplitude is also validated. Glitches in the sample are discussed, including three detected by an ``online'' glitch detection pipeline using a hidden Markov model (HMM). In total 17 glitches are discussed, one of which, in PSR J1902+0615, has not been reported elsewhere. An ``offline'' glitch search pipeline using the HMM framework is used to search for previously undetected glitches. Systematic upper limits are set on the size of undetected glitches. The mean upper limit is $\Delta\nu^{90\%}/\nu = 6.3 \times 10^{-9}$ at 90\% confidence.