The ONs and OFFs of Pulsar Radio Emission: Characterizing the Nulling Phenomenon

TL;DR

This work addresses the challenge of characterizing pulsar nulling, particularly in weak sources where per-pulse signal-to-noise hinders traditional ON/OFF classification. It introduces the $\mathbb{N}$sum algorithm, which uses sums of pulses to improve distinguishability from noise and provides robust estimates of the nulling fraction $n_{\rm F}$, benchmarked against Gaussian Mixture Modelling on simulated and real data. The study extends to quasi-periodicity in nulling, reporting measurements for five pulsars (including first measurements for some) and examining connections to spin-down energy loss $\dot E$ and prior results. Overall, the approach enhances robustness of nulling measurements across diverse pulsar brightnesses and informs how nulling and quasi-periodicity relate to pulsar evolution and death lines.

Abstract

Radio emission from pulsars is known to exhibit a diverse range of emission phenomena, among which nulling, where the emission becomes temporarily undetectable, is an intriguing one. Observations suggest nulling is prevalent in many long-period pulsars and must be understood to obtain a more comprehensive picture of pulsar emission and its evolution. One of the limitations in observational characterisation of nulling is the limited signal-to-noise, making individual pulses often not easily distinguishable from noise or any putative faint emission. Although some of the approaches in the published literature attempt to address this, they lose efficacy when individual pulses appear indistinguishable from the noise, and as a result, can lead to less accurate measurements. Here we develop a new method (the $\mathbb{N}$sum algorithm) that uses sums of pulses for better distinguishability from noise and thus measures the nulling fraction more robustly. It can be employed for measuring nulling fractions in weaker pulsars and observations with a limited number of observed pulses. We compare our algorithm with the recently developed Gaussian Mixture Modelling approach, using both simulated and real data, and find that our approach yields consistent results for generic and weaker pulsars. We also explore quasi-periodicity in nulling and measure the related parameters for five pulsars, including PSRs~J1453$-$6413, J0950$+$0755 and J0026$-$1955, for which these are also the first such measurements. We compare and contrast our analysis of quasi-periodic nulling with previously published work and explore the use of spin-down energy loss ($\dot E$) to distinguish between different types of modulation behaviour.