Algorithmic Pulsar Timing

TL;DR

The paper introduces the Algorithmic Pulsar Timer (APT), an automated method that accurately performs pulsar phase connection and timing, reducing manual effort and enabling scalable analysis of pulsar data.

Contribution

This work presents APT, the first algorithm that automates pulsar phase connection using statistical tests and decision logic, achieving high success on simulated data.

Findings

99% success rate on 100 simulated systems

Automates the traditionally manual phase connection process

Sets foundation for automated binary pulsar timing

Abstract

Pulsar timing is a process of iteratively fitting pulse arrival times to constrain the spindown, astrometric, and possibly binary parameters of a pulsar, by enforcing integer numbers of pulsar rotations between the arrival times. Phase connection is the process of unambiguously determining those rotation numbers between the times of arrival (TOAs) while determining a pulsar timing solution. Pulsar timing currently requires a manual process of step-by-step phase connection performed by individuals. In an effort to quantify and streamline this process, we created the Algorithmic Pulsar Timer, APT, an algorithm which can accurately phase connect and time isolated pulsars. Using the statistical F-test and knowledge of parameter uncertainties and covariances, the algorithm decides what new data to include in a fit, when to add additional timing parameters, and which model to attempt in subsequent iterations. Using these tools, the algorithm can phase-connect timing data that previously required substantial manual effort. We tested the algorithm on 100 simulated systems, with a 99% success rate. APT combines statistical tests and techniques with a logical decision-making process, very similar to the manual one used by pulsar astronomers for decades, and some computational brute-force, to automate the often tricky process of isolated pulsar phase connection, setting the foundation for automated fitting of binary pulsar systems.