An Improved Spin-Down Rate for the Proposed White-Dwarf Pulsar AR Scorpii

TL;DR

This study refines the white dwarf's spin-down rate in AR Scorpii, confirms its increasing spin period, and investigates color variations and timing offsets of the pulses, supporting the synchrotron emission model.

Contribution

It provides an improved measurement of the white dwarf's spin-down rate and analyzes color and timing variations, offering insights into emission mechanisms and observational corrections.

Findings

White dwarf spin-down rate measured with 4% uncertainty

Color index of primary pulses varies significantly across orbit

Wavelength-dependent timing offsets can reach 6 seconds

Abstract

We analyze rapid-cadence, multiwavelength photometry of AR Scorpii from three observatories, covering five observing seasons. We measure the arrival times of the system's beat pulses and use them to compute an updated ephemeris. The white dwarf spin-down rate is estimated with an uncertainty of only 4%. These results confirm, beyond any doubt, that the white dwarf's spin period is increasing at the rate consistent with by that of Stiller et al. (2018). We study the evolution of the beat pulse's color index across the orbit. The color of the primary pulse maxima varies significantly across the orbit, with the peaks being bluer after superior conjunction than in the first half of the orbit. Specifically, at orbital phase 0.5, the color index of the primary pulse shows a very sharp discontinuity towards bluer indices. This supports the Potter & Buckley (2018b) synchrotron emission model where the two emitting poles differ significantly in color. However, no corresponding jump in the color of the secondary pulses is seen. Furthermore, our analysis reveals that the arrival times of the pulses can differ by as much as 6s in simultaneous $u$ and $r$ photometry, depending on the binary orbital phase. If left uncorrected, this wavelength-dependent timing offset could lead to erroneous measurements of the spin-period derivative, particularly with heterogeneous datasets.