Measuring pulsar profile variations with 2D Gaussian process regression

Researchers using advanced Gaussian process regression techniques have identified significant pulsar profile variations in 21 out of 26 observed pulsars, with 11 showing previously documented correlations to spin-down rates and seven exhibiting new evidence of such correlations. The study, conducted at Jodrell Bank Observatory, employed a computationally efficient 2D Gaussian process model to enhance sensitivity by simultaneously tracking time- and phase-correlated signals in pulsar magnetospheres. Notably, PSR B0740–28 displayed complex profile shape changes that switch between quasi-stable states, challenging current understanding of the underlying physical processes.

Keith, M. J., Spiewak, R., Lyne, A. G., Weltevrede, P., Antonopoulou, D., & Stappers, B. (2025). Measuring pulsar profile variations with 2D Gaussian process regression. Monthly Notices of the Royal Astronomical Society, 540(3), 2486–2505. https://doi.org/10.1093/mnras/staf886