Artificial Precision Polarization Array: Sensitivity for the axion-like dark matter with clock satellites

TL;DR

The paper tackles the challenge of detecting ultralight axion-like DM via axion-induced birefringence by proposing the Artificial Precision Polarization Array (APPA), a space-based network of clocked transmitters and a central receiver near the Sun–Earth L2 point. It develops a dual-analysis framework—Bayesian upper limits and a frequentist Generalized Lomb-Scargle Periodogram (GLSP)—to quantify sensitivity to the axion-photon coupling $g_{a\gamma}$ in the mass window $m_a \sim \mathcal{O}(10^{-22}-10^{-19})$ eV. The Bayesian approach leverages a Gaussian likelihood with cross-covariance $\mathbf{\Sigma}$ and a $\langle q \rangle = 2.71$ criterion to yield a 95% CL upper bound $g_{95\%}$, while the GLSP-based frequentist analysis demonstrates improved detectability due to regular space-based sampling that reduces spectral leakage. The results indicate that APPA can achieve tighter $g_{95\%}$ constraints and enhanced detection sensitivity compared with ground-based observations, highlighting APPA’s potential to advance axion DM searches and motivate future space-based polarization timing initiatives.

Abstract

The approaches to searching for axion-like signals based on pulsars include observations with pulsar timing arrays (PTAs) and pulsar polarization arrays (PPAs). However, these methods are limited by observational uncertainties arising from multiple unknown and periodic physical effects, which substantially complicate subsequent data analysis. To mitigate these issues and improve data fidelity, we propose the Artificial Pulsar Polarization Arrays (APPA): a satellite network comprising multiple pulsed signal transmitters and a dedicated receiver satellite. In order to constrain the axion-photon coupling parameter $g_{aγ}$, we generate simulated observations using Monte Carlo methods to investigate APPA's sensitivity via two complementary approaches: Bayesian analysis and frequentist analysis. Simulations indicate that for axion mass $m_{a}\sim\mathcal{O}\big(10^{-22}-10^{-19}\big)$ eV, APPA yields a better upper limit on $g_{aγ}$ (at the 95\% confidence level) than conventional ground-based observations and achieves better detection sensitivity.

Figures (9)

• Figure 1: Conceptual diagram for APPA. The satellite network consists of a single receiving satellite $O$ and several transmitting satellites $S_{i} (i=1,...,6)$, each equipped with an ultra precision clock. This diagram is purely illustrative and does not represent the actual physical configuration.
• Figure 2: Bayesian 95% C.L. upper limit $g_{95\%}$ on the function $g_{a\gamma}(m_{a})$, which is calculated from $\langle q \rangle =2.71$ and Eq. (\ref{['q']}).
• Figure 3: Numerical simulation: The GLSP (blue curve) derived from simulated ground-based radio telescope data. The simulated dataset contains only Gaussian white noise with $\text{S/N}=1$. The vertical dashed line marks the target signal frequency corresponding to $m_{a}=10^{-22}\mathrm{eV}$. The red horizontal line indicates the 10% (FAP) threshold. Since the peak at the target frequency lies below this threshold, the likelihood of the detected peak representing a genuine signal is less than 90%.
• Figure 4: Numerical simulation: The GLSP (blue curve) derived from APPA observational data. The simulated signal includes only Gaussian white noise with S/N=1. The vertical dashed line marks the target signal frequency of the axion with mass $m_{a}=10^{-22} \mathrm{eV}$. The horizontal lines indicate FAP levels of 10% (red), 5% (yellow), and 1% (green).
• Figure 5: Detection range and sensitivity curve of the APPA satellite network. The detection sensitivity is set to $\phi_{a}=0.01^{\circ}$ in Eq. (\ref{['deltaphisim']}), consistent with current observational limits on linear polarization angle rotation, with minor fluctuations attributed to instrumental noise. The black dashed line marks the critical boundary beyond which the condition $\mathrm{cos}(m_{a}T)\simeq1$ is no longer valid.