Axion search with telescope for radio astronomy (ASTRA): forecast for observations between 0.5 and 4~GHz

Abstract

Axion dark matter (DM) is predicted to convert into radio waves in neutron star magnetospheres. We assess the detectability of this signal using a 5 m radio telescope to be installed at the Fan Mountain Observatory, operating in the UHF, L- and S-bands from 0.5 to 4~GHz. We demonstrate that such a telescope can search new parameter space for axion-like particles over a broad range from $2\,μ\text{eV}<m_a<17\,μ\text{eV}$ for axion-photon couplings $g_{aγγ} \gtrsim 2\times 10^{-12}\text{ GeV}^{-1}$ with a three year observing period assuming the standard halo model -- improving neutron star observations by more than an order of magnitude. The search is broadband and is thus complementary to other techniques in the same frequency range. We describe in detail our neutron star population model, noise model, and proposed observing strategy. Most constraining power comes from neutron stars at the Galactic centre, where the smooth DM halo is densest. If a DM spike exists at the Galactic centre, the search is sensitive in the QCD axion model band. UHF and L-band observations (0.5 to 2~GHz) represent the pathfinder phase of a wider program we call ``Axion Search with Telescope for Radio Astronomy'' (ASTRA). Future higher mass searches aimed at discovery potential for the post-inflation axion require further hardware development to cover S, C, X and Ku bands (2 to 18~GHz).

Figures (6)

• Figure 1: Left: Spatial distribution of NSs generated by PsrPopPy with an age cut of $\leq10^7$ years. The NS population is predominantly old and is scattered throughout the MW galaxy. The apparent extension of the Galactic disk arises from NSs displaced to large radii by natal kicks. The inset shows a zoomed-in view of the GC, and shows the paucity of NSs near the GC in the PsrPopPy radial model of Ref. yusifov_revisiting_2004, illustrating the need for our separate GC model. Right: Intensity of the axion-photon conversion signal from the population of stars generated by PsrPopPy at 1 GHz, with signal smoothed using the telescope beam. The signal shows a strong enhancement at the GC due to the high DM density.
• Figure 2: Left: A model of the telescope we will use for ASTRA. Light from the sky is focused into the receiver with the 5-m parabolic reflector. The digital spectrometer hardware will be mounted in an electronics enclosure. More detail is given in the Supplementary Material. Right: An aerial view of Fan Mountain Observatory in Virginia, which is in the United States Radio Quiet Zone. The red dot marks the site for the ASTRA telescopes.
• Figure 3: The expected brightness temperature of the axion--photon conversion signal (solid) and the rms noise temperature (dotted) as a function of time for a representative day in January. Since the spiral arm survey continuously observes a sky patch that remains visible for the entire day, the resulting brightness temperature is independent of time. The silver patch shows the expected signal from the GC survey, which is substantially higher due to high DM density. For illustration, we assume that the axion has a mass consistent with each frequency shown: in reality the signal will occupy a single spectral channel for a one component axion DM model.
• Figure 4: Forecasted constraints on the axion--photon coupling compared with existing limits and forecasts. Our best forecast sensitivity across the full frequency range comes from the GC survey, which is assumed to be carried out for $\sim$3 hours per day for three years. The spiral arm survey assumes observing the best candidate NS (and all other starts in the model captured in the telescope beam) in a region that is almost always visible.
• Figure 5: Histogram showing the radial distribution of NS in the GC. The distribution follows the birth-rate prescription given in Eq. (\ref{['eq:BL NS birthrate']}).