Searching for Long-Period Radio Transients in ASKAP EMU Data with 10-Second Imaging

TL;DR

This work targets long-period transients (LPTs) by performing a large-scale image-plane search with 10-second ASKAP EMU imaging, pushing into the minute-to-hour timescale that traditional time-series and deep-imaging surveys miss. The authors implement sky-model subtraction via CASA, create model-subtracted visibilities, and analyze 3600+ short-interval images per beam with the VASTER pipeline, applying dual statistical tests and a robust candidate-classification scheme. Although no new LPTs are found, six stellar radio flares are identified and characterized, and the study derives a 10-second transient surface density limit of $\rho = 2.21^{+2.60}_{-1.40}\times10^{-6}\ \rm deg^{-2}$ with a sensitivity of $14.6$ mJy, alongside luminosity limits for potential detections. The results demonstrate the feasibility of minute-to-hour transient searches with ASKAP and provide concrete guidance for pipeline improvements (e.g., ML-based artefact filtering, joint Stokes I/V analyses) and observational strategies to enhance LPT detectability in future surveys.

Abstract

Long-period radio transients (LPTs) are a recently identified phenomenon that challenge our current understanding of compact objects and coherent radio emission mechanisms. These objects emit radio pulses similar to those of pulsars, but at much longer periods -- on the order of minutes to hours. With duty cycles of only a few percent, individual pulses have been observed to last between 10 and 1000 seconds. This places LPTs in a timescale gap between the two main techniques used in transient radio searches: time-series analysis at millisecond to second timescales, and image-plane searches sensitive to variability on the scale of days. As a result, LPTs remained undetected until recently, and only a handful are currently known. To increase the sample of known LPTs, we conducted a dedicated search using 200 hours of archival data from the ASKAP Evolutionary Map of the Universe survey, covering 750 deg$^2$ of sky at the shortest possible imaging time step of 10-seconds. This represents the first large-scale search using ASKAP data at second-scale resolution. Although no LPTs were detected, we identified flares from six stars, at least one had never been detected in the radio regime before. We placed a lower limit on the transient surface density of $2.21\times10^{-6}$ deg$^{-2}$ at a 10-second timescale, with a sensitivity of 16.9 mJy. Our findings evaluate the feasibility of detecting radio transients using 10-second imaging with ASKAP and provide insights into improving detection pipelines and observation strategies for LPTs.

Figures (11)

• Figure 1: EMU fields used in this study, along with the positions of pulsars, magnetars, and LPTs. We plot the figure in Galactic coordinates using the Mollweide projection and focus on the Galactic plane. The blue patches indicate the EMU fields processed in this study, which have Galactic latitudes of $|b|<10\degree$. The background shows the Galactic emission at 887.5 MHz modelled by GalacticRadioEmission. Only white dwarf binaries with periodic radio emission are included in this plot. We also plotted the positions of pulsars and magnetars according to the ATNF Pulsar Catalogue (version 1.70) ATNFPulsarCatalogue and the McGill Online Magnetar Catalog MagnetarCatalogue, respectively.
• Figure 2: Flowchart of the candidate classification pipeline. olive arrows indicate a true outcome, while red arrows indicate a false outcome. Candidates that pass only one of the two statistical tests are convolved with five boxcar filters using a 5$\sigma_{\rm conv}$ threshold. Candidates that pass both tests ($\sigma_{\rm log~\eta} > 6$ and $\sigma_{\rm log~S/N} > 6$) are first filtered based on their median flux density. Those with median flux below 10 mJy are further convolved using a 10$\sigma_{\rm conv}$ threshold. Candidates classified as Transients will be inspected manually while the rest are discarded. Figure \ref{['fig:lcexample']} shows two example light curves for each type of candidates.
• Figure 3: Light curves and dynamic spectra of Beta Centauri (left) and HD 105386 (right). In both plots, the top panel shows the light curve averaged in 5-minute intervals, while the middle and bottom panels display the dynamic spectra of the Stokes I and Stokes V parameters, respectively. In panel (a), the observation ended before the full transient event could be captured. While low-level emission appears prior to the peak, its flux density is comparable to the noise level, and the Stokes V dynamic spectrum does not confirm whether these features are genuine. This transient is only detected in the lower-frequency band. HD 105386 exhibits a double-peaked structure separated by approximately three hours, with strong circular polarisation and a sign reversal between the peaks.
• Figure 4: Light curves and dynamic spectra of Gaia DR3 5853594572486546176 (left) and HD 110244 (right), with panel arrangements matching those in Figure \ref{['fig:combined_2']}. The transient associated with Gaia DR3 5853594572486546176 is 50% circularly polarised and shows a significant linear polarisation fraction, with a pulse width exceeding four hours. The transient from HD 110244 spans nearly six hours and exhibits a strong circular polarisation fraction, with emission confined to the higher-frequency bands.
• Figure 5: Light curve and dynamic spectrum of IO Vel. The top panel shows the light curve averaged in 2-minute time steps to improve visibility of the pulse, with total intensity in black, linear polarisation in red, and circular polarisation in blue. The middle and bottom panel shows the dynamic spectrum of the Stokes I and Stokes V parameter, respectively. The pulse width is around 100 minutes and the circular polarisation is 77%.