Aql X-1 from dawn 'til dusk: the early rise, fast state transition and decay of its 2024 outburst

TL;DR

This work presents a dense, multiwavelength campaign of Aql X-1's 2024 outburst, leveraging EP's early X-ray sensitivity alongside NICER, NuSTAR, Swift, and LCO data to trace the rise, peak, and decay. By combining phenomenological rise analyses and Gaussian Process modeling, the authors constrain the optical-to-X-ray delay to be at most about 3 days, with the X-ray onset likely occurring after optical activation. Broadband spectroscopy across epochs reveals a rapid hard-to-soft transition (~12 hours) accompanied by the emergence of boundary/spreading layers, a shrinking NS-emitting region, and corona cooling, followed by a prolonged soft-state plateau and eventual decay back to quiescence. The results provide direct insight into the time scales and physical processes governing state transitions in neutron-star LMXBs and highlight the power of early-rise, multi-band campaigns in constraining accretion geometry at low luminosities.

Abstract

Transient Low-Mass X-ray Binaries (LMXBs) are usually first detected by all-sky X-ray monitors when they enter new outbursts, typically at X-ray luminosities above $\sim$10$^{36}$ erg/s. Observations of these sources during the early rise of the outbursts have so far been very limited. However, the launch of the Einstein Probe (EP) has greatly improved our ability to detect fainter X-ray activity, unlocking access to the outburst early rise. In September 2024, EP detected the early onset of a new outburst from the neutron star LMXB Aql X-1, catching the source at a luminosity below 10$^{35}$ erg/s. In this paper we present results from a comprehensive, multi-wavelength campaign of this event, combining data from EP, NICER, NuSTAR, Swift and Las Cumbres Observatory covering the full outburst from its early rise through its decay. By comparing X-ray and optical light curves obtained with Las Cumbres Observatory during the initial rise, we show that the start of the X-ray emission lagged the optical rise by, at most, 3 days. Time-resolved X-ray spectroscopy revealed how the geometry and the physical properties of the accretion flow evolve during this early stage of the outburst, as well as at higher luminosities as the source transitioned through the canonical X-ray spectral states - hard, intermediate and soft. These data show that the source underwent a very rapid, about 12-h long, transition from the hard to the soft state about two weeks after the optical onset of the outburst. The evolution of the temperature and physical sizes of both the inner region of the disk and a black body near the NS surface suggest that at the state transition, a boundary and spreading layer likely formed. We discuss these results in the context of time-scales for outburst evolution and state transitions in accreting neutron stars and black holes.

Figures (10)

• Figure 1: EP/WXT light curve of the 2024 outburst of Aql X-1. The luminosity has been extrapolated to the 0.5-10 keV range (see text for more details). The quiescence level, measured by Cackett2011, is marked with a horizontal solid cyan line, while the vertical dotted gray line indicates $T_0$, the time of the first bright (significance above 3-$\sigma$) detection by EP. The first two points correspond to non-detections and the corresponding upper limits are marked with vertical downwards arrows. The third and fourth points are instead marginal detections (significance between 2 and 3-$\sigma$) and are therefore represented with circular points but with smaller size compared to the remaining points.
• Figure 2: Multi-band light curves of the outburst rise of Aql X-1, including: EP/WXT (blue) and MAXI (magenta) in the Top panel, Swift/UVOT in different filters in the Middle panel and LCO in the Bottom panel. The horizontal solid lines in the Top panels are set at the equivalent WXT and MAXI count-rates for the quiescent luminosity in Cackett2011. The phenomenological trends estimated for each data set are shown with superimposed dashed lines. Upper limits are marked with vertical downwards arrows.
• Figure 3: NICER light curve (Top) and HID (bottom). The total count rate has been extracted in the 1-10 keV band. The hardness is defined as the ratio between the hard band (6-10 keV) and the soft band (2-3.5 keV) count rates. A color map going from blue to red and to green is used to indicate the time evolution. The vertical lines in the top panel mark the times of the NuSTAR (dashed) and EP/FXT (dotted) observations. The labels on top of selected circular data points in the bottom panel indicate the corresponding Epoch.
• Figure 4: Broadband NICER (different shades of blue to distinguish between the NICER spectra used in the same Epcoh) and NuSTAR (different shades of green to distinguish FPMA and FPMB) spectra for Epoch 3 (left, fitted with Model H) and Epoch 28 (right, fitted with Model S) and residuals. Different line styles were adopted to distinguish between the different components: dash for diskbb, dot for thcomp$\times$bbodyrad and dash-dot for relxillCp (left panel) and relxillNS (right panel).
• Figure 5: Evolution of the main spectral-timing parameters of Aql X-1 during the 2024 outburst. In the top six panels (Luminosity in the 1 - 10 keV abnd, $\Gamma$ index of the comptonizing medium, temperature and radius of the black body component, temperature and radius of the disk component), each point corresponds to a different Epoch. In the bottom panel (fractional rms), we distinguish instead between values obtained for NICER (black diamonds) and EP/FXT (gray circles). The fractional rms has been calculated considering the 0.01-1000 Hz frequency range and the 1-10 keV energy range. Parameters that were kept frozen during the fit are reported as gray diamonds.