Accretion Geometry of the New Galactic Black Hole Candidate AT2019wey in the Hard State

TL;DR

This study investigates the accretion geometry of the Galactic black hole candidate AT2019wey during its 2022 hard-state outburst using quasi-simultaneous broadband data from NICER, Swift/XRT, and NuSTAR. The authors find that a single Comptonizing region cannot reproduce the observed spectrum and instead require two distinct Comptonizing components, each with associated reflection, plus thermal disk emission, consistently placing the inner disk at $R_{ m in} \sim 16-56\;r_{\rm g}$ and the source at $L \sim 1.9\%\,L_{\rm Edd}$. Timing analysis reveals strong broadband noise with hard lags at low frequencies and a frequency-dependent lag-energy spectrum that shows a log-linear trend above 1 keV and a low-energy turnover consistent with thermal reverberation, supporting a propagating-fluctuation picture and disk–corona coupling in the hard state. The results reinforce a dual-corona geometry in which the hard Comptonizing region dominates the flux and produces distant reflection, while a cooler, optically thick region contributes less to the overall emission. Together, these findings advance our understanding of disk truncation and corona-disk interactions in low-mass X-ray binaries at relatively low Eddington luminosities, with implications for the interpretation of hard-state variability and reverberation features in BHXRBs.

Abstract

We perform broadband spectral and timing studies of the Galactic low-mass black hole candidate AT2019wey using quasi-simultaneous NICER, Swift, and NuSTAR observations obtained in 2022. The long-term MAXI light curve, along with the hardness-intensity diagram (HID), indicates that the source remained in the hard state and did not switch to the soft state. Spectral modeling using two different model combinations reveals that the broadband spectrum is best described by two distinct Comptonizing regions, associated reflection components, and thermal emission from the disk. The harder Comptonizing region dominates ($\gtrsim80\%$) the total flux and is primarily responsible for the observed reflection features from the distant part of the disk. We find that the accretion disk is truncated at a radius of $\sim16-56~r_{\rm{g}}$, while the luminosity is $\sim1.9\%$ of the Eddington limit, assuming a black hole mass of $10 ~ M_\odot$ and distance of 8 kpc. Our spectral results also show consistency in the estimated inner disk radius obtained through two independent methods: modeling the disk continuum and the reflection spectrum. The variability studies imply the presence of intrinsic disk variability, likely originating from an instability in the disk. We also detect hard time lags at low frequencies, possibly arising from the inward propagation of mass accretion rate fluctuations from the outer to the inner regions of the accretion disk. Moreover, an observed deviation of the lag-energy spectrum from the log-linear trend at $\lesssim 0.7$ keV is most likely attributed to thermal reverberation, arising from the reprocessing of hard coronal photons in the accretion disk.

Figures (9)

• Figure 1: Long-term MAXI lightcurve of AT2019wey in the $2-20$ keV band binned with 40 days, showing the entire duration of the source activity including the onset of the first outburst, a brief quiescent period, and resumption in early 2022 lasting for another $\sim 3$ years. The vertical lines indicate the timings of the observations used in this work. The shaded region marks the period from the onset to the decay of the outburst, during which the HID is derived (see Figure \ref{['fig:hid_nicer']}).
• Figure 2: HID of AT2019wey constructed using MAXI observations (red) taken during the shaded interval shown in the MAXI light curve in Figure \ref{['fig:maxi_ltcrv']}. The MAXI data have been binned for clarity. Black circles show the 'q' - shaped HID of GX 339--4, derived from MAXI observations covering its full or successful outburst in 2021.
• Figure 3: Residual of NuSTAR FPMA spectral data in $3-80$ keV to the fit with tbfeo(powerlaw+diskbb). Spectral fitting is performed only in the $3-5$ and $9-12$ keV bands. Presence of the iron line excess and reflection hump are clearly visible in both the dataset.
• Figure 4: Left panel: Joint Swift/XRT (red), Nustar/FPMA (black), and FPMB (blue) spectral data fitted with Model-1. Right panel: Joint NICER (red), NuSTAR/FPMA (black), and FPMB (blue) spectral data fitted with the same model. In both panels, solid lines represent the extrapolated models of the continuum fit. The dashed blue and magenta curves denote the hard and soft Comptonizing components, respectively. The black dashed curve corresponds to the disk emission, while the grey dashed line indicates the Gaussian component.
• Figure 5: Left panel: Joint Swift/XRT (red), Nustar/FPMA (black), and FPMB (blue) spectral data fitted with Model-2. Right panel: Joint NICER (red), NuSTAR/FPMA (black), and FPMB (blue) spectral data fitted with the same model. In both panels, solid lines represent the extrapolated models of the continuum fit. The black dashed curve shows the thermal Comptonization from the hard Comptonizing region, while the blue dashed curves represent the corresponding reflection component. The magenta dashed curve denotes the thermal Comptonization and associated reflection from the soft Comptonizing region.