On the distance to the black hole X-ray binary Swift J1727.8$-$1613

TL;DR

This study integrates MeerKAT HI absorption measurements with HST near-UV reddening to constrain the distance to Swift J1727.8−1613, a dynamically confirmed black hole X-ray binary. The HI data set a near-side lower bound of $d_{\rm near} = 3.6$ kpc, while the near-UV reddening $E(B-V) = 0.37$ leads, under a K4$\pm$1V donor assumption, to a best-distance estimate of $d \approx 5.5$ kpc and a natal kick of $v_{\rm kick} \approx 190$ km s$^{-1}$. The results favor a Galactic-disk origin with an asymmetric supernova birth, though the exact distance remains sensitive to donor evolution and system inclination. The work highlights the complementary power of HI-absorption and reddening analyses for rapid distance constraints to Galactic transients and underscores the need for improved measurements of the donor star and system inclination to tighten the distance. Overall, the paper provides a coherent, multi-method distance framework that informs jet energetics, Eddington fractions, and the natal-kick history for this prominent BH X-ray binary.

Abstract

We review the existing distance estimates to the black hole X-ray binary Swift J1727.8$-$1613, present new radio and near-UV spectra to update the distance constraints, and discuss the accuracies and caveats of the associated methodologies. We use line-of-sight HI absorption spectra captured using the MeerKAT radio telescope to estimate a maximum radial velocity with respect to the local standard of rest of $24.8 \pm 2.8 \, {\rm km\,s^{-1}}$ for Swift J1727.8$-$1613, which is significantly lower than that of a nearby extragalactic reference source. From this we derive a near kinematic distance of $d_{\rm near} = 3.6 \pm 0.3 \, ({stat}) \pm 2.3 \, ({sys}) \, {\rm kpc}$ as a lower bound after accounting for additional uncertainties given its Galactic longitude and latitude, $(l, b) \approx (8.6^{\circ}, 10.3^{\circ})$. Near-UV spectra from the Hubble Space Telescope's Space Telescope Imaging Spectrograph allows us to constrain the line-of-sight colour excess to $E(B\!-\!V) = 0.37 \pm 0.01 \, ({stat}) \pm 0.025 \, ({sys})$. We then implement this in Monte Carlo simulations and present a distance to Swift J1727.8$-$1613 of $5.5^{+1.4}_{-1.1} \, {\rm kpc}$, under the assumption that the donor star is an unevolved, main sequence K3-5V star. This distance implies a natal kick velocity of $190 \pm 30 \, {\rm km\,s^{-1}}$ and therefore an asymmetrical supernova explosion within the Galactic disk as the expected birth mechanism. A lower distance is implied if the donor star has instead lost significant mass during the binary evolution. Hence, more accurate measurements of the binary inclination angle or donor star rotational broadening from future observations would help to better constrain the distance.

Figures (5)

• Figure 1: Our H i absorption spectra along of the line of sight towards J1727. The top two plots are the 32k-S spectra, and the bottom two plots are the 32k-Z spectra, with LSR velocity bin widths that are 5.6 and 0.7 km s$^{-1}$ wide respectively. The dashed lines indicate the origins of the $x$- and $y$-axes. The $y$-axis represents the residuals after subtracting the continuum emission from the data, with $3\sigma$ uncertainties indicated by the shaded areas, both of which are presented as a percentage of the continuum flux density of the source. The vertical dotted lines indicate the maximum velocity taken as the mid-point of the bin at which significant ($>\!3\sigma$) absorption is observed. In chronological order, these maximum velocities are $19.2 \pm 2.8$, $6.1 \pm 0.4$, $24.8 \pm 2.8$, and $18.0 \pm 0.4$ km s$^{-1}$. The significantly greater SNR of the 14-10-2023 spectrum meant we were able to observe significant ($>\!3\sigma$) H i absorption to greater velocities than observed in the 32k-Z spectra.
• Figure 2: A consolidated view of the H i spectra for both the target, J1727 (red, left), and our reference source, J1733 (blue, right). We observed no significant ($>\!3\sigma$) absorption outside the velocities chosen as the $x$-axis range. The $y$-axis represents the H i absorption percentage. The dashed lines indicate the $x$- and $y$-axes origins. The vertical dotted lines indicate the maximum velocity at which significant absorption is observed. For J1727, these are $24.8 \pm 2.8$ and $18.7 \pm 0.4$ km s$^{-1}$. For J1733, these are $30.4 \pm 2.8$ and $32.7 \pm 0.4$ km s$^{-1}$. The top-left plot is the 32k-S spectrum for J1727 from 14-10-2023, while the bottom-left plot is the mean weighted J1727 spectrum achieved by combining both 32k-Z spectra from Figure \ref{['fig:HI_Spectra_SwiftJ1727']}. We generated the 32k-Z (32k-S) spectra for J1733 from observations on 06-09-2023 (14-10-2023) with the peak J1733 flux density measured to be $6.1 \pm 0.1$ Jy ($6.0 \pm 0.1$ Jy), providing the significantly greater SNR compared to that of the J1727 observations.
• Figure 3: The near-UV spectrum of J1727 as seen by HST/STIS (black), with a reddened power-law fit (green) to the data (Castro Segura et al. in prep.) The uncertainty around the fit is indicated as a shaded region. The vertical shaded regions were masked during the fit to avoid Fe and Mg lines that may influence the fit.
• Figure 4: Estimates of $E(B\!-\!V)$ for distances between $0 \text{--} 7.5$ kpc along the line of sight to J1727. The 3D Galactic dust maps include BayestarGreen_etal_2019_3D_dust_map_Gaia_Pan-STARRS1_2MASS and GaiaEdenhofer_etal_2024_pc_scale_3D_dust_map. We also use an update of the 2D "SFD" dust map Schlegel_Finkbeiner_David_1998_SFD_dust_map_via_IR_emission_for_reddening_and_CMBRFs to remove extragalactic contamination that is referred to as the "corrected SFD" Chiang_2023_Corrected_SFD_CSFD_dust_map_minimal_extragalactic_contamination. To investigate the likely maximum $E(B\!-\!V)$ values within this distance range, we use CSFD; as this only reports the total value for $E(B\!-\!V)$ along the line of sight, it serves as a good estimate of the maximum $E(B\!-\!V)$ in the direction of J1727 at high distances. From this we derive a maximum $E(B\!-\!V) \approx 0.4$. Despite each having significant caveats and the disagreement between Bayestar and Gaia in how gas is distributed along the line of sight, both appear to accumulate to roughly this maximum value. Lastly, our estimate of $E(B\!-\!V)$ (pink dotted line) is derived from our near-UV observations. We use this estimate to conclude that values of $E(B\!-\!V) \gg 0.4$ may not be reliable, and suggest a distance to J1727 of $5.5^{+1.4}_{-1.1}$ kpc with $1\sigma$ uncertainties, which are included for reference (orange dot-dash line and shaded region respectively).
• Figure 5: The distance posterior distribution produced from Monte Carlo calculations using the near-UV constraint on $E(B\!-\!V)$. The solid vertical line represents the 0.5000 quantile and the dashed vertical lines represent the 0.1585 and 0.8415 quantiles.