Context. The quadruple lensed system WGD2038-4008 (zs = 0.777 ± 0.001) has recently been discovered with the help of new techniques and observations. Black hole masses have been estimated for lensed quasars, but they have mostly been calculated for one broad emission line of one image. However, the images could be affected by microlensing, which changes the results.
Aims: We present black hole mass (MBH) estimations for images A and B of WGD2038-4008 using the three most prominent broad emission lines (Hα, Hβ, and Mg II) obtained in one single-epoch spectra. This is the first time the mass has been estimated in a lensed quasar in two images, allowing us to disentangle the effects of microlensing. The high S/N of our spectra allows us to get reliable results that can be compared with the existing data in the literature.
Methods: We used the X-shooter instrument mounted on the Very Large Telescope at Paranal Observatory to observe this system, taking advantage of its wide spectral range (UVB, VIS, and NIR). The sky emission correction was performed using principal component analysis as the nodding was small compared to the image separation. We compared the lines profiles to identify the microlensing in the broad-line region and corrected each spectra by the image magification. Using the flux ratio of the continuum to the core of the emission lines, we analyzed whether microlensing was present in the continuum source.
Results: We obtained MBH using the single-epoch method with the Hα and Hβ emission lines from the monochromatic luminosity and the velocity width. The luminosity at 3000 Å was obtained using the spectral energy distribution of image A, while the luminosity at 5100 Å was estimated directly from the spectra. The average MBH between the images obtained was log10(MBH/M⊙) = 8.27 ± 1.05, 8.25 ± 0.32, and 8.59 ± 0.35 for Mg II, Hβ, and Hα, respectively. We find Eddington ratios similar to those measured in the literature for unlensed low-luminosity quasars. Microlensing of −0.16 ± 0.06 mag in the continuum was found, but the induced error in the MBH is minor compared to that associated with the macromodel magnification. We also obtained the accretion disk size using the MBH for the three emission lines, obtaining an average value of log10(rs/cm)=15.3 ± 0.63, which is in agreement with theoretical estimates.