A spectral survey of WASP-19b with ESPRESSO
High-resolution precision spectroscopy provides a multitude of robust techniques for probing exoplanetary atmospheres. We present multiple VLT/ESPRESSO transit observations of the hot-Jupiter exoplanet WASP-19b with previously published but disputed atmospheric features from lowresolution studies. Through spectral synthesis and modelling of the Rossiter-McLaughlin (RM) effect we calculate stellar, orbital and physical parameters for the system. From narrow-band spectroscopy we do not detect any of HI, FeI, MgI, CaI, NaI, and KI neutral species, placing upper limits on their line contrasts. Through cross-correlation analyses with atmospheric models, we do not detect Fe I and place a 3σ upper limit of log (XFe/X⊙) ≈ -1.83 ± 0.11 on its mass fraction, from injection and retrieval. We show the inability to detect the presence of H2O for known abundances, owing to lack of strong absorption bands, as well as relatively low S/N ratio. We detect a barely significant peak (3.02±0.15 σ) in the cross-correlation map for TiO, consistent with the sub-solar abundance previously reported. This is merely a hint for the presence of TiO and does not constitute a confirmation. However, we do confirm the presence of previously observed enhanced scattering towards blue wavelengths, through chromatic RM measurements, pointing to a hazy atmosphere. We finally present a reanalysis of low-resolution transmission spectra of this exoplanet, concluding that unocculted starspots alone cannot explain previously detected features. Our reanalysis of the FORS2 spectra of WASP-19b finds a ~100× sub-solar TiO abundance, precisely constrained to log XTiO ≈ -7.52 ± 0.38, consistent with the TiO hint from ESPRESSO. We present plausible paths to reconciliation with other seemingly contradicting results. © 2021 The Author(s). ; We acknowledge the use of the following PYTHON packages in addition to those explicitly mentioned in the text: NUMPY (Harris et al. 2020), MATPLOTLIB (Hunter 2007), SCIPY (Virtanen et al. 2020), PANDAS (McKinney et al. 2010), EMCEE (Foreman-Mackey et al. 2013), SCIKIT-LEARN (Pedregosa et al. 2011), NUMBA (Lam, Pitrou & Seibert 2015), CORNER (Foreman-Mackey 2016), and ASTROPY (Price-Whelan et al. 2018). This work has made use of the VALD database, operated at Uppsala University, the Institute of Astronomy RAS in Moscow, and the University of Vienna. ES acknowledges support from ESO through its fellowship program. RB acknowledges support from FONDECYT Project 11200751, and CORFO project No. 14ENI2-26865. RB acknowledges support from ANID – Millennium Science Initiative – ICN12_009. ; With funding from the Spanish government through the Severo Ochoa Centre of Excellence accreditation SEV-2017-0709. ; Peer reviewed
