The Hubble/STIS Near-ultraviolet Transmission Spectrum of HD 189733b

TL;DR

This study presents near-ultraviolet transmission spectra of exoplanet HD 189733b using Hubble/STIS, revealing broad absorption features possibly linked to iron, with implications for atmospheric composition and mass-loss processes.

Contribution

First near-ultraviolet spectral analysis of HD 189733b with Hubble/STIS, identifying absorption features and constraining magnesium and iron presence in the atmosphere.

Findings

Broad absorption features at 2350 and 2600 Å suggest complex atmospheric composition.

No significant magnesium ion absorption detected at the core of resonance lines.

Hydrodynamic models cannot fully explain observed iron-related features without additional mechanisms.

Abstract

The benchmark hot Jupiter HD 189733b has been a key target to lay out the foundations of comparative planetology for giant exoplanets. As such, HD 189733b has been extensively studied across the electromagnetic spectrum. Here, we report the observation and analysis of three transit light curves of HD 189733b obtained with {\Hubble}/STIS in the near ultraviolet, the last remaining unexplored spectral window to be probed with present-day instrumentation for this planet. The NUV is a unique window for atmospheric mass-loss studies owing to the strong resonance lines and large photospheric flux. Overall, from a low-resolution analysis ($R=50$) we found that the planet's near-ultraviolet spectrum is well characterized by a relatively flat baseline, consistent with the optical-infrared transmission, plus two regions at $\sim$2350 and $\sim$2600 {\AA} that exhibit a broad and significant excess absorption above the continuum. From an analysis at a higher resolution ($R=4700$), we found that the transit depths at the core of the magnesium resonance lines are consistent with the surrounding continuum. We discarded the presence of \ion{Mg}{ii} absorption in the upper atmosphere at a $\sim$2--4$\sigma$ confidence level, whereas we could place no significant constraint for \ion{Mg}{i} absorption. These broad absorption features coincide with the expected location of \ion{Fe}{ii} bands; however, solar-abundance hydrodynamic models of the upper atmosphere are not able to reproduce the amplitude of these features with iron absorption. Such scenario would require a combination of little to no iron condensation in the lower-atmosphere, super-solar metallicities, and a mechanism to enhance the absorption features (such as zonal wind broadening). The true nature of this feature remains to be confirmed.