= Solution
In the 2016 observational context, appropriate choices are as follows.
For atomic species, use optical transit spectra of Na I near \b[$0.589\,\mu\mathrm m$] and K I near \b[$0.770\,\mu\mathrm m$], with <Hubble Space Telescope>/STIS or ground-based high-resolution instruments on <Very Large Telescope>/<Keck telescope>-class telescopes. Neutral/ionized ultraviolet metal lines offer additional diagnostics with <Hubble Space Telescope>. Resolving a line's shape and distinguishing stellar contamination matters as much as locating its wavelength.
For molecules, use near/mid-IR spectroscopy: <water> bands near \b[$1.4$, $1.9$ and $2.7\,\mu\mathrm m$], CO near \b[$2.3$ and $4.6\,\mu\mathrm m$], <methane> near \b[$3.3\,\mu\mathrm m$], and $\mathrm{CO}_2$ near \b[$4.3$ or $15\,\mu\mathrm m$]. <Hubble Space Telescope>/WFC3 samples <water> around $1.1$--$1.7\,\mu\mathrm m$; ground-based <Very Large Telescope>/CRIRES or <Keck telescope>/NIRSPEC spectra target accessible near-IR windows such as CO at $2.3\,\mu\mathrm m$. Warm <Spitzer Space Telescope> provides $3.6$ and $4.5\,\mu\mathrm m$ photometry, while cryogenic <Spitzer Space Telescope> spectra/longer-wavelength data are archival by 2016. A broadband eclipse point alone does not uniquely identify a molecule.
For <atmospheric escape>, use ultraviolet transit spectroscopy of H I Ly$\alpha$ at \b[$121.6\,\mathrm{nm}$], and suitable C/O/metal lines, with <Hubble Space Telescope>/STIS or COS. Extended, velocity-shifted absorption beyond the optical planet suggests an escaping envelope, although stellar variability, geocoronal emission and interstellar absorption require treatment. Ground-based H$\alpha$ at $656.3\,\mathrm{nm}$ probes excited hydrogen and is a complementary rather than interchangeable measure of neutral-hydrogen escape.
Back to article page
