Why is it so Cold in Here?: Explaining the Cold Temperatures Retrieved from Transmission Spectra of Exoplanet Atmospheres. (arXiv:2003.11548v1 [astro-ph.EP])
<a href="http://arxiv.org/find/astro-ph/1/au:+MacDonald_R/0/1/0/all/0/1">Ryan J. MacDonald</a>, <a href="http://arxiv.org/find/astro-ph/1/au:+Goyal_J/0/1/0/all/0/1">Jayesh M. Goyal</a>, <a href="http://arxiv.org/find/astro-ph/1/au:+Lewis_N/0/1/0/all/0/1">Nikole K. Lewis</a>
Transmission spectroscopy is a powerful technique widely used to probe
exoplanet terminators. Atmospheric retrievals of transmission spectra are
enabling comparative studies of exoplanet atmospheres. However, the atmospheric
properties inferred by retrieval techniques display a significant anomaly: most
retrieved temperatures are far colder than expected. In some cases, retrieved
temperatures are ~1000 K colder than T_eq. Here, we provide an explanation for
this conundrum. We demonstrate that erroneously cold temperatures result when
1D atmospheric models are applied to spectra of planets with differing
morning-evening terminator compositions. Despite providing an acceptable fit,
1D retrieval techniques artificially tune atmospheric parameters away from
terminator-averaged properties. Retrieved temperature profiles are hundreds of
degrees cooler and have weaker temperature gradients than reality. Retrieved
abundances are mostly biased by > 1$sigma$ and sometimes by > 3$sigma$, with
the most extreme biases for ultra-hot Jupiters. When morning-evening
compositional differences manifest for prominent opacity sources, H$_2$O
abundances retrieved by 1D models can be biased by over an order of magnitude.
Finally, we demonstrate that these biases provide an explanation for the cold
retrieved temperatures reported for WASP-17b and WASP-12b. To overcome biases
associated with 1D atmospheric models, there is an urgent need to develop
multidimensional retrieval techniques.
Transmission spectroscopy is a powerful technique widely used to probe
exoplanet terminators. Atmospheric retrievals of transmission spectra are
enabling comparative studies of exoplanet atmospheres. However, the atmospheric
properties inferred by retrieval techniques display a significant anomaly: most
retrieved temperatures are far colder than expected. In some cases, retrieved
temperatures are ~1000 K colder than T_eq. Here, we provide an explanation for
this conundrum. We demonstrate that erroneously cold temperatures result when
1D atmospheric models are applied to spectra of planets with differing
morning-evening terminator compositions. Despite providing an acceptable fit,
1D retrieval techniques artificially tune atmospheric parameters away from
terminator-averaged properties. Retrieved temperature profiles are hundreds of
degrees cooler and have weaker temperature gradients than reality. Retrieved
abundances are mostly biased by > 1$sigma$ and sometimes by > 3$sigma$, with
the most extreme biases for ultra-hot Jupiters. When morning-evening
compositional differences manifest for prominent opacity sources, H$_2$O
abundances retrieved by 1D models can be biased by over an order of magnitude.
Finally, we demonstrate that these biases provide an explanation for the cold
retrieved temperatures reported for WASP-17b and WASP-12b. To overcome biases
associated with 1D atmospheric models, there is an urgent need to develop
multidimensional retrieval techniques.
http://arxiv.org/icons/sfx.gif
