Probing thermal gradients of habitable-zone rocky planets as an anti-indicator of a global surface ocean using mid-infrared direct imaging

Abstract

Future direct-imaging missions such as the Large Interferometer for Exoplanets (LIFE) aim to observe thermal emission from potentially habitable planets to characterize their surface environments and search for signs of life. Previous studies of directly imaged Earth-like planets have mainly examined the signatures of atmospheric composition, often using 1D models, while the effect of horizontal temperature gradients has received limited attention. Because a pronounced horizontal temperature gradient may signal the absence of a global ocean, we investigate its detectability through thermal-emission direct imaging. Adopting Teegarden's Star b (zero-albedo equilibrium temperature ~280 K) as a benchmark, we compute 3D atmospheric structures with and without a global ocean using the ROCKE-3D general circulation model and simulate geometry-dependent thermal emission spectra. We show that the temperature gradients that disfavor a global-ocean scenario manifest in both orbital phase variation and spectral shape of the snapshot spectra. The phase variation is more readily detectable: one-day integrations with LIFE at two orbital phases would reveal flux variations in no-ocean cases with 1-10 bar atmospheres, depending on background atmospheric composition. Shapshot spectra provide complementary diagnostics, including the running brightness temperature of the continuum and detailed absorption band shapes, but detecting these features requires longer integration times. These 3D effects, if neglected, can bias interpretations based on 1D models, and could also lead to misidentification of the H$_2$O band beyond 10 $\mu$m. We also assess the detectability of these 3D effects on other nearby planets. Our results highlight the need to incorporate 3D atmospheric structures when characterizing rocky exoplanets, both to constrain surface conditions and to avoid misinterpretation of spectral data.