Modeling Earth as an exoplanet in reflected and polarized light

Roccetti, Giulia (2025): Modeling Earth as an exoplanet in reflected and polarized light. Dissertation, LMU München: Fakultät für Physik

Abstract

The next generation of ground- and space-based observatories will enable the detailed characterization of rocky exoplanets within the habitable zones of their host stars. With improved sensitivity, these facilities allow us to observe Earth-like planets and study their reflected light, which provides valuable information about planetary albedo, a result of the interplay between surface and atmospheric properties. For the first time, we will be able to assess the habitability of nearby non-transiting rocky exoplanets, such as Proxima b, and explore whether they might retain liquid water on their surfaces. Observing Earth as if it were an exoplanet offers crucial insights into how we assess exoplanetary habitability. Earthshine, the sunlight reflected from Earth onto the darker portion of the visible Moon, provides an opportunity to study Earth in a way that closely mirrors how exoplanets are observed in reflected light. Unlike satellite observations, Earthshine captures the complex scattering and reflection geometries encountered when observing exoplanets. In this thesis, I model Earth as an exoplanet using advanced 3D Monte Carlo radiative transfer codes in the visible and near-infrared spectral ranges. My approach builds on codes originally developed for Earth’s remote sensing, using 3D atmospheres and 2D surface albedo maps. By combining state-of-the-art knowledge from Earth observations and reanalysis product, I generate spatially unresolved spectra and phase curves of Earth. I address two notorious challenges: the accurate representation of surface properties and the complex behavior of clouds. Earth’s surface albedo is highly variable across space and time, with also wavelength-dependent features like the Vegetation Red Edge (VRE), a peak in vegetation reflectivity around 700 nm caused by chlorophyll absorption. Satellite data provides detailed albedo maps at only a few wavelengths in the visible and near-infrared, limiting our ability to fully simulate these features. To extend this data, I employ a Principal Component Analysis (PCA) algorithm to generate hyperspectral albedo maps, greatly improving surface representation. Clouds pose an even greater challenge due to their complexity. Standard models using cloud properties from satellite observations and weather forecasts significantly overestimate the planet’s global reflectivity, an issue also seen in climate models. To address this, I develop a 3D cloud generator algorithm that creates finer-grid cloud patterns, representing their patchy nature. This approach allows for a more precise representation of Earth's clouds and their influence on reflectivity, enhancing our ability to model Earth as an exoplanet. I validate my model, particularly its treatment of clouds and surface albedo, using a decade-long dataset of Earthshine observations that encompass a variety of planetary geometries and cloud conditions, for both intensity and polarization. Polarization offers deeper insights into the planet’s physical properties and has the advantage of not requiring atmospheric correction. I successfully validate the surface and clouds models in both intensity and polarized light, creating a robust framework capable of accurately representing Earth as an exoplanet. My model is far superior to previous descriptions in the literature because it matches both the continuum features and the absorption lines of Earthshine observations in a way that has never been achieved by earlier attempts. This work also proposes an optimal strategy for detecting liquid water on exoplanets. Ocean glint, observable at high phase angles through phase curve variability or water lines, is best detected via polarization, which is highly sensitive to ocean surfaces. Polarization also reveals liquid water in clouds by identifying rainbows at smaller phase angles, offering detailed insights into cloud droplet properties. Combining polarization with traditional spectroscopy enhances the precision of exoplanet habitability assessments. This study paves the way for future instruments on observatories like the Extremely Large Telescope (ELT) and upcoming space missions such as the Habitable Worlds Observatory (HWO). Additionally, the advanced modeling developed here contributes to both exoplanet science and climate research.

• Dokumententyp: Dissertationen (Dissertation, LMU München)
• Keywords: Earth, exoplanet atmospheres, Earth's remote sensing, radiative transfer, astrobiology
• Themengebiete: 500 Naturwissenschaften und Mathematik ; 500 Naturwissenschaften und Mathematik > 530 Physik
• Fakultäten: Fakultät für Physik
• Sprache der Hochschulschrift: Englisch
• Datum der mündlichen Prüfung: 14. Juli 2025
• 1. Bericht­erstatter:in: Ercolano, Barbara