Coupled atmospHere Interior modeL Intercomparison (CHILI) Protocol Version 1.0: A CUISINES Intercomparison Project of Magma Ocean Models

TL;DR

The CHILI Version 1.0 protocol addresses the need for systematic intercomparison of magma ocean interior–atmosphere coupling across multiple exoplanet and Solar System scenarios. By organizing participating codes into evolutionary and static categories within the CUISINES framework, CHILI defines standardized inputs, outputs, and benchmark cases for Earth, Venus, and select exoplanets, including TRAPPIST-1 b/e and a lava-world analog TRAPPIST-1 α. Initial preliminary results reveal meaningful divergences in surface temperature, melt fraction evolution, and atmospheric composition across models, underscoring the importance of an organized intercomparison to identify robust physical prescriptions and key uncertainties. The project aims to enable efficient interpretation of JWST data, constrain plausible volatile inventories and redox histories, and connect magma ocean physics to planetary habitability assessments, while fostering broad community participation and data sharing through open repositories. Overall, CHILI establishes a structured pathway to reconcile interior and atmospheric modeling across diverse planetary regimes, with results to be disseminated in follow-up intercomparison papers and leveraged by related CUISINES projects.

Abstract

Spectroscopic characterization of rocky exoplanets with the James Webb Space Telescope has brought the origin and evolution of their atmospheres into the focus of exoplanet science. Time-evolved models of the feedback between interior and atmosphere are critical to predict and interpret these observations and link them to the Solar System terrestrial planets. However, models differ in methodologies and input data, which can lead to significant differences in interpretation. In this paper, we present the experimental protocol of the Coupled atmospHere Interior modeL Intercomparison (CHILI) project. CHILI is an (exo-)planet model intercomparison project within the Climates Using Interactive Suites of Intercomparisons Nested for Exoplanet Studies (CUISINES) framework, which aims to support a diverse set of multi-model intercomparison projects in the exoplanet community. The present protocol includes the initial set of participating magma ocean models, divided into evolutionary and static models, and two types of test categories, one focused on Solar System planets (Earth & Venus) and the other on exoplanets orbiting low-mass M-dwarfs. Both test categories aim to quantify the evolution of key markers of the links between planetary atmospheres and interiors over geological timescales. The proposed tests would allow us to quantify and compare the differences between coupled atmosphere-interior models used by the exoplanet and planetary science communities. Results from the proposed tests will be published in dedicated follow-up papers. To encourage the community to join this comparison effort and as an example, we present initial test results for the early Earth and TRAPPIST-1 b, conducted with models differing in the treatment of energy transport in the planetary interior and atmosphere, surface boundary layer, geochemistry, and the in- and outgassing of volatile compounds.

Figures (4)

• Figure 1: Illustration describing the processes treated by several models participating in CHILI. The leftmost seven models are evolutionary, where dynamics and composition change with time. The rightmost four models are static, where the planets are implicitly assumed to be at steady-state. MORE is shown between evolutionary and static models, as it not fully time-dependent (see main text). The six radial domains of the models include: the core, solid mantle, liquid mantle, surface boundary layer, atmosphere, and the exosphere or outer-space. Interactions coupling these layers are denoted by arrows.
• Figure 2: Evolutionary models example test outputs for the nominal Earth case, with inputs defined in Table \ref{['tab:inputs_solar_system']}. The figure shows surface temperature in Kelvin (panel A, left), volumetric melt fraction of the whole mantle (panel B, center), and partial pressure of H2O in bars (panel C, right). All model results are truncated at the model-specific endpoint time. Only a subset of the evolutionary models that will participate in CHILI are shown here.
• Figure 3: Evolutionary models example test outputs for the TRAPPIST-1 b case, with inputs defined in Table \ref{['tab:inputs_trappist1']}. The figure shows surface temperature in Kelvin (panel A, left), volumetric melt fraction of the whole mantle (panel B, center), and partial pressure of H2O in bars (panel C, right). Dots show the predictions from two static models, MOAChi and LavAtmos at $10^5$ yrs.
• Figure 4: Static model example test with MOAChi for the TRAPPIST-1 b case at $\tau_5 = 10^5$ yr for the output of the evolutionary model PROTEUS. The figure shows surface temperature as a function of pressure (panel A) and altitude (panel B).