OPAL Key Science: Exoplanet Atmospheres

• OPAL Key Science Project is an integrated modeling campaign that simulates exoplanet formation by tracing stellar and disk chemistry through atmospheric evolution.
• It employs a seven-stage computational pipeline—spanning from disk evolution to N-body dynamics and chemical kinetics—to capture intricate formation pathways.
• High-performance computing enables the generation of a robust synthetic-atmosphere library, essential for precise validation of Ariel’s atmospheric retrieval techniques.

The Origins of Planets for Ariel (OPAL) Key Science Project is an end-to-end numerical campaign developed to support the ESA Ariel mission. By constructing a physically consistent, high-fidelity library of synthetic exoplanetary atmospheres, OPAL enables precise testing and validation of Ariel’s retrieval pipelines, and delivers essential groundwork for interpreting the mission’s large-scale atmospheric surveys. OPAL distinguishes itself through an integrated modeling strategy, linking host-star photospheric chemistry, protoplanetary disk evolution, planet formation and migration, and resulting atmospheric composition, with explicit tracking of both equilibrium and disequilibrium chemistry for hundreds of molecular and atomic species (Polychroni et al., 23 Jan 2026).

1. Scientific Objectives

OPAL’s primary objective is to determine how the chemical compositions of gas-giant atmospheres encode their entire accretion and migration histories. Rather than restricting analysis to post-disk-dispersal conditions or adopting oversimplified atmospheric models, OPAL reconstructs the genetic pathway from primordial stellar abundances and disk physics through to the observable atmospheric properties of giant planets.

Specific aims include establishing which atmospheric elemental ratios—such as C/O, C/N, and Si/O—or molecular tracers can diagnose planet-forming environments or differentiate between otherwise degenerate formation pathways. Key questions addressed include: what fraction of heavy elements is accreted as solids versus gas ($f_s \equiv M_\mathrm{solids} / M_\mathrm{total}$), and how do factors like disk ionization, grain size, and core growth timescale imprint on atmospheric diagnostics?

2. End-to-End Pipeline and Building Blocks

The OPAL pipeline comprises seven major computational stages, summarized in the table below.

Module	Role in Pipeline	Key Outputs and Features
GGChem	Partition host-star elements between gas, condensates, ices	Initial disk composition by species and reservoir
JADE	Evolve disk structure and chemistry (0–3 Myr)	Radial and temporal disk chemical maps, grain growth, planetesimals
GroMiT	Monte Carlo sampling of planet formation pathways	Distributions of final planet mass (M_f) and orbit (a_f)
Mercury-Arχes	N-body planet+planetesimal formation/migration	Dynamical/accretion histories tagged by source region
Hephaestus	Merge accretion histories and disk chemistry	Complete elemental inventory of formed planet
FastChem	Solve for equilibrium chemistry in atmosphere	Atmospheric composition, molecular abundances (∼500 species)
Vulcan	Integrate atmospheric disequilibrium kinetics	Disequilibrium molecular profiles for ∼100 species

The process initiates with GGChem [Woitke et al. 2018], which assigns elements to gaseous and solid phases according to the host star’s abundances, defining the base composition of the nascent disk. JADE [Pacetti et al. 2025] advances the evolving disk using coupled viscous diffusion and two-phase astrochemistry for 0–3 Myr, with 8,000 reactions among 600 species, tracking dust growth and planetesimal formation.

GroMiT [Polychroni et al. 2023] samples 10⁵ pebble-accretion formation scenarios, randomizing key initial conditions (e.g., embryo insertion epoch, radial location, disk viscosity α) and generating empirical distributions for final planetary locations and masses. Mercury-Arχes (Turrini et al.) applies N-body dynamics, migration, and mass growth (via solid- and runaway gas-accretion regimes) to form one or more planets, explicitly tagging accreted planetesimals by formation radius to preserve chemical provenance.

Merging chemical evolution and accretion history, Hephaestus [Pacetti & Turrini 2022] synthesizes a complete elemental inventory for each planet. FastChem [Stock et al. 2018, 2022] computes the equilibrium molecular composition for specified 1D pressure–temperature profiles ($P = 10^{-6}$–$10^3$ bar), while Vulcan [Tsai et al. 2017, Simonetti et al. in review] solves the coupled kinetics and vertical transport equations for disequilibrium molecular abundances, including refractory–volatile pairs and vertical mixing ($K_{zz} = 10^{6}$–$10^{10}$ cm² s$^{-1}$).

3. Tracing the Genetic Link: Methodology

A central methodology within OPAL is the explicit tracing of elements from stellar origin through planet assembly—a “genetic” approach. For every modeled planet, the integration follows:

• Bulk planetary metallicity: $$Z \equiv M_Z/M_\mathrm{total} = [\sum_{X \ne \mathrm{H},\mathrm{He}} M_X]/M_p$$
• Key ratios (e.g., $(\mathrm{C}/\mathrm{O})_p$, $(\mathrm{C}/\mathrm{N})_p$) compared to the star ($(C/O)_\star$) to diagnose enrichment.
• The envelope elemental content is a time-integral over the accretion streams, with solids and gas traced separately through

$$\Delta X_\mathrm{solid} = \int \dot{M}_\mathrm{solid}(t)\, X_\mathrm{solid}(r(t),t)\,dt$$

$$\Delta X_\mathrm{gas} = \int \dot{M}_\mathrm{gas}(t)\, X_\mathrm{gas}(r(t),t)\,dt$$

• Disk chemistry (e.g., α-viscosity law $\nu = \alpha c_s H$) controls volatile and refractory partitioning, directly affecting enrichment pathways and observable statistics.

This element tagging through all stages allows the OPAL library to maintain “provenance” for each spectrum, supporting robust testing of atmospheric retrieval diagnostics.

4. Synthetic Atmosphere Library Characteristics

OPAL’s synthetic atmosphere library encompasses:

• 25 tracked elements and approximately 400 neutral and 100 charged gas-phase species per modeled atmosphere.
• Both chemical equilibrium (FastChem) and disequilibrium (Vulcan) end-members.
• $P$–$T$ profiles appropriate to hot Jupiters ($T_\mathrm{eq} \sim 1000$–$2000$ K).
• Opacities sourced from ExoMol, HITRAN, HITEMP, and, where necessary, refractories (e.g., MgSiO$_3$, Fe).
• Forward radiative transfer implemented by

$F_\lambda = \int_{0}^{\infty} B_\lambda[T(z)]\, e^{-\tau_\lambda(z)}\, d\tau_\lambda(z}$

where $\tau_\lambda$ is line-by-line optical depth.

• Parameter space spanning a wide diversity of disk initial conditions, growth timescales, and migration tracks.

For each synthetic planet, dual models (equilibrium/disequilibrium) provide bracketing cases, supporting sensitivity studies and validation of retrieval codes’ assumptions about atmospheric chemical processes.

5. Parameter Space Exploration and Early Results

Preliminary sparse sampling, exemplified by applications to WASP-69 b analogs, demonstrates a compositional diversity resulting from the wide degeneracy in plausible planet formation pathways. Bulk atmospheric C/O ratios range from heavily sub-solar ($\sim$0.2) to super-solar ($>$1.0), with the fraction of carbon accreted as solids dictating the locus in C/O–C/N space. Gas-only accretion events generate atmospheres with low C/N and high C/O, while those with significant solid accretion yield high C/N and diverse oxygen content.

The oxygen deficit metric,

$$D_O \equiv 100\,\times \left[1 - \frac{O_\mathrm{refractory}}{O_\mathrm{total}}\right]$$

displays a wide dynamic range (from $D_O < 0$ for carbon-dominated to $D_O > 20\%$ for refractory-rich outcomes), and approximately 30% of cases yield atmospheres with super-solar C/O. There is no single invariant molecular tracer able to uniquely recover formation history; rather, discriminating planet formation processes requires considering a multi-dimensional ratio vector.

6. High-Performance Computational Framework

OPAL’s computational infrastructure leverages the Leonardo pre-exascale Euro-HPC system at CINECA. The hardware comprises DCGP nodes each equipped with dual 56-core Intel Xeon Platinum 8480+ CPUs and 512 GB DDR5. Efficiency optimizations are achieved by domain-decomposing (JADE astrochemistry) and by running parallel simulations with moderately sized core allocations (Mercury-Arχes on 14 cores for 7–10 days per job). FastChem and Vulcan atmospheric runs are modest in computational cost and can be scheduled to exploit node spare capacity.

Workflow optimizations target I/O streamlining, integrated management of Python/Fortran boundary interactions, and future acceleration via GPU porting—Mercury-Arχes being the immediate focus, with JADE identified as a subsequent target. Scaling tests confirm that linear speedup is maintained in core-dominated stages.

7. Implications for Exoplanet Taxonomy and Mission Science

The synthetic-atmosphere library produced by OPAL will be a primary validation set for the Ariel mission’s retrieval suite, allowing controlled end-to-end tests under physically consistent initial conditions. By directly comparing retrieval outputs for elemental ratios and molecular abundances to known OPAL inputs, systematic biases and dominant uncertainties can be quantified, and optimal observational strategies designed.

OPAL’s results indicate that simultaneous measurement of C/O and C/N (and, where feasible, Si/O or Fe/O) with precision of $\lesssim 0.1$ is required to disentangle formation inside versus outside major snowlines, evaluate solid-to-gas enrichment ($f_s$), and test core accretion and migration models in exoplanet populations. The project is poised to establish the first compositional taxonomy of exoplanets that is grounded in the physics of planet formation, directly linking present-day atmospheric chemistry to the primordial environments and dynamics in which those ecosystems arose (Polychroni et al., 23 Jan 2026).