Forming Worlds Lab

3 July

Hello, and welcome! 👋 We're the Forming Worlds Lab at @kapteynastro.bsky.social, @rug.nl. We study how rocky planets form and evolve, and what makes a world habitable, from magma oceans 🌋 to the skies of distant exoplanets 🪐, and we build the open-source PROTEUS framework our science runs on. 🔭

New paper led by PhD student Mariana Sastre (@marianasastre.bsky.social): "Thermal Evolution of Lava Planets Across System Ages: Predictions for Hell of a Survey", now accepted in The Astrophysical Journal 🌋 https://arxiv.org/abs/2609.20155 (1/6)

View full thread

TOI-431 b & TOI-2431 b only match their densities with iron cores of 70 to 80% by radius. For systems ≥320 Myr old, escape efficiency ≥10^-2 strips the atmosphere (Fig. 3). Where one survives (Fig. 5), molten mantles outgas H2O, H2, CO2 & CO, while S2 builds up only when solid. (5/6)

Because sulphur stays dissolved until late crystallisation and heavy S2 resists escape, detecting S2 points to a solidified interior, while a small phase curve amplitude flags a thick atmosphere over a molten mantle 🔭 https://arxiv.org/abs/2609.20155 #Exoplanets #PlanetaryScience #JWST (6/6)

Nicholas Connors, Ian Crossfield, Lina D'Aoust, Oliver Herbort, Leoni Janssen, Mathilde Kervazo, Owen Lammert, @astroyami.bsky.social, @climatebook.bsky.social, Allona Vazan, Joost P. Wardenier & @sebizieba.bsky.social (The Hell of a Survey collaboration). (3/6)

JWST Program 8864 observes MIRI phase curves of 5 ultra-short-period rocky planets (~200 Myr to >5 Gyr). In coupled interior-atmosphere models (Fig. 6 in post 1), the 3 older worlds (>2 Gyr: TOI-6255 b, TOI-431 b, TOI-2431 b) cool into solid bare rocks, while the 2 young ones stay molten 🌍 (4/6)

With @timlichtenberg.bsky.social, @astrolisadang.bsky.social, Anjali Piette, Haiyang S. Wang, Mercedes López-Morales, Thomas Wilson, Charles-Edouard Boukaré, Mahesh Herath, Nicolas Cowan, Md Abdullah Al Zaman, @madysonbarber.bsky.social, @spaceycaseylynn.bsky.social, (2/6)

Posted to arXiv last month and submitted to The Astrophysical Journal, work led by PhD student Emma Postolec: 🌋 "Evolutionary pathways toward survival of a thick CO2- or SO2-rich atmosphere on the lava world TOI-561 b" Read: https://arxiv.org/abs/2609.03144 #Exoplanets #PlanetaryScience

View full thread

Every surviving run still sits on a global magma ocean under 1,000 to 10,000 bar of heavy gas (first plot in post 1). They split into two archetypes set by how much sulfur the planet started with (plot here): a CO2-dominated super-Venus or an SO2-dominated super-Io.

Late inward migration works too if the planet moves in within 1 to 100 Myr, though it isn't required: in-situ and migrated histories look the same today. Telling the CO2-rich and SO2-rich atmospheres apart is the next job for observations. 🔭

TOI-561 b orbits an 11-Gyr-old star every 11 hours, where a rocky planet should have been stripped bare ages ago. Instead its 4.3 g/cm³ density is far below bare rock (5.5 g/cm³), and JWST eclipse and phase-curve data (Teske+, Boucher+) point to a thick atmosphere over a magma ocean.

Emma ran coupled interior-atmosphere simulations over the full 11 Gyr to see what keeps an envelope alive that long. Over half the grid strips to bare rock. Survivors need a small iron core (≤40% in radius), an oxidised mantle, up to 200 Earth oceans of H, and slow escape (ε ≤ 0.001). 🌍

With @timlichtenberg.bsky.social, @johannateske.bsky.social, @nichollsh.bsky.social, @marattia.bsky.social, Anjali Piette, @astrolisadang.bsky.social, @nicolewallack.bsky.social, Mykhaylo Plotnykov, @alexmcginty.bsky.social, Samuel Boucher, @planetarybo.bsky.social & Diana Valencia.

15 Sept

🌍✨ Last week, the FormingWorlds group attended the Europlanet Science Congress (#EPSC2026) in The Hague. We had a great time with Karen Stuitje, Saphira Brandenberger, @marianasastre.bsky.social, Emeline Decocq, @marattia.bsky.social, Lorenzo Cesario, Ema Jungová & @timlichtenberg.bsky.social!

15 Sept

We showcased our planetary science work at the Kapteyn Astronomical Institute: - Hadean Earth redox & climate (Emeline Decocq) - PALEOS: Rocky planet equations of state (@marattia.bsky.social)

15 Sept

- fO2-dependent melting on exoplanets (@marianasastre.bsky.social) - Redox-driven atmospheric inflation (Lorenzo Cesario & @timlichtenberg.bsky.social) 🌋🔭✨

8 Sept

Accepted earlier this year in A&A, work led by Mara Attia (@marattia.bsky.social) tackles a core problem in exoplanet interiors: "PALEOS: Multiphase equations of state and mass-radius relations for exoplanet interiors" 🌍 Read: https://arxiv.org/abs/2605.03741 #Exoplanets

Lead images: Mara Attia and the paper abstract. Below: core size vs temperature (measured radius as contour, solutions A and B), and mass-radius curves at four temperatures. Behind it is PALEOS, 17 phases, checked against Earth to 0.3%: https://github.com/maraattia/PALEOS

8 Sept

A point on the mass-radius diagram is not an interior until you know the temperature. 🔭 With @timlichtenberg.bsky.social, Ema Jungová and @marianasastre.bsky.social at @kapteynastro.bsky.social.

Heat it and the rock expands. For low-mass silicate planets the radius grows by more than 1% above 1500 K and up to 16% at 4000 K. That is as large as the spread from changing bulk composition, and as large as the typical transit radius uncertainty. 🌋

8 Sept

So the same mass and radius can be two different planets. WASP-47 e and TOI-1807 b each admit two purely rocky solutions, identical in mass and radius: one fully solid, one with a deep magma ocean on a liquid iron core.

Measure an exoplanet's mass and radius and you get its density, and density is usually read as composition: how much iron, how much rock, how much water. That reading assumes the inside is cold and solid.

28 Aug

The JWST campaign on TOI-561 b, led by @johannateske.bsky.social, continues to deliver surprises! Samuel Boucher (supervised by @astrolisadang.bsky.social) just posted a new phase curve study on this lava world.

The phase curve shows a high Bond albedo and moderate global heat transport. When you feed that into general circulation models, the best fit is a highly reflective atmosphere loaded with silicate clouds (like SiO2 and MgSiO3) near the dayside terminator.

28 Aug

Keeping that atmosphere intact against escape means the gas must be actively replenished. The surface magma ocean is likely exchanging volatiles with the envelope to maintain it over time. Accepted in The Astronomical Journal. Congrats to Samuel on his work with Lisa! 🌋🔭

Emma Postolec, @timlichtenberg.bsky.social, and @nichollsh.bsky.social joined the team to look at the interior-atmosphere exchange. Read the paper: https://arxiv.org/abs/2608.21519

28 Aug

Last year, Teske et al. (2025) showed this super-Earth held onto a thick volatile atmosphere despite the intense radiation. Now, these new 37-hour continuous NIRSpec observations map the planet across its entire orbit.

24 Aug

Earlier this summer, Mariana Sastre (@marianasastre.bsky.social) posted her first paper as lead author 🌋 "Geophysical and atmospheric implications of fO2-dependent melting on rocky exoplanets" Read the paper: https://arxiv.org/abs/2606.20249 #Exoplanets #PlanetaryScience

Test case is the short-period super-Earth GJ 1132 b. The two atmospheres are distinct enough for JWST to tell apart. 🔭 The first image shows the melting curves themselves: reduced in blue melts hotter, oxidised in pink melts cooler.

24 Aug

Accepted in Astronomy & Astrophysics. With @timlichtenberg.bsky.social, Laurent Soucasse (@esciencecenter.bsky.social), Dan Bower, @nichollsh.bsky.social (@cambridgeastro.bsky.social) and Inga Kamp, at @kapteynastro.bsky.social.

You would expect that to rewrite the cooling history. It only does when the planet is volatile-poor. Give it enough water and the outgassed atmosphere holds the mantle above its melting point whatever the curve says.

24 Aug

What the planet breathes out is a different matter. Reduced surfaces give hydrogen and carbon monoxide, oxidised ones water and carbon dioxide. Plot: outgassed composition against mantle oxidation state, for three starting water inventories. 🌍

Models of young molten planets normally carry one melting curve, the temperature at which mantle rock starts to melt, taken from oxidised Earth-like rock. How oxidised the rock actually is moves that curve by hundreds of kelvin.

New paper from Lorenzo Cesario 🔭 His third first-author paper, and he only finished his MSc this year. "Characterizing the oxidation state of rocky exoplanets with the Large Interferometer for Exoplanets (LIFE)" Read the paper: https://arxiv.org/abs/2607.28121 #Exoplanets #PlanetaryScience

Accepted at Monthly Notices of the Royal Astronomical Society. With @timlichtenberg.bsky.social, @exoclouds.bsky.social (@sronspace.bsky.social), Lena Noack and Caroline Brachmann (@freieuniversitaet.bsky.social), Eleonora Alei, Sascha Quanz (@ethz.ch) and the LIFE Collaboration.

It's a proof of concept: one star type, one orbit, no photochemistry or clouds. Adding those is the next step. 🌍 Code and data: https://github.com/lorlo9999/LIFE_RedoxStateCharacterization

The six states can be told apart. Carbon dioxide dominates the oxidised cases, ammonia the reduced ones, and methane peaks in the middle, which is what separates the two most oxidised cases. Plot: retrieved abundance of each gas, black bars for the true values.

Reduced atmospheres are harder to retrieve than oxidised ones, because their spectra are flatter. In the spectra on the first post: top row reduced, bottom oxidised, red the input model, blue the retrieved range, grey the noise. Hydrogen and nitrogen were not useful.

The gases a rocky planet releases depend on how oxidised its mantle is. 🌋 That matters for biosignatures: you can't say a gas is unusual without knowing what a lifeless planet would produce. And you can't measure mantle chemistry directly from ten parsecs away.

So we tested whether it can be measured indirectly. Atmospheres for six mantle oxidation states, twelve orders of magnitude in oxygen fugacity, then a simulation of what LIFE would see for Earth-sized planets at 1 AU around a Sun-like star at 10 parsecs.

arxiv.orgConstraining the lives and times of exoplanets through evolutionary Bayesian retrievalsStatic retrieval frameworks are leading tools for interpreting exoplanet observations, yet time-independent modelling leaves them prone to degeneracy and unable to resolve exoplanets' histories. The c...

Harrison worked closely with our group all through his PhD, and still does. His new paper asks what you can learn about an exoplanet by fitting its whole life instead of a snapshot of today. 🔭 https://arxiv.org/abs/2607.25845 #Exoplanets #PlanetaryScience

Harrison Nicholls@nichollsh.bsky.social

New paper on arXiv today! On the efficacy of using novel Bayesian machine learning to dispatch 'expensive' forward models of planetary evolution. This connects with #PLATO ageing and can help solve degeneracies in current static-retrieval approaches which treat #exoplanets as snapshots.

100 simulations per retrieval, a few minutes each, hours of wall-clock on fewer than ten CPU cores. On a warm terrestrial that recovers the volatile inventory the planet formed with to within ~20%, and constrains its mantle redox state and metallic core fraction.

Pink is everything a static fit allows for surface pressure and temperature, black the evolving best fit, the ring the truth. Sub-Neptunes stay hard: core fraction and volatile budget still trade off. ⏳ Proof of concept on synthetic planets, built into PROTEUS. @kapteynastro.bsky.social

Harrison Nicholls@nichollsh.bsky.social

Retrievals fit a planet as it is now, because the model has to run thousands of times and so has to be cheap. That leaves ambiguity: a sub-Neptune's density suits a steam atmosphere over a small core about as well as a hydrogen envelope over a dense one. Figure: the three test planets.

So fit the whole life instead, magma ocean to today. Every guess is then a full coupled interior-atmosphere run. 🌋 Asynchronous Bayesian optimisation makes that affordable: each worker starts its next simulation the moment its last one ends, instead of idling for the batch.

Harrison Nicholls@nichollsh.bsky.social