Abstract

JWST has provided a new wealth of information on exoplanet atmospheres, including one of the most critical components, that being the composition of cloud particles and their 3D global distribution.

In this review talk, I will first outline the basic theories and multiple approaches applied cloud formation in exosolar atmospheres, as well as their general effects on the observed properties of exoplanet atmospheres. Moving through the historical context in brown dwarf atmosphere science to contemporary modelling efforts will serve to provide the audience with a broad literature review and context to the field. A simple interactive python based demonstration will be included to aid understanding of the basic methodology.

Moving into 3D, I will discuss how the field approaches this tricky topic using general circulation models (GCMs) coupled to various flavours of cloud models, from simple diagnostic efforts to complex microphysical models. I will showcase recent progress in the field using these techniques and how they have been used to uncover the complex interactions and feedbacks that clouds can induce in these atmospheres. Another interactive session will elucidate the time-dependent nature of cloud formation in 3D and the modelling of microphysics.

Lastly, I will suggest what types of observational data and programs would best constrain the 3D properties of clouds in exoplanet atmospheres, as well as discuss gaps in theory and data that can be can tackled in future endeavours in the field.

Abstract

Abstract

This study aims to characterize the internal structure of the atmospheres in the benchmark brown dwarfs binary system, WISE J104915.57--531906.1AB. This binary system is the closest and brightest to us and is composed of a primary L7.5 (WISE 1049A) and a secondary T0.5 (WISE 1049B). As both objects are in the L/T transition, we are interested in studying their variability because although both components vary significantly, B presents higher variability than A. We use data from 7h of time-resolved spectroscopy (~57100 spectra for each object) to study their variability, taken by NIRSpec/PRISM BOTS on board JWST (GO 2965).

We select 10 molecular bands in each object's spectrum and calculate their respective light curves. We find the best fit of a sum of sinusoidal functions at the light curves using statistical methods like the f-test. The best fit gives us information about the number of bands present in the atmosphere for that specific depth. Studying the residuals, we found regions of the light curves that could not be accounted for with our sum of sinusoidal functions. We interpret this behavior as spots in the object's atmosphere, which may repeat in the light curve with each successive rotation of the brown dwarf.

From analyzing the contribution functions generated by radiative transfer models, we locate the molecular bands at different pressure depths to generate the first atmospheric map using the JWST data. Thanks to these data, we infer a preliminary vertical structure and have a 3D view of a brown dwarf for the first time.

Spectroscopic phase curves are treasure troves of information, enabling us to make 3D scans of planetary atmospheres and observe the longitudinally dependent distributions of chemical species. This is important, because the atmospheric composition of tidally locked gas giants is determined by locally varying parameters such as the temperature, the wind dynamics, and the incident radiation coming from the host star. As such, the planet shows distinct atmospheric signatures dependent on the phase during which it is observed. Now, with the unprecedented sensitivity of the James Webb Space Telescope, we can finally gain insight in the physical and chemical processes that occur in a planetary atmosphere.

During this presentation, we will present a suite of one-, two-, and three-dimensional photochemical models, exploring in detail the chemistry of the hot Jupiter WASP-43 b. We will match our photochemical results to constraints offered by the mid-infrared phase curve that was obtained using the James Webb Space Telescope (Bell et al. 2024) and derive insights into the metallicity, vertical mixing rate, and wind dynamics of WASP-43 b. Intriguingly, we determine both dynamical and chemical mechanisms that cause a depletion of methane. This presentation will highlight the information content provided to us by phase curves and enable future studies of the chemical distributions of tidally locked exoplanets.

Clouds play a crucial role in shaping the atmospheres of exoplanets, influencing their albedo, heat distribution, and spectrum. While most past studies focused on hot and ultra-hot planets, JWST now allows an in-depth characterisation of warm objects, where the interactions between cloud, circulation, and radiative transfer are yet to be studied in detail.

In this study, we integrate physically motivated, radiatively active, tracer-based clouds in General Circulation Models (GCMs) to analyze the atmosphere of WASP-80b, a warm Jupiter orbiting an M-dwarf star. This planet, with an equilibrium temperature of ~800 K, is analogous to warm sub-Neptunes, making it a crucial target for understanding sub-Neptune atmospheres. We take advantage of the high-quality dataset from JWST obtained through the MANATEE GTO collaboration, allowing us to jointly interpret both high-quality panchromatic emission and absorption spectra ranging from 2.5 to 12 microns with the outputs from a GCM. We provide an in-depth characterisation of the planet's atmospheric dynamics and molecular distribution, including unique photochemical molecules. Realistic insights into the distribution of clouds composed of various species, such as silicates, sulfides, and chlorides, with varying particle sizes were obtained.

While both emission and transmission spectra are very well fitted by cloudless GCMs, the data also appears compatible with small KCl cloud particles, but Na2S condensates can be ruled out due to the strength of their radiative feedback. This showcases the unique insights that can be obtained from global modelling of exoplanet atmospheres.

Our results point towards a homogeneous atmosphere with minimal temperature contrast between the day and night sides, suggesting efficient heat redistribution. Additionally, the presence of photochemical products such as CS2 and the relatively low abundance of CH4 point to active atmospheric chemistry and the possibility of a high internal heat flux. This work not only provides a comprehensive framework for interpreting JWST observations but also enhances the capabilities of GCMs in characterizing the global atmospheres of exoplanets in the JWST era.

Sub-Neptunes with substantial atmospheres are expected to possess magma oceans in contact with the overlying gas. Chemical interactions between the atmosphere and interior could potentially play a critical role in shaping the planet's atmospheric composition and interior structure. Early JWST observations have found a number of highly metal-enriched sub-Neptune atmospheres, which may result either from accretion of heavy elements at formation or from magma-atmosphere interactions. Most previous work examining these interactions has been limited to studying conditions at the atmosphere-mantle boundary, without considering implications for the upper atmosphere, which is probed by spectroscopic observations.

In this talk, we present a modeling architecture to determine observable signatures of magma-atmosphere interactions. We combine an equilibrium chemistry code which models reactions between the core, mantle and atmosphere with a radiative-convective model that determines the composition and structure of the observable upper atmosphere. We examine how different conditions at the atmosphere-mantle boundary as well as different core and mantle compositions impact the resulting upper atmospheric composition.

We compare our findings to recent JWST observations of sub-Neptune atmospheres. In particular, we find that the upper atmosphere of TOI-270 d as measured from its JWST transmission spectrum can be explained as the outcome of interactions between the atmosphere, core and mantle, assuming that the planet originally accreted a solar composition envelope. We discuss the implications of this finding, as well as the further modeling improvements that will be required to fully understand the impact of these processes on atmospheres of the broader sub-Neptune population.

Assessing the prevalence of atmospheres on rocky planets around M-dwarf stars is a top priority of exoplanet science. High-energy activity from M-dwarfs can destroy the atmospheres of these planets, which could in turn explain the lack of atmospheric detections to date. Volcanic outgassing has been proposed as a mechanism to replenish the atmospheres of tidally heated rocky planets. Similarly, L98-59b, a sub-Earth transiting a nearby M dwarf, was recently identified as the most promising exoplanet to detect tidally driven volcanism. We present the transmission spectrum of L98-59b from four transits observed with JWST/NIRSpec. We find a 3.6\( \sigma \) preference for an SO2 atmosphere over the no-atmosphere model based on the Bayesian evidence. The best-fit atmospheric model is in excellent agreement with literature predictions that modeled tidal dissipation on this planet. Such an atmosphere would likely be in a steady state where volcanism balances escape. If so, L98-59b must experience at least eight times as much volcanism and tidal heating per unit mass as Io. If volcanism is driven by runaway melting of the mantle, we predict the existence of a subsurface magma ocean in L98-59b extending up to 60-90% of the planet radius. An SO2-rich volcanic atmosphere on L98-59b would be indicative of an oxidized mantle with an oxygen fugacity of fO2>IW+2.7, and it would imply that L98-59b must have retained some of its volatile endowment despite its proximity to its star. Our findings suggest that volcanism may revive secondary atmospheres on tidally heated rocky planets around M-dwarfs.

One of the key questions in exoplanetary science is determining whether close-in rocky planets can retain atmospheres. A particularly exciting case is that of ultra-short-period, highly irradiated rocky planets, which are predicted to give rise to exotic atmospheres unlike those observed in our Solar System. These planets provide a unique opportunity to explore the predicted compositional diversity resulting from outgassing and surface evaporation. Exposed to intense irradiation that likely strips away their primordial atmospheres, these planets may host secondary atmospheres enriched with species such as Na, O, and SiO.

Here, we present the first near-infrared transmission and emission spectra of the lava world K2-141b, a 1.58 \( R_\oplus \), 2.31 \( M_\oplus \), T~2000K hot rocky exoplanet orbiting a K-dwarf star with an ultra-short orbital period of only 7 hours. These precise measurements, obtained using the NIRSpec instrument onboard JWST (PID 2159; PI: Espinoza), allow us to place strong constraints on the atmospheric make-up of this lava world and provide an unprecedented view of these worlds between 3-5 µm. These findings contribute to understanding the evolutionary pathways of ultra-short-period rocky planets (P < 1 day, RP < 2 \( R_\oplus \)), thought to be the stripped cores of mini-Neptunes, and offer valuable insights into planetary formation and survival in highly irradiated environments.

The ice giants in our solar system (Uranus and Neptune) and their metal-rich atmospheres hold important clues about how our solar system formed and evolved. Exo-Neptunes likewise hold critical information about planet formation beyond our solar system, providing an important bridge between the Jupiter-sized and abundant sub-Neptune-sized exoplanet populations. Unlocking the hidden secrets of exo-Neptunes requires careful consideration of their complex carbon chemistry, which is shaped by both atmospheric metallicity and disequilibrium processes.

In this talk, I will present the transmission spectrum of the stand-out ~800K exo-Neptune HAT-P-11b as observed with JWST, with the aim of refining its unexpectedly sub-solar metallicity as observed by Hubble when compared to planets of similar mass, and reconciling its low IR transit depths as observed by Spitzer. We also highlight the necessary synergy between Hubble and JWST observations when interpreting the spectra of a planet orbiting an active host star. Our NIRSpec/G395H transmission unlocks a comprehensive view of the carbon inventory of the atmosphere for the first time, allowing us to test the validity of prior understanding, examining and perhaps disentangling the chemical and dynamical processes occurring within exoplanet atmospheres near the CO-CH4 transition.

Clouds have long been thought to shape the climates of exoplanets as much, if not more so, than clouds in the atmospheres of solar system objects. Clouds interact with atmospheric radiative transfer, chemistry, and dynamics such that they are both of first-order importance in understanding climate and can be used as tracers of fundamental planetary properties. With the advent of JWST, it is now apparent that non-gray clouds are required to interpret the atmospheres of nearly all planets, even those previously thought to be free of observable clouds (e.g., WASP-39b).

Despite a myriad of predictions that clouds are highly sensitive to atmospheric dynamics such that cloud-cover is likely to vary spatially, directly observing inhomogeneous aerosol coverage has been unfeasible for the vast majority of exoplanet phase-space until recently. With precise JWST data, it is now possible to separate limb-resolved transmission spectra that probe exoplanet atmospheres at spatially distinct locations for planets with diverse periods, temperatures, radii, and more.

Here, I present the theoretical underpinnings and initial results from the JWST Mornings & Evenings Program (PID 3969) which targets limb-asymmetry spectra of 6 hot Jupiters with equilibrium temperatures spanning 1000-2000 K. I will present the primary motivation for this program: new 2D models of microphysical clouds on hot Jupiters that predict distinct observable trends in inhomogeneous cloud coverage. I will discuss how the 2D interaction between cloud microphysics and dynamical transport substantially alters predictions about cloud-coverage and observed atmospheric properties from the 1D case and how these recent observations are able to place novel constraints on cloud formation and evolution processes. I will describe how the initial results from this population-level view of limb asymmetries on hot Jupiters both confirms and challenges state-of-the-art theoretical models — giving rise to new testable questions about the nature of these extreme worlds.

Observations and modeling of temperate exoplanets reveal potentially hazy atmospheres, coming as no surprise given the prevalence of photochemical hazes in our Solar System on bodies such as Titan. These hazes not only affect spectral features but also influence chemical and physical processes in both atmospheres and on planetary surfaces. Titan, the best-studied hazy world, forms haze through photolysis of its N2-CH4 atmosphere, producing complex organic aerosols that deposit on its the surface. These particles may interact with liquid surface water, made available through cryovolcanism or impacts on the water ice shell, leading to the formation of prebiotic molecules like amino acids and nucleotides. Laboratory studies suggest that a variety of exoplanet atmospheres could similarly produce organic haze. On water-rich planets, haze particles likely interact with liquid water after deposition, potentially driving prebiotic chemistry. Here, we perform a series of hydrolysis experiments on laboratory-simulated hazes representative of water-rich exoplanet and Titan-like atmospheres. Exoplanet haze analogs are formed through photolysis of gases at 1000x solar metallicity, while Titan haze analogs are formed through N2-CH4 photolysis. We determine the optical constants and chemical compositions of these hydrolyzed haze analogs, providing critical data for interpreting exoplanet observations and uncovering their potential biological relevance.

We present a population-level view of volatile evolution during the sub-Neptune to rocky planet transition, revealing in detail the dynamic nature of small planet atmospheric composition. We couple the atmospheric escape model IsoFATE with the magma ocean-atmosphere equilibrium chemistry model Atmodeller to simulate interior-atmosphere evolution over time for sub-Neptunes around G, K and M stars. Our simulations fully capture the transition from sub-Neptune to rocky composition atmospheres and reveal an oxidation gradient straddling the radius valley. We discover a key mechanism shaping the oxidation landscape is dissolution of water into the molten mantle, which shields oxygen from early escape, buffers the escape rate, and leads to highly oxidized secondary atmospheres following mantle outgassing. Our simulations robustly produce a prominent population of He-rich worlds along the upper edge of the radius valley as well as a broad population of O2-dominated atmospheres on close-in planets around low mass stars. These planets present a high-priority opportunity for the first-ever atmospheric O2 detection and have implications for biosignature definition. We motivate future atmospheric characterization surveys by providing a target list of planet candidates predicted to have O2-, He-, and deuterium-rich atmospheres. Finally, we present high-resolution transmission spectra that test our model predictions for a new class of planets: helium worlds.

We present the first results of the 50-hour JWST/NIRSpec phase-curve observation of the canonical ultra-hot Jupiter WASP-76b (data taken in January 2025). WASP-76b is one of the best-characterized exoplanets to date, having been observed with at least 10 different ground-based instruments. To date, 22 different atoms and molecules have been detected in its atmosphere, while 13 optical species (refractories and alkalis) have measured abundances. While such abundance measurements provide key insights into the formation/migration history of ultra-hot Jupiters, it is well-known that the extreme day-night contrasts of these planets can bias retrievals by orders of magnitude. This makes the interpretation of transmission and emission spectra a lot more challenging.

In this contribution we will present the white-light phase curve of WASP-76b, as well as initial reductions/analyses of the transmission and dayside emission spectra.

The aim of our JWST/NIRSpec program is to characterize the 3D thermochemical structure of WASP-76b in unprecedented detail and put constraints on atmospheric dynamics and heat redistribution. Additionally, we can use this 3D information to remove potential biases from retrievals based on previous ground-based observations. By combining these "updated" refractory measurements with constraints on the volatile budget from JWST, we can reliably trace back the formation/migration history of WASP-76b.

Another particularly elusive aspect of WASP-76b from previous ground-based observations is its asymmetric transmission signature, which seems to hint at stark thermochemical differences between the morning and evening terminator. Our phase-dependent transmission spectra from JWST will shed light on the physical mechanisms governing this asymmetry.

With JWST online, we are entering the era of cold substellar atmosphere characterization. The population of nearby isolated late T and Y dwarfs provide an invaluable resource to anchor cold atmospheric models ahead of the more difficult-to-observe true exoJupiters. Their high-SNR high-resolution JWST spectra encode exquisitely detailed information about atmospheric properties and processes. One such process is disequilibrium chemistry. As temperatures lower, atmospheric mixing can outpace chemical reactions and drive abundances away from thermochemical equilibrium. This phenomenon can reveal mixing strengths in radiative and convective structures below the visible photosphere, and plays an important role in interpreting observations. We conducted a comprehensive forward-model analysis, leveraging three state-of-the art model grids and JWST spectra from ~30 late T and Y type brown dwarfs in order to map atmospheric mixing strengths across the population. Observed nonequilibrium abundances support 1D-RCE models' prediction that objects cooler than 400-500 K have much shallower radiative-convective boundaries compared to 500-800 K objects. Results also imply that mixing in radiative layers is several orders of magnitude weaker than mixing in convective layers, and suggest that, even in convective layers, mixing is weaker than standard mixing length predictions.

To investigate this further, we have begun adapting the non-hydrostatic 3D modeling code CANOE to simulate the combined effects of convection, rotation, and condensation on such atmospheres, and determine whether the observed trends in disequilibrium chemistry can be replicated. 3D modeling will also enable predictions for viewing angle effects and help inform urgently needed updates to 1D-RCE models. Results of our analysis showed systematic discrepancies in the spectral residuals. The structure of data-model discrepancies hint that, at depths of 5-15 bars, either there is a missing cloud suppressing some NIR flux or temperature-pressure profiles deviate from a dry adiabat. Preliminary results from this non-hydrostatic 3D modeling will be available by July.

Reflected light is a powerful but underexploited window onto the composition, clouds and climates of other worlds. In the coming decades, it will be used to characterize some of the first potentially Earth-like worlds, with the Extremely Large Telescope, in search of biosignature gases. Further into the future, the Habitable Worlds Observatory will use reflected light phase curves to hunt for exo-rainbows and ocean glint, telltale signs of liquid water--a key ingredient for life as we know it. To achieve these future science cases requires pioneering this technique with existing instrumentation. Here I demonstrate the use of reflected light at high spectral resolution to unravel the mysteries of unusual hot Neptune LTT-9779 b.

LTT-9779 b is a rare denizen of the Neptune desert and is one of only a handful of exoplanets known to be highly reflective. JWST observations revealed a relatively flat spectrum consistent with a variety of cloudy and cloud-free atmospheric scenarios. We aim to disentangle these using high-resolution spectra, in order to learn more about this enigmatic world and to illuminate its survival in the desert.

Obtaining a high-resolution spectrum of this world required combining the Very Large Telescope's four individual telescopes into the world's largest optical telescope using the ESPRESSO spectrograph's 4UT mode. Here, I will show this data and the constraints it places on the atmosphere of LTT-9779 b, which improves upon JWST observations by disfavoring a family of proposed models. I will discuss what this means for this planet's survival in the Neptune desert and the future of reflected light studies of exoplanets with current instrumentation and with upcoming observatories.

Abstract

We analysed recent 15\( \mu \)m eclipse photometry of LHS-1140c from MIRI/IMAGING as part of the Hot Rocks Survey to determine whether it hosts an atmosphere. By comparing the level of thermal emission to various atmospheric models, we were able to rule out a wide range of atmospheres including 10 mbar pure CO2 atmospheres and 1 bar H2O atmospheres to greater than 3 sigma. We instead found our eclipse depth was most consistent with a low albedo bare rock finding a brightness temperature of 561\( \pm \)44K, close to the theoretical maximum. This should help inform atmospheric escape models which have previously suggested an atmosphere on LHS-1140c is possible (Kite & Barnett 2020; Chatterjee & Pierrehumbert 2024). This has strong implications for the Cosmic shoreline model of atmospheric escape as both LHS-1140c and its less irradiated sibling LHS-1140b are some of the most favourable super-Earths to have retained atmospheres around an M-dwarf host, making this a potential benchmark system for atmospheric escape models.

We introduce a new technique to analyse these data by joint-fitting individual pixel light curves. Our method has the advantage of being able to weight away from particular pixels displaying increased systematics/correlated noise. This helped reveal a persistence effect from a cosmic ray which could be weighted away from using this technique. We also identified a pattern in the initial ~30-60 minutes of detector settling which appears strongly connected to the previous filter used by MIRI. This helps explain the significant differences in initial detector settling seen across datasets in the Hot Rocks Survey and is an important effect to understand given the 500 hour Rocky Worlds DDT survey extending this technique to more targets.

Abstract

Abstract

Abstract

Abstract

Abstract

Atmospheric characterisation of temperate sub-Neptunes is the new frontier of exoplanetary science with recent JWST observations of K2-18 b. Accurate modelling of atmospheric processes is essential to interpreting high-precision spectroscopic data given the wide range of possible conditions in the sub-Neptune regime, including on potentially habitable planets. 3D General Circulation Models (GCMs) can provide the sophistication to model fully self-consistent atmospheres, but care needs to be taken with their modelling of complex physical effects. Notably, convection is an important process which can operate in several very different modes across sub-Neptune conditions. We develop a new mass-flux scheme which can capture these variations and simulate convection over a wide range of parameter space for use in 3D general circulation models (GCMs). We first perform an initial validation where we show that the convection model performs near-identically to Earth-tuned models in an Earth-like TRAPPIST-1 e convection case. We then perform simulations of a mini-Neptune atmosphere. Assuming the bulk properties of K2-18 b with a deep H2 - rich atmosphere, we demonstrate the capability of the scheme to reproduce non-condensing convection. We find convection occurring at pressures above 0.3 bar and the dynamical structure shows high-latitude prograde jets. We discuss the dynamical structure and the effect of convection on the wider climate state. Next, we assume that K2-18 b has a shallow H2-rich atmosphere with a liquid water ocean underneath. We discuss the general climate, dynamical climate and convective behaviour present for a range of albedos and surface pressures, and impose constraints on the possibility of a surface liquid water ocean.

Tracking the formation history of planets requires a holistic picture of their building blocks: ices, rocks, and gases. However, achieving this objective is challenging for many exoplanets, as well as for planets in our own solar system. This difficulty arises because, at typical exoplanet temperatures, most rock-forming species condense into deep interiors which makes them nearly impossible to detect.

In this regard, Ultrahot Jupiters offer a unique testbed for capturing information from both rock-forming "refractory" elements and ice-forming "volatile" elements. Their highly irradiated dayside hemispheres are in a temperature regime where both refectory species and volatile species are in gaseous form, and therefore can be sensed via emission spectroscopy. The ratio of volatile elements, for e.g., the standard C/O ratio, obtained via compositional assessment does not provide information on the solids accreted during planet formation. In contrast, refractory elements mostly remain as solids in protoplanetary disk and therefore a refractory-to-volatile ratio (tracer of rock-to-ice content) provides a complete picture on early stages of planet formation.

I will present a detailed overview of our campaign to uncover the rock-to-ice content of several benchmark ultrahot Jupiters - MASCARA-1b, WASP-121b, HAT-P-70b, and WASP-189b. Our IGRINS observations offer a unique combination of high-spectral resolution (R ~ 45000), wide wavelength coverage (1.45 - 2.45 microns), and large telescope diameter (8m). We detect a plethora of refractory (Fe, Mg, Ti, Si, Ca) and volatile (H2O, CO, OH) species using the high-resolution cross correlation spectroscopy technique, showcasing the compositional diversity in our targets. Beyond detections, I will also summarize our latest findings on the atmospheric dynamics, temperature structure, and elemental composition of the atmosphere of MASCARA-1 b. Our comprehensive analysis provides the first estimate of a super-solar refractory-to-volatile ratio of MASCARA-1 b, providing key insights into its formation and history.

Temperate sub-Neptune exoplanets could contain large inventories of water in various phases, such as water--worlds with water--rich atmospheres or even oceans. Both space-based and ground-based observations have shown that many exoplanets likely also contain photochemically-generated hazes. Haze particles are a key source of organic matter and may impact the evolution or origin of life, and their optical properties are imperative in aiding interpretations of observations using theoretical atmospheric modeling. Modelers have thus far had to assume haze optical constants which may not match sub-Neptune atmospheric compositions. Often orbiting close to M-dwarf stars, these planets receive large amounts of radiation, especially during flaring events, which may strip away their atmospheres and have the potential to accelerate atmospheric escape. Critically, it remains unknown how stellar flaring affects the formation, evolution, and optical properties of exoplanet hazes. In this work, we present laboratory investigations of stellar flare energies on laboratory-produced exoplanet hazes under conditions analogous to water-world atmospheres. We subjected two water-world haze samples to varying UV irradiation environments to assess how the hazes evolve over time. We measured the transmittance and reflectance spectra and calculated the optical properties of the hazes before and after UV irradiation across a broad wavelength range (from Far Ultraviolet to mid--IR, 0.2 - 9 μm), which overlaps with HST, JWST, and the upcoming Habitable Worlds Observatory instruments. We find that simulated flares alter the transmittance and reflectance spectra of the hazes, and also reveal that the haze optical properties change during irradiation, which may change our interpretations of exoplanet atmospheric characterization. Overall, these laboratory-made hazes show signs of degradation over the simulated flaring period. Our results provide insight into the effects that stellar flaring events have on potential exoplanet haze composition, optical properties, and the ability for water-world like exoplanets to retain their atmospheres.

As the field of exoplanet science transitions from detection to detailed characterization, instruments like JWST are revolutionizing our understanding of exoplanetary atmospheres and interiors. However, compared to atmospheres, the interiors of exoplanets are poorly understood, despite their role in interpreting atmospheric data accurately. In this talk, I will show how recent advances, including atmospheric metallicity measurements and Love number determinations for hot Jupiters, open new pathways for exploring planetary interiors. We present a retrieval framework that uses both atmospheric properties and Love numbers to constrain the internal structures of exoplanets. The Love number, which quantifies a planet's tidal deformation, provides direct insight into its internal mass distribution. While only a few Love number determinations currently exist, JWST is expected to improve the precision of the measurements and increase the sample size, thereby facilitating a better determination of interior structures. Our framework is applied using two different models: a simple two-layer model and a more complex dilute core model. We show the extent to which the Love number measurements improve our ability to infer interior properties for the two different models and identify the Love number precision required for reliable interior constraints. Finally, we apply our model to hot Jupiters with existing Love number measurements, providing new insights into their internal structure.

During the early stages of planetary formation, planets suffer severe changes in their physical and orbital properties due to internal and external forces, which also affect their primordial atmospheres composed mainly of hydrogen and helium. The study of planets at early stages and their comparison with the already mature planet population is crucial for a better comprehension of different processes, such as: planet formation, evaporation of primary atmospheres of rocky planets, gas accretion or inflation.

As part of a large project to study the evolution, outflow, and evaporation processes of exoplanet atmospheres across different planet types and host stars, we conducted high-resolution transmission spectroscopy of more than 30 exoplanets to explore their atmospheres through the two principal ground-accessible evaporation tracers: Halpha in the visible, and He triplet in the infrared. We focused on young planets, in particular the sub-Neptune and mini-Neptune population with well constrained ages of less than 1Gyr. We complemented our survey with (nearly) all the available literature results, building the largest dataset of evaporation observations. We compared the observational results between young and old planet populations, using a sample that covers a wide range of the Mass-Radius-Period parameter space. This broad range of studied planets allows us to obtain a general understanding of atmospheres across different ages, radii, and temperatures, while minimizing observational biases in the sample. In addition, this large sample helps to identify what the next steps should be to understand the processes and mechanisms of evaporation.

This talk is based on published and new results from the MOPYS (Measuring Outflows in Planets around Young Stars) survey, complemented by the observations from the large HET/HPF helium survey and preliminary results from helium observations with the new high-resolution spectrograph NIRPS.

Ultrahot Jupiters present extreme physics and dynamics not found in the solar system. Possible effects include magnetic drag disrupting the atmospheric circulation, clouds confined to the nightside, and many molecules dissociating on the dayside. Spectroscopic phase curves let us measure these spatially inhomogeneous features and comparison with 3D models help us interpret these measurements and understand the physical phenomena behind them. However, previous phase curves with Hubble were limited to looking at a narrow wavelength range, covering one water band, which limited our ability to break the degeneracy between possible physical interpretations of observed effects. JWST presents the ability to measure spectroscopic phase curves over a broad wavelength range, allowing us to observe a broader range of spectral features. We will present how phase curves of the ultrahot Jupiter WASP-121b from JWST's NIRISS instrument compare to a range of cutting-edge 3D models from different groups. While no single model includes all of the possible physics at work, by comparing varying models we can interpret what physical assumptions and effects best reflect the observations. The models used vary in metallicity, radiative transfer schemes, and physical phenomena account for such as magnetic fields, active cloud formation, and hydrogen dissociation. This work highlights the surprising ways in which the 3D models and the spectroscopic phase curves of WASP-121b match each other and it's implications. This analysis also serves as a guide for how complex atmospheric features will present in JWST spectroscopic phase curves.

Transmission spectroscopy studies benefit from both ground-based and space-based observations as the two offer complementary advantages to one another, and capitalising on this synergy is becoming increasingly important as the field of exoplanet atmosphere characterisation continues to mature. In this study, we build on results from previous space-based spectroscopic observations from HST and JWST and we observe two transits of the warm Neptune-like exoplanet WASP-107 b with ground-based VLT/CRIRES+. Our observations were taken at a resolution of R>120,000 in the K-band (2.0-2.5 um), which is a wavelength range at which this target has never been studied before. Using the cross-correlation technique, we are able to both confirm and expand upon previous detections of several molecular species in the atmosphere of our target, providing further indirect insight into possibly ongoing photochemistry. Considering the unusual parameters of WASP-107 b (whose orbit is eccentric, oblique, and polar), we demonstrate that common assumptions--like the use of cloud-free models as cross-correlation templates, and certain treatment of orbital parameters like eccentricity and argument of periastron--can severely impact the Kp-Vsys detection map. With the aid of simulations, we show that these factors can lead to notable artefacts that hinder the interpretation of planetary signals and may even lead to spurious/incorrect non-detections. This is a key finding as the field continues to move away from primarily studying hot Jupiters in order to also explore smaller and cooler exoplanets of relatively weak signals, and we provide strategic recommendations for future studies of similar targets for both space-based and ground-based observations.

Efforts to infer the atmospheric structure and composition of sub-Neptunes have often been hampered by the presence of feature-attenuating aerosols. Though the exact nature of these aerosols remains enigmatic, both condensate clouds and photochemical hazes have been proposed as possible culprits for the observed featureless transmission spectra of exoplanets like GJ 1214b. Although models that consider clouds or hazes in isolation have struggled to reproduce these spectra, cloud-haze interactions may play a pivotal role in resolving this issue.

During cloud formation via heterogeneous nucleation, cloud species condense on cloud condensation nuclei (CCN) before growing further. On Earth, as on Titan, complex hydrocarbon aerosols significantly contribute to the global CCN budget and thus play a key role in setting the atmospheric albedo and cloud coverage. Analogously, photochemical hazes on sub-Neptunes may act as crucial CCN catalyzing cloud formation. Interactions between clouds and hazes are thus likely to control the overall abundance and distribution of aerosols, with direct implications for interpreting atmospheric observations. Thus, a detailed understanding of the impacts of cloud-haze interactions is of first-order importance in interpreting sub-Neptune atmospheric observations.

Motivated by this prospect, we have developed the first 2D bin-scheme microphysical model of cloud formation on sub-Neptunes that includes interactions between clouds and photochemical hazes. Adapted from the Community Aerosol and Radiation Model for Atmospheres (CARMA), our model includes longitudinal and vertical cloud and gas transport. Using this first principles model, I will present trends in aerosol coverage and particle size distributions that emerge across fundamental planetary properties such as temperature, atmospheric composition, and mixing. Our results not only have important bearings on the interpretation of sub-Neptune atmospheric spectra but may also help inform the targets that are ideal for future in-depth characterization with JWST.

Lava planets are among the most peculiar exoplanets discovered to date. Their bulk density suggests an Earth-like structure. With orbital periods of approximately one day, lava planets may serve as better targets for studying rocky planets than other Earth analogs that orbit much farther from their star. Lava planets are tidally locked into synchronous rotation, causing their dayside to be hot enough to melt and vaporize silicate rocks.

Several studies have used melt-vapor equilibrium models to predict the composition and spectral characteristics of lava planet atmospheres. The composition of such an atmosphere is determined by that of the underlying magma ocean, which is intimately linked to the dynamics of the silicate interior. However, the internal dynamic of lava planets is poorly understood.

We conducted multiphase fluid dynamics simulations to investigate the compositional evolution of lava planet interiors. Our model incorporates recent advances in mineral physics, including high-pressure equations of state that control melts and solid densities, and phase diagrams that govern the melting behavior of silicates.

When the magma ocean is global and well-mixed, its composition reflects that of the bulk silicate mantle. However, as the mantle crystallizes, the composition of the magma ocean evolves, leaving a shallow dayside magma ocean enriched in FeO and other incompatible elements. Interestingly, this enriched layer becomes buried in the deep mantle as FeO-rich liquids are denser than silicate solids at high pressure. Our simulations indicate a correlation between the dayside atmospheric composition and the nightside surface temperature, which probes the internal temperature of the planet.

While characterizing the atmospheric composition remains challenging with telescopes such as James Webb, measuring the nightside temperature is readily accessible and might offer invaluable constraints on the thermochemical evolution of these distant rocky worlds.

Sub-Neptunes are the most common short-period (P < 100 d) planets in the galaxy, but their evolutionary histories remain a major mystery for the field. Demographic evidence alone cannot currently distinguish the roles of photoevaporation, core-powered mass loss, and gas-poor formation in sculpting the distribution of sub-Neptune sizes and masses. Observations of atmospheric escape processes in action can help clarify the physical mechanisms responsible for sub-Neptune evolution. We have recently formed the WINERED Helium Consortium to help advance the field towards a better understanding of these processes. Using the Magellan/WINERED instrument, members of our consortium are making precise measurements of metastable helium, a tracer of atmospheric escape that can help diagnose physical conditions in the upper atmospheres of exoplanets. By measuring metastable helium absorption to at least 0.2% precision in a sample of 50 sub-Neptunes, we will help disentangle the roles of photoevaporation and core-powered mass loss in a statistical sample of small planets, all while delivering a sample of planets with known low-Z atmospheres that are suitable for targeted JWST followup. By the start of Exoclimes VII, this effort will be almost halfway done. We will present some of the first results from this extensive observational campaign, including a high-SNR detection of helium in a planet only twice the size of the Earth--the smallest planet to have its atmosphere robustly detected from the ground to date.

Rapid rotated brown dwarfs and exoplanets often exhibit complex atmospheric phenomena such as clouds, storms, and chemical heterogeneity, akin to weather on Earth. These weather patterns could result in as large as 10%-30% near-infrared spectral variability. Understanding the origins of this variability provides key insights into their atmospheric dynamics and structures. Atmospheric retrievals are commonly used to infer quiescent properties from time-averaged, snapshot spectra, but resolving rotational variability demands retrievals on time-series spectra, which presents substantial computational challenges. Here, we present a novel time-resolved spectral inversion strategy that leverages atmospheric retrieval techniques applied to PCA eigen-spectra. Using simulated time-series data, we demonstrate that key drivers of variability, such as inhomogeneities in cloud, chemical abundances, and temperature structures, can be accurately recovered simultaneously. We also present preliminary results from applying this method to time-series data from JWST/NIRISS-SOSS.

By the end of the decade the first light of the Extremely Large Telescope (ELT) will revolutionise exoplanet science, providing unprecedented characterisation of exoplanet atmospheres at high spectral resolution (HRS; R~100,000). To date, HRS has largely been confined to the study of Jupiter-mass planets in the optical and near-infrared (0.4-3.5 microns). However, first light instrumentation on the ELTs (e.g. METIS) will observe at novel M-band wavelengths (3.5-5 microns) and will target sub-Neptunes and super-Earths with high mean molecular weight (MMW) atmospheres. Here I present two projects opening these new parameter spaces for HRS, in preparation for the imminent arrival of the ELTs. I will first present results from our CRIRES+ survey of directly imaged companions at M-band wavelengths, pioneering the wide-scale use of this spectral range. Alongside CO and H2O, the M-band also provides a unique sensitivity to gaseous SiO which acts as both a probe of condensation processes, and as a tracer of the rock/vapour accretion history of giant exoplanets for targets across a range of temperatures. Through precise measurements of SiO abundance we constrain the formation pathway of each target through novel silicate abundance ratios, and benchmark these against established formation tracers including C/O and 13C/12C. Second, I will demonstrate the sensitivity of HRS to sub-Neptune atmospheres around M-dwarf hosts. Using CRIRES+ K-band spectra I constrain the atmosphere of the transiting warm sub-Neptune GJ 3090 b to have a high MMW (>7.1 g/mol) or high-altitude aerosols, further highlighting the diversity of atmospheres within the sub-Neptune population. These results open up the study of sub-Neptune targets from the ground and show that HRS can provide complementary and competitive constraints to JWST. The exploration of new parameter spaces with HRS is vital preparation for the ELTs and yields novel scientific results.

To study Earth as an exoplanet, we analyze its spatially integrated visible reflected spectrum using Earthshine, the sunlight reflected off Earth and observed on the dark side of the Moon. This approach enables observations at varying phase angles as the Sun-Earth-Moon geometry changes. We simulate these observations with the 3D radiative transfer model MYSTIC, incorporating the complexity and variability of Earth's surface and atmosphere. A key element of this work was the development of HAMSTER, the first dataset to capture spatial and temporal variations in Earth's surface albedo across wavelengths. Additionally, we developed a novel 3D cloud generator based on ERA5 reanalysis data, providing accurate models of patchy cloud distributions.

By comparing these simulations with Earthshine observations, we evaluate the sensitivity of various planetary features in characterizing habitability. Decade-long Earthshine monitoring reveals seasonal variability, cloud cover dynamics, and differences in land and ocean geometry driven by planetary rotation. Our findings underscore the necessity of detailed cloud and surface modeling to match observations, highlighting the pivotal role of cloud radiative response and surface albedo in interpreting Earth's reflected light.

Through a comparison of spectroscopy and spectropolarimetry, we propose an optimal strategy for detecting biosignatures on Earth-like exoplanets in future missions. These insights are particularly relevant for upcoming reflected light missions, such as the Habitable Worlds Observatory (HWO), which aim to characterize Earth-like planets. By refining our understanding of Earth's radiative properties, we pave the way for interpreting observations of distant worlds.

To aid the search for atmospheres on rocky exoplanets, we should know what to look for. An unofficial paradigm is to anticipate CO2 present in these atmospheres, through analogy to the solar system and through theoretical modelling. This CO2 would be outgassed from molten silicate rock produced in the planet's mostly-solid interior—an ongoing self-cooling mechanism that should proceed, in general, so long as the planet has sufficient internal heat to lose.

Outgassing of CO2 requires relatively oxidising conditions. Previous work has noted the importance of how oxidising the planet interior is (the oxygen fugacity), which depends strongly on its rock composition. Current models presume that redox reactions between iron species control oxygen fugacity. However, iron alone need not be the sole dictator of how oxidising a planet is. Indeed, carbon itself is a powerful redox element, with great potential to feed back upon the mantle redox state as it melts. Whilst Earth is carbon-poor, even a slightly-higher volatile endowment could trigger carbon-powered geochemistry.

We offer a new framework for how carbon is transported from solid planetary interior to atmosphere. The model incorporates realistic carbon geochemistry constrained by recent experiments on CO2 solubility in molten silicate, as well as redox couplings between carbon and iron that have never before been applied to exoplanets. We also incorporate a coupled 1D energy- and mass-balance model to provide first-order predictions of the rate of volcanism.

We show that carbon-iron redox coupling maintains interior oxygen fugacity in a narrow range: more reducing than Earth magma, but not reducing enough to destabilise CO2 gas. We predict that most secondary atmospheres, if present, should contain CO2, although the total pressure could be low. An atmospheric non-detection may indicate a planet either born astonishingly dry, or having shut off its internal heat engine.

The last few years in exoplanet science has been remarkable due to the transformative nature of JWST to obtain exoplanet spectra. Nonetheless, ground-based high-resolution Doppler-resolved spectroscopy has undergone a quieter yet important revolution. This technique uses the large Doppler-shift of a planet relative to its host star and telluric lines to disentangle the exoplanet's spectrum. For many years, this technique was only applicable to a small number of targets, a small number of species, and could only tell us about the presence or absence of atmospheric species. However, the combination of new instruments, bright targets, and the development of appropriate retrieval techniques has dramatically changed this. In fact, even in the JWST era ground-based telescopes still lead the way in the number of atmospheric species detected, and have also provided the most precise constraints on key observables such as C/O ratios. My group were one of the pioneers of high-resolution retrievals of transmission spectra using a fully Bayesian framework. I will give an overview of our retrieval framework, and discuss new developments that have enabled us to explore the 3D nature and dynamics of the ultra-hot Jupiters WASP-121b and MASCARA-1b in both emission and transmission, as well as obtain precise constraints on their atmospheric compositions. I will also introduce a new filtering technique in development that is designed to more effectively target stellar and telluric signals while preserving the planet's signal. This will be a step towards more precise and robust retrievals and towards applying these techniques to cooler planets with the ELTs.

With the increasing quality and quantity of spectroscopic observations of exoplanet and brown dwarf atmospheres, recent studies suggest that their potentially distinct formation pathways may be imprinted in their elemental and isotopic ratios. This work is part of the ESO SupJup Survey, which aims to disentangle the formation pathways of super-Jupiters and isolated brown dwarfs through the analysis of their chemical and isotopic ratios. In this study, we aim to characterize the atmospheres of two young L4 dwarfs 2MASS J03552337+1133437 (2M0355) and 2MASS J14252798-3650229 (2M1425) in the AB Doradus Moving Group. This includes constraining their chemical composition, 12CO/13CO ratio, pressure-temperature profile, surface gravity, and rotational velocity, among other parameters. We have obtained high-resolution CRIRES+ K-band spectra of these brown dwarfs, which we analyze with an atmospheric retrieval pipeline. Atmospheric models are generated with the radiative transfer code petitRADTRANS, for which we employ a free-chemistry approach, which we couple with the PyMultiNest sampling algorithm to determine the best fit. We report robust detections of 13CO (13 & 8 sigma) and HF (11.6 & 15.8 sigma) in 2M0355 and 2M1425, respectively. Both objects have similar overall atmospheric properties, including their 12CO/13CO isotope ratios of 95.5+/-7 for 2M0355 and 109.6+/-10 for 2M1425. The most notable difference is the robust evidence of CH4 (5.5 sigma) but no H2S (< 2.3 sigma) in 2M1425, in contrast to 2M0355, where we retrieve H2S (4.6 sigma) but not CH4 (< 2.2 sigma). We also find that 2M1425 is 50-200K hotter than 2M0355 and has a higher surface gravity. Future studies will put them into the context of other objects observed as part of the ESO SupJup Survey.

Geochemical evidence suggests that the Moon, Mars, Venus, and Earth went through an early 'magma ocean' (MO) stage. These conditions are also expected to arise early in the lifetimes of all terrestrial-mass exoplanets, whether from accretion or large impacts. Interactions between MOs and overlying atmospheres leads to an evolving feedback of outgassing, greenhouse forcing, and mantle melt fraction.

We have developed a coupled interior-atmosphere model for the evolution of MOs and their atmospheres. Our results thus far have shown that atmospheres overlying MOs have variable compositions during/after solidification due to outgassing of volatiles. Despite much of the existing literature assuming H2O or CO2 to be the dominant gases, H2 rich atmospheres induce a greenhouse effect which can prevent/stall solidification. The geochemical properties of the mantle exert significant control over MO evolution through interactions with the atmosphere.

Previous works have generally assumed that these atmospheres are convective. With a new radiative-convective atmosphere model, we investigated convective shutdown on TRAPPIST-1 c and HD 63433 d. The former is expected to have solidified within 100 Myr, while the latter may maintain a permanent MO to this day, even in the presence of deep isothermal layers. Absorption features of CO2 and SO2 within synthetic emission spectra are associated with mantle redox state, which shows that that future observations of terrestrial-mass planets may provide constraints on the geochemical properties of contemporary MOs analogous with the early Earth. Critically, the synthetic spectra arising from these models show that cool near-isothermal stratospheres can mimic the emission of an atmosphere-less body, despite thick atmospheres and surfaces compatible with a permanent MO. Shutdown of atmospheric convection also has implications for compositional mixing and chemistry in these atmospheres.

https://arxiv.org/abs/2411.19137

https://arxiv.org/abs/2412.11987

There are large uncertainties in cloud properties and their distributions in exoplanet atmospheres. These uncertainties propagate into the impacts clouds have on atmospheric structure via reflection and absorption of light (radiative feedback) and into the spectra we observe. It has been hard to fully explore these uncertainties, however, due to the computational expense of cloudy 3-D models. This makes it difficult to quantitatively link cloud modeling assumptions to the (3-D) outcomes. We will present a large grid of hot Jupiter general circulation models with radiative feedback from parameterized clouds, testing the importance of varying cloud assumptions such as particle size, the set of species considered, the efficiency of vertical mixing, and the total cloud formation efficiency. We will present this analysis for two example planets, one with fairly uniform day-night cloud coverage and one with clouds primarily on the night side. We will present the global properties (Bond albedos, day-night heat transport efficiencies) and cloud structures of the modeled atmospheres, and present preliminary insights into how well JWST transmission or emission spectra can differentiate between cloud properties.

We present the first JWST spectroscopic study of the warm sub-Neptune GJ3090b (2.13 Earth radii, Teq ~ 700 K) which orbits an M2V star, making it a favourable target for atmosphere characterisation. Sub-Neptunes, the most common planet type, remain poorly understood. Their atmospheres are expected to be diverse, but their compositions are challenging to determine, even with JWST.

We observed four transits of GJ3090b; two each using JWST NIRISS/SOSS and NIRSpec/G395H, yielding wavelength coverage from 0.6 - 5.2 microns. We detect the signature of metastable helium escape at a statistical significance of 5.5 sigma with an amplitude of 434 \( \pm \) 79 ppm. Based on predictions from solar-metallicity forward models, this amplitude is significantly smaller than expected, suggesting high-atmospheric metallicity which is expected to lower the mass loss rates.

Furthermore, we find that stellar contamination, in the form of the transit light source effect, dominates the NIRISS transmission spectra, with unocculted spot and faculae properties varying across the two visits separated in time by approximately six months.

In this talk, we will present our observations with its first JWST detection of Helium escape as well as discuss the atmospheric inferences of GJ3090b using both retrieval and self-consistent model grid analyses.

Transit spectroscopy has traditionally relied on the integration of multiple transits to achieve the signal-to-noise ratios necessary to resolve atmospheric features. However, only single-transit observations with high signal-to-noise ratio make it possible to disentangle the complex chemistry and dynamics that go beyond global trends by spatially resolving the limbs.

Using ESPRESSO's superpower—the ability to combine light from all four Unit Telescopes of the Very Large Telescope—we achieved phase-resolved observations of WASP-121 b's transmission spectrum, uncovering new insights into its atmosphere. These observations revealed the presence of neutral titanium, a species previously thought to be depleted due to cold trapping. Together with vanadium, titanium plays a crucial role in hot gas giant atmospheres, as their oxides are key drivers of temperature inversions. By resolving the planetary velocity traces of titanium and vanadium throughout the transit, we gained a unique glimpse into their local spatial and dynamical distribution, contrasting with predictions from General Circulation Models.

This work underscores the transformative potential of upcoming facilities like the ELT to reveal hidden chemical and dynamical features, paving the way for a new era of atmospheric studies with unprecedented precision.

Sub-Neptune planets are among the most common exoplanets, but our understanding of their atmospheric structure and dynamics remains limited. Previous modeling studies of hydrogen-rich atmospheres, including those of the Solar System giant planets, suggest that episodic storms and precipitation occur when the mean molecular weight of the atmosphere is below 6 g/mol.

In this work, we focus on understanding the occurrence and frequency of episodic storms and vertical mixing in hydrogen-rich atmospheres by exploring a wide parameter space. Using 3D convection-resolving simulations, we examine initial surface temperatures ranging from 250 to 800 K to identify the temperature regimes where episodic storms arise. The results provide insights into how episodic mixing influences cloud formation, the distribution and detectability of condensibles, and the atmospheric temperature profile.

Time-resolved observations of directly imaged exoplanets provide unique windows into their atmospheric dynamics and cloud patterns. Inspired by the spectacular weather reports of nearby brown dwarfs delivered by JWST, we have carried out high-contrast imaging and spectroscopic monitoring observations of directly imaged exoplanets using NIRCam coronagraphic mode and NIRSpec IFU. These observations enable rotational phase mapping to identify cloud and disequilibrium structures, and detection of weather patterns through measurements of the planets' spectral and color evolution.

My presentation will focus on a 20-hour high-contrast dual-band monitoring campaign of the iconic directly imaged planet Beta Pic b using JWST's NIRCam coronagraph. I will demonstrate the order-of-magnitude improvement in high-contrast imaging precision for detecting planetary variability signals enabled by JWST and discuss new insights into the atmospheric dynamics and clouds of this young giant world gained from the planet's light curves. With JWST emerging as an "exoplanetary weather satellite," I will use this pilot study to discuss the future prospects of a high-contrast monitoring survey to systematically unveil the dynamics, climate, and weather in directly imaged exoplanets.

JWST has reopened our window to mid-infrared studies of substellar atmospheres. Of particular interest are the young low-mass brown dwarfs, which are physical analogs of the directly imaged planets. In the absence of a host star to contribute photon noise, these "rogue" or "free-floating" planets can be studied with exquisite signal-to-noise data. Here we present our study of the variable free-floating planet PSO 318 with JWST spectra from 1-20 µm. We show that the atmosphere is characterized by patchy silicate clouds and that the data are sensitive to the structural properties of the cloud grains, such as their shape and degree of polimerization of the SiO4 tetrahedrons. At the same time, we find that the data quality and required model complexity challenge our state-of-the-art inference techniques, such as nested sampling retrievals. We discuss the implications of this finding and how it might be addressed in the future. An alternative that is already available today is to partition the data into wavelength intervals that allow answering "isolated" questions that do not depend on the full wavelength range. Using such an approach, we show how cloud properties can be characterized without performing a full retrieval, thereby speeding up analyses by many orders of magnitude. JWST and future ELT data require us to rethink how we characterize planets, and our approach is one example of how this may happen.

Temperate sub-Neptunes, such as K2-18b, located within the Habitable Zone, have gained significant attention due to their potential to support life. The cool equilibrium temperatures of these planets slow down chemical reactions (chemical timescales >>10^10 s) in the upper atmosphere (< 1 bar), making the chemical abundances in this region strongly influenced by vertical transport from deeper layers. In this study, we used a three-dimensional (3D) general circulation model to investigate the transport-induced chemistry and vertical mixing in temperate gas-rich sub-Neptunes, using the parameters of K2-18b as an example. We model K2-18b assuming 180 times solar metallicity and consider it as either a synchronous or an asynchronous rotator, exploring spin-orbit resonances of 2:1, 3:1, 6:1, and 10:1. We find that vertical transport significantly affects the chemical structure, with quenching pressures between 10 and 1 bar, increasing the abundances of CO2 and CO (~10^-3) in the upper atmosphere compared to their equilibrium values (< 10^-10), and horizontal winds further homogenize the chemical composition zonally in this region. Molecular abundances in the photosphere remain consistent across different rotation periods. We employ a passive tracer in the model to estimate the one-dimensional (1D) equivalent eddy-diffusion coefficient (Kzz) of K2-18b, providing a parameter useful for future 1D atmospheric models. Additionally, synthetic transmission spectra generated from our model are compared with recent JWST observations. This work offers a 3D perspective on transport-induced chemistry in temperate sub-Neptunes and derives vertical mixing parameters to support 1D modeling.

The growing number of discovered hot and super-dense sub-Neptunes, found in the hot-Neptune desert, offers us an unprecedented opportunity: that of probing the potential exposed cores and mantles of giant planets. Indeed, it is difficult to explain how such massive planets can avoid triggering runaway gas accretion and remain small. Their large masses also make them mostly resilient to photoevaporation, meaning that either they lost their extended hydrogen envelope through energetic events such as giant planetary impacts, or they formed later in the lifetime of the protoplanetary disk, when most of the gas had dissipated. The hot and dense sub-Neptune TOI-824b was recently highlighted as a key target to study this population of planets, thanks to Spitzer secondary eclipse observations revealing a hot dayside and hints of a metal-rich envelope on this planet, consistent with expectations for an exposed mantle scenario. In this talk, we present the very first thermal emission spectrum of a hot and dense sub-Neptune, TOI-824b, using two JWST eclipse observations of the planet with the NIRSpec/G395M instrument. Early analyses of TOI-824b's 2.7-5.2 um emission spectrum are in agreement with the hot dayside temperature and metal-rich envelope hinted at by the Spitzer eclipses. The emission spectrum also shows hints of carbon dioxide absorption, opening the door to temperature structure and C/O measurements, that will allow us to characterize the formation history and nature of this potential exposed planetary mantle.

JWST has enabled the first in-depth characterisations of the atmospheres of sub-Neptune exoplanets, revealing diverse chemistry and clouds that challenge our theoretical understanding of these objects. Multi-planet systems offer an intriguing opportunity to compare planets whilst controlling for the conditions of the protoplanetary disk in which they formed. I will present the results from our JWST program on the TOI-125 system, showing one transit each of planets b and c, two similar-sized sub-Neptunes in a resonant system, recently observed with NIRSpec G395. Preliminary reductions show substantial molecular features that differ between the two planets, revealing an exciting compositional diversity in their atmospheres. I will show results from our retrieval analysis and compare our abundance constraints to predictions from theoretical models. I will also discuss the implications of our results on the formation history of these planets, as well as conclusions we can draw about their interior compositions. In addition, I will present results from our JWST program on another sub-Neptune multi-planet resonant system, and compare our constraints and conclusions across the two systems. Finally, I will also present theoretical predictions of sub-Neptune chemistry and compare these to all the JWST observations so far, reviewing the results and summarising the most promising avenues of research for developing our understanding of these atmospheres.

Until recently, the three-dimensional nature of close-in exoplanet atmospheres has only been accessible with phase curve measurements. However, the advancement of ground-based, ultra-stable, high-resolution spectrographs alongside the launch of JWST has now enabled the study of this regime with both transit and eclipse observations.

As a planet transits its host star, the morning limb passes the stellar disk before the evening limb. For tidally-locked planets, the morning limb is permanently redshifted, whereas the evening limb is permanently blueshifted. This separation in velocity allows the line contributions of these regions to be modelled independently - mapping the longitudinal variation of atmospheric parameters. We successfully applied this technique to multiple transits of the ultra-hot Jupiter WASP-76b with ESPRESSO and have consistently constrained the abundances of multiple atomic and molecular species across the planet's hemispheres.

Alternatively, at low-resolution, this spatial variation across the planet's hemispheres imprints asymmetries on the measured transit light curves. Using transmission strings (Grant & Wakeford, 2022) we can measure the varying scale heights of the morning and evening limbs and infer the longitudinally varying condensation chemistry. We demonstrate this by exploring variations in cloud chemistry from 0.6 to 12.0 μm along the terminator angle of the hot Jupiter WASP-17b. With the enhanced sensitivity of JWST, this technique can also be applied to colder, more distant worlds, whose multi-dimensionality is inaccessible via phase curves. We aim to map these longitudinal asymmetries for the Neptune-mass planet HAT-P-26b.

A primary motivation behind the measurement of exoplanet atmospheres is to learn about planet formation and evolution. However, this is challenged by the uncertainties and degeneracies inherent to protoplanetary disc composition, planet formation, and planetary evolution. To determine whether atmospheric composition is actually a reliable tracer of formation history, we are undertaking a survey with JWST (BOWIE-ALIGN) that is comparing the compositions of planets which we are confident have different evolutionary histories due to their different orbital alignments (obliquities) around hot stars, where tidal re-alignment is inefficient. It is believed that aligned planets are the outcome of disc migration, while misaligned ones arise from high-eccentricity migration. This dichotomy leads to differences in the material they accrete during their evolution which, theory predicts, should lead to differences in their atmospheric compositions. I will present the results for the first BOWIE-ALIGN target, the misaligned hot Jupiter WASP-15b. Our JWST transmission spectrum reveals significant H2O and CO2 absorption consistent with super-solar metallicities and solar C/O, together implying planetesimal accretion, along with evidence for complex sulphur photochemistry. Once complete, BOWIE-ALIGN will test if composition depends on migration and thus if composition can be used to make robust inferences about formation.

By resolving individual molecular lines, high resolution spectroscopy (HRS) of exoplanets can constrain chemical compositions, thermal structure and velocities in their atmospheres. Providing favourable contrasts and large orbital velocities, Hot Jupiters (HJs) have been particularly fruitful targets for HRS, producing high quality datasets in both emission and transmission. Today, a key goal for HRS study of HJs is to robustly extract spatial and dynamical information revealing their 3D structure and climate patterns. Particularly in emission, a significant challenge to this goal lies in the need to sum signal over many phases of the planets' orbits, such that extracted spectra are averaged over the planetary surface. Eclipses, where the planet is progressively hidden/revealed from behind its star, could provide a unique solution for obtaining spatially resolved spectra. While it has been successfully applied at low spectral resolutions to map temperatures on several HJs, eclipse mapping has yet to be attempted with HRS.

Here, we present preliminary results from the first observational campaign to perform high resolution eclipse mapping on the ultra-hot Jupiter (UHJ) WASP-33b. Based on injection-recovery in real data observed through an ongoing large program with SPIRou, we find that 20 eclipses would be required to confidently constrain the planet's rotation and map its CO emission. With a total of 8 eclipses observed so far, we strongly detect the planet's CO emission while it is being occulted, in line with our expectations. Our results demonstrate the great potential of eclipses with HRS to map molecules and winds on exoplanets in the near future, paving the way for the upcoming ELT which will be able to perform this on a much wider population of HJs. Additionally, leveraging much shorter time series than typical HRS observations, our analysis yields interesting lessons on optimal stellar correction and pre-processing applicable to slow-moving planets.

Giant planets are some of the easiest planets to detect, but it is perhaps the planets with the largest signals that are the hardest to explain. Giant Exoplanets around M-dwarf Stars (GEMS) are generally easily detectable and characterizable given their advantageous radius and mass ratios compared to their host stars, but are some of the hardest planets to explain within the framework of current planet formation models. These massive planets must have formed with exceptional efficiency, given the small protoplanetary disks surrounding M-dwarfs, sometimes requiring planet formation efficiencies seemingly greater than 100%. Yet, these planets do exist. It was understanding this unique population of planets that motivated the JWST GEMS program, the largest JWST Cycle 2 GO program dedicated to exoplanet science, to observe seven of these mysterious objects with NIRSpec/PRISM to constrain their atmospheric compositions and provide insight into their formation histories. Much like small planets around M-dwarfs, these planets also show strong stellar contamination signals, but unlike many of their smaller counterparts, these giant planets still have strongly detectable atmospheric features. Our observations reveal atmospheric compositions that are markedly different from those of many giant exoplanets orbiting earlier-type stars, with very little detectable atmospheric water, high C/O ratios, and very low atmospheric metallicities. Here, I will present results from several planets in our sample, highlighting spectra that are as puzzling as the formation histories of these enigmatic worlds and early lessons learned.

Ultra Hot Jupiters (UHJs) are unique extrasolar laboratories excellent for studying extreme climate conditions. Their complex atmospheres are 3D with regional differences in temperature-pressure structure, chemistry, and wind pattern. High-resolution spectroscopy is highly sensitive to the line position, strength, and shape, which are affected by these 3D effects. To analyze UHJ atmospheres with high-resolution spectroscopy, retrieval frameworks have been increasingly deployed. However, these frameworks rely on computationally fast 1D models despite the inherent 3D nature of UHJ atmospheres. In this study, we assessed the robustness of 1D retrievals by simulating emission observations using a 3D Global Circulation Model of the well-known UHJ WASP-76 b. We find that although a 1D retrieval can broadly recover atmospheric conditions, constraints are not a homogeneous average of all regional and phase-dependent information. Instead, they are more sensitive to those regions with large thermal gradients that do not necessarily coincide with the strongest emitting regions. This results in a biased view of the bulk atmospheric properties that need to be addressed to correctly interpret planet formation pathways and their atmospheric dynamics.

For three decades, the exoplanetary community has chased after the same key questions: How do planets form? Is our solar system unique? Where should we look for life? To answer these, we've turned to exoplanetary atmospheres, which are both informative and accessible. By fitting spectral models to observed atmospheric spectra (a process often called a "retrieval"), we have been able to measure the temperature structure and chemical abundances of dozens of exoplanets. Now, JWST's unprecedented data precision and broad wavelength coverage invite us to begin answering the questions we have had since the beginning.

However, interpretation of JWST's powerful data comes with its own challenges, including unexpected absorbers at wavelengths never previously explored and stellar contamination in multiple exoplanetary spectra. Models not equipped to handle these new data unknowns, or those that use incomplete physics, can inadvertently bias our inferences and lead us astray. We therefore run the risk of misunderstanding dozens of new observations and limiting our science return from JWST. Due to this limitation, we require a more flexible modeling approach that can account for unknowns as they arise.

We introduce a method to provide reliable inferences by adapting the data-model comparison used in retrievals. With non-parametric models, we can identify and marginalize over regions of data-model disagreement, allowing us to identify deficiencies in our models and provide more credible estimates of planetary properties. We apply our new framework to well-validated Cycle 1-3 JWST observations that have resulted in conflicting interpretations among the community. We demonstrate the ability to reconcile these interpretations by marginalizing over both data unknowns and model deficiencies to provide new, unbiased inferences. This method will be crucial to determining improvements needed in modeling efforts, as well as observational priorities and capabilities for future observatories such as HWO and Pandora.

Planetary atmospheres are probably one of the most diverse type of astronomical objects to study. For the particular case of gas giant exoplanets exoplanets, HST and Spitzer gave us a broad view of their diversity at the population level whereas JWST is now confirming the insights on selected exoplanets, while building a population on its own.

Understanding the global properties of hot Jupiter atmospheres require complex, 3D models and, so far, the modelling framework has never been scaled up to study the whole population. Here, I will present a population study of hot Jupiters, both from a global circulation modelling perspective and a data perspective.

From the model side, I will showcase a new publicly available grid of 500 global circulation models. I will show that the combined effect of parameters such as rotation period, metallicity or atmospheric drag can impact the day/night heat transport of these planets as much as the changing irradiation temperature. I will further show how population level statistical trends observed with Spitzer can be recovered by the models, hinting towards a sharp onset of gaseous TiO around an equilibrium temperature of 1800K.

From the data side, I will discuss how recent JWST phase curve observations and ground-based high-spectral resolution observations of transiting ultra-hot Jupiters allows us to answer some questions that were not accessible with previous telescopes. In particular, I will show observational proof of circulation-driven disequilibrium chemistry in NGTS-10b and the presence of a trend in directly measured wind-speeds vs. equilibrium temperature that confirms the presence of a magnetically-driven atmospheric circulation regime in the population of ultra-hot Jupiters.

With the increased sensitivity and wavelength coverage provided by JWST, performing retrievals on its datasets becomes increasingly time consuming. The timely analysis of these observations requires speeding up retrievals, so that multiple retrievals of varying complexity can be run for each individual dataset. In recent years, machine learning has positioned itself as the most promising method to overcome this issue.

In this talk, we introduce FlopPITy, a machine learning based retrieval method that accelerates retrievals by up to 40 times. FlopPITy is a likelihood-free inference method, based on iteratively training a normalising flow. We also discuss the benefits of the lack of likelihood and why the community might want to consider moving in this direction.

The significant speed-up achieved by FlopPITy also allows for the use of more complex and computationally expensive forward models in retrievals, enabling new science that is not feasible with current methods. In particular, it enables the comparison of self-consistent models, that are too slow for traditional retrievals, with observations. This serves as a test-bed for our knowledge of physical processes of exoplanet atmospheres.

To showcase the power of this new technique, we present an analysis of the complete 0.5-12 microns JWST transmission spectrum of WASP-39b. This is one of the first simultaneous retrievals on the broadband transmission spectrum of a hot Jupiter, and the first ever retrieval on transmission spectra with self-consistent models. In particular, we assume radiative-convective equilibrium, microphysical cloud formation, and chemical equilibrium adjusted to account for the effect of vertical mixing and photochemistry.