Since the first observations of exoplanet atmospheres, clouds have had a major impact on the observed spectra. It is therefore essential to understand the processes that affect their formation, as well as the details of the cloud particles themselves and the impact it can have on atmospheric opacity. The wide range of planetary parameters seen in exoplanets offers a unique laboratory to explore chemistry and stellar environments that are not observable in our solar system. Furthermore, we now know that most exoplanets do not look like solar system planets, instead they are a size intermediate between the Earth and Neptune. These so called ‘Mini-Neptunes’ and ‘Super-Earths’, are what I ultimately aim to study.


Mineral Snowflakes:

Cloud particles are often modelled as compact spheres. This is expected to not be the case, instead being non-compact and irregularly shaped.

Illustration of a compact, spherical, mixed material cloud particle. Materials shown are crystals of materials found in our models, but we do not expect them to look like this.
Illustration of a – perhaps – more realistic cloud particle. At the interface between materials micro-porosity is expected to occur, and the non-spherical shape is modelled by a distribution of hollow spheres.


To explore the inclusion of empty space into the cloud particle, which we term ‘micro-porosity’ we introduce a parameter (fpor) to modify the effective density of the cloud particles, constant across the atmosphere. Optical affects are modelled by including the volume as vacuum, using Mie and effective medium theory.

Refractive indices for Mg2SiO4 with increasing micro-porosity fractions included using effective medium theory.

To include details on irregularly shaped particles we use a statistical distribution of hollow spheres (DHS), a hollow sphere is made up of a vacuum core and a material mantle. Each hollow sphere in the distribution is defined by the ratio of the volume of the material mantle fhol. This distribution is then averaged over using Mie theory for optical properties.

Illustrative figure of hollow spheres with increasing fhol. The volume of the mantle is conserved for all hollow spheres in the distribution.

Looking at the vertically integrated optical depth for a ‘warm gas giant’ Drift-Phoenix profile, we see that micro-porosity affects the silicate features at ~10 μm. As well as this, micro-porosity makes clouds optically thick for wavelengths up to 100 μm, compared to 20 μm for the compact case. Irregularly shape effects (through a DHS) also increases the optically thick wavelength for both cases. Such affects are in the wavelength regime of MIRI and thus may be observable with JWST.

Pressure at which micro-porous and non-spherical clouds become optically thick for a Drift-Phoenix with Teff = 1800 K, log(g [cms-2]) = 3.0 atmosphere. If the cloud never becomes optically thick the bottom of the atmosphere is returned ( ~2 bar). JWST instrument wavelengths are shown by the coloured bars, artist impression of the deployed JWST credit: NASA.

Further details on the albedo of such cloud particles, as well as micro-porous one dimensional cloud structures for a range of planetary parameters, and a Gaussian cloud particle size distribution are explored in the Paper



Global Cloud Distribution on HAT-P-7b:

As part of a collaboration among attendees at Cloud Academy I, we modelled the global distribution of clouds on the Ultra-hot Jupiter HAT-P-7b. We used 97 one dimensional (p, T, Vz)-profiles extracted from a three dimensional, cloud-free, global circulation model.

These profiles were used as input into a kinetic cloud formation model. We find cloud formation on the night-side, but a near absence of cloud formation on the day-side, which can be up to 2500 K hotter.

Local gas temperature around the equator of HAT-P-7b, the star is located to the right and the terminator is indicated by the green dashed line.
Same as left plot, but showing the ratio of cloud particle mass density to atmospheric mass density.

Equatorial super-rotation causes the evening terminator (ɸ = 90.0°) to be warmer than the morning terminator (ɸ = 270.0°), this produces different chemistry on the two terminators. It also causes an asymmetry in clouds, which affects the optically observable depth of the atmosphere.

Optical depth of gas species and clouds, for four key points around the equator of HAT-P-7b

Thus one dimensional transmission spectra retrievals of atmospheric parameters will be biased by ignoring this effect. Further results involving global maps of cloud composition and effect on C/O ratio are explored in Paper I. Disequilibrium processes are explore in Paper II.