Heavy element enriched atmospheres and where they are born

Abstract

The heavy element content of giant exoplanets, inferred from structure models based on their radius and mass, often exceeds predictions based on classical core accretion. Pebble drift, coupled with volatile evaporation, has been proposed as a possible remedy to this with the level of heavy element enrichment a planet can accrete, as well as its atmospheric composition, being strongly dependent on where in the disc it is forming. We use a planet formation model which simulates the evolution of the protoplanetary disc, accounting for pebble growth, drift and evaporation, and the formation of planets from pebble and gas accretion. The growth and migration of planetary embryos is simulated in 10 different protoplanetary discs which have their chemical compositions matched to the host stars of the planets which we aim to reproduce, providing a more realistic model of their growth than previous studies. The heavy element content of giant exoplanets is used to infer their formation location and thus make a prediction of their atmospheric abundances. We focus here on giants more massive than Saturn, as we expect that their heavy element content is dominated by their envelope rather than their core. The heavy element content of 9 out of the 10 planets simulated is successfully matched to their observed values. Our simulations predict formation in the inner disc regions, where the majority of the volatiles have already evaporated and can thus be accreted onto the planet via the gas. As the majority of the planetary heavy element content originates from water vapour accretion, our simulations predict a high atmospheric O/H ratio in combination with a low atmospheric C/O ratio, in general agreement with observations. For certain planets, namely WASP-84b, these properties may be observable in the near future, offering a method of testing the constraints made on the planet's formation.

Figures (9)

• Figure 1: Growth tracks showing the orbital evolution of a planet simulated in the disc of WASP-84 for 3 different disc viscosities: $\alpha=10^{-4}$, $\alpha=5\times10^{-4}$ and $\alpha=10^{-3}$. The simulations shown are for a planet which originates at an orbital radius of 1AU and migrates inwards. The top plot shows the mass evolution of the planet with the planet's heavy element mass plotted on the colourbar, The middle plot shows the mass evolution of the planet with the planet's atmospheric Carbon to Oxygen ratio plotted on the colourbar and the bottom plot shows the heavy element mass evolution of the planet with the planet's atmospheric Carbon to Oxygen ratio plotted on the colourbar. The locations of the C and H$_2$O evaporation fronts are shown for the 3 different viscosities. The horizontal red dashed line and shaded region shows the observed mass (top and middle) and the observed heavy element mass (bottom) and their uncertainties. The coloured triangles in the bottom plot show the final simulated heavy element mass for each viscosity. Where the growth tracks are plotted as solid lines indicates that the planet is undergoing pebble accretion while dotted lines indicate gas accretion.
• Figure 2: Mass vs Heavy Element Mass relation for the 10 planets simulated. Blue points show simulations where the initial formation location matched the observed heavy element mass most closely. Data from Thorngren2016 is shown in red (T16). Planets which are simulated in this work are plotted as red squares while those which are not simulated in this work, either because they are less massive than 0.5$M_J$ or because the stellar chemical abundances of their host star are not included in Teske2019, are plotted as pale red diamonds.
• Figure 3: Top row: Heavy element masses of the planets simulated in the discs of WASP-84 (left) and CoRoT-9 (right) as a function of the initial position of the planetary embryo. The observed heavy element mass is shown as a dashed red line, with its uncertainty indicated by a red shaded region. Bottom row: Predicted atmospheric C/O number ratios of the same planets as a function of initial position. The red dashed line shows the best-fit formation location based on the heavy element mass, and the grey region indicates the constrained range of possible formation locations. H$_2$O and C evaporation fronts are shown as dashed black lines in all panels. All simulations use $\alpha = 1\times10^{-4}$.
• Figure 4: Heavy element masses of the planet simulated in the disc of WASP-84 for varying initial positions of the planetary embryo in a disc of $\alpha=1\times10^{-4}$ with a refractory carbon content of 20%. The simulated heavy element mass of the planet does not fall within the uncertainty range of observations for any initial position.
• Figure 5: Predicted initial positions which simulate the observed heavy element masses and the corresponding predicted C/O ratios for the 5 planets simulated which had good constraints on the heavy element content. All simulations use $\alpha = 1\times10^{-4}$. Square points represent the 'inner' formation regions and circles represent the 'outer' formation regions. The points are placed at the initial position which led to the simulations most closely matching the observed heavy element mass. The error bars on the x axis represent the width of the constrained possible formation regions of each planet while the error bars on the y axis show the range in the predicted C/O ratio within each region. WASP-84b is the focus of this work due to its potential observability and is marked with red points and error bars.