Table 1 free parameters in our model, the nominal parameter value for illustrative calculations (Figs. 24), and the full Monte Carlo range shown in later calculations (Fig. 5)

From: The erosion of large primary atmospheres typically leaves behind substantial secondary atmospheres on temperate rocky planets

Parameter name

Variable

Nominal point estimate

Monte Carlo range

Explanation/Reference

Initial conditions

Initial carbon, C (kg)

4 × 1020

1020–1021.5

Approximate bulk silicate Earth carbon inventory111,112,113

Initial free oxygen, O (kg)

6 × 1021

1021–1022

Approximately Earth-like (ensures post-solidification mantle redox around quartz-fayalite-magnetite buffer if no nebular atmosphere).

Initial hydrogen, H (kg)

4 × 1020

1020–1023

Point estimate is approximate bulk silicate Earth inventory, range encompasses large primary atmosphere.

Initial radionuclide U, Th, and K inventory (relative Earth)

1.0

0.33–30.0

Scalar multiplication of Earth’s radionuclide inventories in Lebrun, et al.27. Allows for modest tidal heating.

Initial mantle FeO (mole fraction)

0.06

0.06

Earth-like value assumed for nominal calculations, but supplementary material investigates broad range of FeO (0.02 -0.2)

Stellar evolution and escape parameters

TRAPPIST-1 XUV saturation time,\({t}_{sat}\)

3.14

\({3.14}_{-1.46}^{+2.22}\) Gyr

XUV evolution parameters drawn randomly from joint distribution101.

Post saturation phase XUV decay exponent, \({\beta }_{decay}\)

−1.17

\(-{1.17}_{-0.28}^{+0.27}\)

XUV evolution parameters drawn randomly from joint distribution101.

Saturated log10(FXUV/FBOLOMETRIC) flux ratio

-3.03

\(-{3.03}_{-0.23}^{+0.25}\)

XUV evolution parameters drawn randomly from joint distribution101.

Escape efficiency at low XUV flux, \({\varepsilon }_{low}\)

0.2

0.01–0.3

See escape section in Krissansen-Totton, et al.28.

Transition parameter for diffusion limited to XUV-limited escape, \({\lambda }_{tra}\)

1.0

10−6–101*

See escape section in Krissansen-Totton, et al.28.

XUV energy that contributes to XUV escape above hydrodynamic threshold, \({\zeta }_{high}\)

50%

0–100%

See escape section in Krissansen-Totton, et al.28.

Cold trap temperature variation, \(\varDelta {T}_{cold-trap}\)

0 K

−30 to +30 K

Cold trap temperature, \({T}_{cold-trap}\), equals planetary skin temperature plus a fixed, uniformly sampled variation, \({T}_{cold-trap}={T}_{eq}{(1/2)}^{0.25}+\varDelta {T}_{cold-trap}\). Here, \({T}_{eq}\) is the planetary equilibrium temperature given assumed albedo.

Thermosphere temperature, \({T}_{thermo}\)

1000 K

200–5000 K*

114,115,116

Interior parameter

Mantle viscosity coefficient

10 Pa

101–103 Pa s*

Solid mantle kinematic viscosity, \({\nu }_{rock}\), (m2/s) is given by the following equation:\({\nu }_{rock}={V}_{coef}3.8\times {10}^{7}\exp (\frac{350000}{8.314{T}_{p}})/{\rho }_{m}\)Here \({T}_{p}\) is mantle potential temperature (K) and \({\rho }_{m}\) is mantle density (kg/m3). See Krissansen-Totton, et al.28.

Albedo

Bond albedo during magma ocean solidification

0.2

0.0–0.2

85,86

  1. *Denotes this variable was sampled uniformly in log space. All others (except stellar XUV parameters) sampled uniformly in linear space.