Charles Le Losq,Clément Ferraina,Paolo A. Sossi,Charles-Édouard Boukaré
Abstract:Ultra-short-period exoplanets like K2-141 b likely have magma oceans on their dayside, which play a critical role in redistributing heat within the planet. This could lead to a warm nightside surface, measurable by the James Webb Space Telescope, offering insights into the planet's structure. Accurate models of properties like viscosity, which can vary by orders of magnitude, are essential for such studies.
We present a new model for predicting molten magma viscosity, applicable in diverse scenarios, including magma oceans on lava planets. Using a database of 28,898 viscosity measurements on phospho-alumino-silicate melts, spanning superliquidus to undercooled temperatures and pressures up to 30 GPa, we trained a greybox artificial neural network, refined by a Gaussian process. This model achieves high predictive accuracy (RMSE $\approx 0.4 \log_{10}$ Pa$\cdot$s) and can handle compositions from SiO$_2$ to multicomponent magmatic and industrial glasses, accounting for pressure effects up to 30 GPa for compositions such as peridotite.
Applying this model, we calculated the viscosity of K2-141 b's magma ocean under different compositions. Phase diagram calculations suggest that the dayside is fully molten, with extreme temperatures primarily controlling viscosity. A tenuous atmosphere (0.1 bar) might exist around a 40° radius from the substellar point. At higher longitudes, atmospheric pressure drops, and by 90°, magma viscosity rapidly increases as solidification occurs. The nightside surface is likely solid, but previously estimated surface temperatures above 400 K imply a partly molten mantle, feeding geothermal flux through vertical convection.
What problem does this paper attempt to address?
### What problem does this paper attempt to solve?
This paper aims to solve the problem of predicting the viscosity of lava oceans on the surfaces of ultra - short - period exoplanets (such as K2 - 141 b). Specifically, the author attempts to develop a general machine - learning model to predict the viscosity of aluminosilicate melts and apply it to the study of the surface properties of dry lava planets.
#### Main problems and background
1. **Importance of lava oceans**:
- Ultra - short - period exoplanets (USP) such as K2 - 141 b may have huge lava oceans on their day - side, and these lava oceans play a crucial role in the redistribution of internal heat within the planet.
- The properties of lava oceans (such as viscosity) are essential for understanding the planetary structure and atmospheric composition. For example, the James Webb Space Telescope (JWST) can indirectly understand the internal state of the planet by measuring the night - side temperature.
2. **Limitations of existing models**:
- Most current viscosity models focus on predicting the viscosity of crystal - free melts within specific composition, temperature, and pressure ranges and cannot cover the widely varying composition, temperature, and high - pressure conditions on USP planets.
- A new model is needed to handle a wider range of composition, temperature, and pressure, especially for lava oceans under high - temperature and high - pressure conditions.
#### Goals of the paper
1. **Develop a new model**:
- Use a database containing 28,898 viscosity measurement data to train a hybrid model that combines a grey - box artificial neural network (ANN) and a Gaussian process (GP).
- The model can handle various compositions from SiO₂ to multi - component melts and industrial glasses and consider the pressure effect up to 30 GPa.
2. **Application and verification**:
- Apply the new model to the lava ocean of K2 - 141 b and calculate the viscosity under different compositions.
- Evaluate the influence of the internal temperature on the night - side surface, thereby inferring the possibility of the internal state of dry lava planets.
Through these studies, the author hopes to provide an accurate and widely applicable melt viscosity prediction tool to support further research on USP planets and the analysis of observational data.
#### Key formulas
- The Vogel - Tamman - Fulcher (VFT) equation used by the grey - box artificial neural network (Greybox ANN):
\[
\log \eta = A+\frac{B}{T - C}
\]
where \( T \) is the temperature (in K), \( A \) is a common adjustable parameter, and \( B \) and \( C \) are adjustable parameters depending on the melt composition \( X \) and pressure \( P \).
- Definition of the Gaussian process (GP):
\[
f(x)\sim\text{GP}(m(x), k(x, x'))
\]
where \( m(x) \) is the mean function and \( k(x, x') \) is the kernel function (covariance function).
Through these methods, the author hopes to provide a reliable and flexible tool for studying the lava oceans of USP planets.