PlanetProfile: Self-consistent interior structure modeling for terrestrial bodies in Python

The open-source PlanetProfile framework was developed to facilitate investigation of the interior structure of icy moons based on self-consistency and comparative planetology. The software, originally written in Matlab, relates observed and measured properties, assumptions such as the type of materials present, and laboratory equation-of-state data through geophysical and thermodynamic models to evaluate radial profiles of mechanical, thermodynamic, and electrical properties, as self-consistently as possible. We have created a Python version of PlanetProfile. In the process, we have made optimization improvements large and small and built in parallelization and parameter-space search features to take advantage of fast operation for investigating unresolved questions in planetary geophysics, in which many model inputs are poorly constrained. The new Python version links to a variety of other scientific software packages, including for evaluating equation-of-state data, magnetic induction calculations, seismic calculations, and spacecraft data analysis. Geophysical and thermodynamic models implemented in PlanetProfile have been reconfigured so as to improve self-consistency and generate the most realistic relationships between depth, pressure, temperature, density, electrical conductivity, etc. Here, we describe the software design and algorithms in detail, summarize model inputs and outputs for major moons across the outer solar system, and discuss new inferences about the interior structure of several bodies. Chiefly, the high values and narrow uncertainty ranges reported for the axial moments of inertia for Callisto, Titan, and Io are difficult to reconcile with self-consistent models, requiring highly porous rock layers that are equivalent to incomplete differentiation for Callisto and Titan, and a high rock melt fraction for Io. This difficult matching, which may imply the bodies are not in hydrostatic equilibrium, is even more pronounced with the more realistic models in the new Python version. Radial profiles for each model are provided as a Zenodo archive.