Enceladus Plume

A model of the time-variable water-vapor plume of Enceladus.

Liquid water in the tidally-flexing south-polar cracks rises and falls as the walls widen and narrow over the 33-hour diurnal cycle. As the walls are periodically covered and uncovered, sublimation and condensation drive a vapor flow out of the crack. The model couples three physical processes — liquid-column dynamics, gas-column dynamics, and wall heat diffusion — to compute the water level and the water-vapor mass flux, and to explain the observed two-peak diurnal plume profile (main peak near mean anomaly 180°, secondary peak near 50°).

The accompanying manuscript is "Aerosols in the vacuum: modelling the time variability of the plume of Enceladus" — see writing/.

Repository layout

enceladus_plume/     Python solver (current version)
cpp/                 C++ performance core + pybind11 bindings (optional, accelerates hot loops)
writing/
  manuscript/        Working LaTeX manuscript (current version)
  reference.tex      Source of the previous draft
  Enceladus_Draft.pdf  Rendered previous draft

The original MATLAB implementation has been retired. It remains available in the git history (removed after the Python port reached parity).

Physics modules (enceladus_plume/enceladus_plume/)

Module	Role
liquid_dynamics/	Water level + entrance velocity solver (adaptive RK, friction integral)
gas_dynamics/	Gas-column solver and pre-computed lookup-table interpolation
heat_diffusion/	1-D wall thermal-conduction layer
physics.py	Shared thermodynamics (vapor pressure, evaporation, surface temperature)
friction.py	Reynolds-dependent friction factors
config.py	Typed dataclass config loaded from YAML
workflows/	full_solver (2022 coupled) and modular_solver (2023 lookup-based)

Quick start (Python)

cd enceladus_plume
pip install -r requirements.txt        # or: pip install -e .
pytest -q                              # run the test suite (use -m "not slow" to skip slow tests)

Run the parameter-sweep pipeline:

./run_pipeline.sh                                  # default sweep -> results/
python run_pipeline.py --config config/run_config_zoom.yaml -v   # zoomed sweep -> results_zoom/

The pipeline runs the liquid dynamics for each (wmin, width-ratio) case, builds or reuses the gas lookup table, interpolates the gas quantities, and writes per-case .npz files plus plot grids and heatmaps. See enceladus_plume/README.md for the full module reference and CLI options.

C++ performance core (cpp/)

The numerically heavy inner loops — the liquid RK integrator with its friction integral, and the gas lookup-table generation (RK4 column solve inside a bisection) — dominate runtime. cpp/ provides a C++ implementation of these loops, exposed to Python via pybind11.

The Python package imports the compiled extension when it is available and falls back to the pure-Python implementation otherwise, so the C++ core is an optional accelerator, never a hard dependency. See cpp/README.md for build instructions.