SW1D: A Python package for seismic surface wave solutions in 1-D elastic media

The 30-second take

• What: The authors wrap a classic 1-D surface-wave eigenvalue and excitation solver in Python, covering Rayleigh and Love dispersion plus frequency-domain seismograms from a moment-tensor point source.
• Why it matters: Reliable surface-wave forward modeling is still scattered across legacy codes; a checked Python package makes that scarce tooling more default in modern seismology workflows.
• Who should care: Seismologists modeling layered Earth structure, ambient-noise and lateral-discontinuity groups, and Python-stack geophysicists.

What the paper actually did

SW1D is a Python version of a legacy code for forward computation of Rayleigh and Love wave dispersion curves and synthetic seismograms in elastic, isotropic, laterally homogeneous media. It solves the surface-wave eigenvalue problem in 1-D with a propagator-matrix technique, then computes surface-wave excitations in the frequency domain for a point source given by its moment tensor. The package is validated with theoretical analyses and benchmarked against a well-established seismological software package. The authors place it in a Python seismology ecosystem aimed at niche jobs such as modeling surface-wave propagation across lateral discontinuities and forward modeling of ambient-noise cross-correlations with arbitrary spatial distributions of ambient sources.

What makes this disruptive

This is infrastructure, not a new Earth model. The disruptive move is making a trusted propagator-matrix stack importable, checkable, and composable with other Python tools. Benchmarking against an established package is the trust claim. Naming specific downstream uses (lateral discontinuities; ambient-noise cross-correlations with arbitrary source maps) tells you why a 1-D solver still matters in a 3-D world: as a building block. Abundance here is access — fewer groups stuck on unmaintained binaries.

Why it matters (outside the lab)

Abundance lens: accurate seismic modeling and monitoring capacity is still unevenly distributed. A documented, benchmarked Python forward solver lowers the cost of teaching, noise modeling, and method papers. Horizon is mid: measurement and open tools, not a climate product. Near-term, adopt it where you already wanted this legacy physics. Medium-term, ecosystem uptake decides if it becomes a default dependency.

Limitations & open questions

The physics is elastic, isotropic, and laterally homogeneous — no anelasticity, anisotropy, or 3-D structure inside the solver itself. Lateral-discontinuity and ambient-noise uses are described as ecosystem roles, not as fully solved 3-D engines in this package. “Well-established” benchmark partner should be identified in the PDF. A port can inherit legacy numerics as well as legacy correctness. Frequency-domain moment-tensor excitation is the source model; other sources need extra work. Preprint / software paper.