From dispersion to interpretable depth constraints

What changed
beneath your array?

Work through the questions behind a velocity-change curve: which depths are sampled, which changes can be distinguished, and how water and acquisition geometry affect the inference.

Start with the critical zone →Explore offshore DAS

Five executed notebooks • 18 computed figures • synthetic experiments at field scales

Overlapping surface-wave sensitivity across the critical zone and volcanic edificeA schematic cable and land sensors above three depth regions, with short and long wavelength curves sampling different overlapping depths.Volcanic edificeOffshore DASWeathered layerFractured rockSubstrateOverlapping depth sensitivityFrequency ≠ a unique depth
Conceptual depth schematic. Numerical sensitivities and resolution appear in the notebooks.

Five questions. Five experiments.

Read the rendered results or download an executable notebook.

01 / CRITICAL ZONE

Which shallow layer changed?

A 10–18 Hz weathered-column experiment: rainfall-like softening, a competing shallow oscillation, and depth resolution with correlated frequency errors.

Computed Rayleigh dispersion, integrated layer sensitivities, and relative eigenfields for a synthetic critical zone
Read the experiment →Notebook ↓
02 / VOLCANO MONITORING

Shallow forcing or deeper change?

Compare elastic-change hypotheses in a basaltic edifice. Inspect frequency coverage, averaging matrices, and posterior depth tradeoffs before assigning a process.

Computed fundamental dispersion and depth sensitivities for a synthetic volcanic edifice
Read the experiment →Notebook ↓
03 / FLUID–SOLID INTERFACE

Could the water explain it?

Verify a finite-water coupled branch against an analytical secular equation, then carry water-column uncertainty into seabed velocity-change inference.

Finite-water Scholte dispersion and relative acoustic pressure shapes
Read the experiment →Notebook ↓
04 / TIME-LAPSE TOMOGRAPHY

What did your map covariance lose?

Compare joint path–depth inference with a two-stage phase-map inversion. They agree when the full covariance is retained in this linear acquisition experiment.

Synthetic station network and straight paths through a four-cell tomography grid
Read the experiment →Notebook ↓
05 / COOK INLET · KKFL-S DAS

From a Scholte ridge to seabed stiffness

Published KKFL-S acquisition parameters meet a synthetic layered seabed. Correct along-cable projection, examine gauge response, fit absolute dispersion, then invert small seabed changes with water uncertainty.

Field acquisition facts are cited. The seabed geology and dispersion picks are explicitly synthetic.

DAS apparent phase speed and axial-strain response versus frequency, incidence angle, and gauge length
Read the offshore DAS experiment →Notebook ↓

Read a velocity-change curve as an inverse problem.

Define the observableThe monitoring operator models direct fundamental-mode phase changes. Coda stretching and MWCS require an additional estimator-specific sensitivity model.

Ask about resolutionLayer sensitivities overlap. Good data fit does not establish a uniquely localized depth change. Examine uncertainty, averaging, and alternative backgrounds.

Separate elasticity from causeTemperature, effective stress, damage, and fluid effects may yield similar elastic changes. Interpretation needs environmental or volcanic observations.

Each notebook is rerun from a fresh kernel before publication; assertions gate deployment. These are small elastic reference calculations, conditional on the stated models and priors. Local execution instructions · Machine-readable execution report