DASCore

A technical preview of version 1.0

Derrick Chambers

14th EGU Galileo Conference — Aussois, 31 Aug – 4 Sep 2026

Acknowledgements

DASCore Contributors:

The people who have contributed to DASCore.

As well as:

INERIS, NIOSH, NSF (grant 2148614), Colorado School of Mines and University College Dublin

The problem

DAS data are awkward long before they are interesting

  • There are many different data formats
  • Managing large file archives is a pain
  • Simple processing functions are often not simple to write
  • Tracking and integrating metadata is challenging
  • A large amount of duplication in the ecosystem

What DASCore gives you

Accelerates the science by providing a solid base:

  • IO support for many DFOS formats
  • Archive management
  • Common processing and visualization routines
  • Support for units
  • A description of the observing system
  • Extensible plugin architecture

Patch and Spool

A Patch holds data, coords and attrs, and is one entry in a Spool's contents and index; both expose their own methods.

Inventory

The inventory's two branches: point sensors on the left, fiber on the right, meeting at a shared network.

The inventory describes the observing system

  • What fiber was measured, and where it was
  • Which interrogator settings produced which recordings
  • All of it a function of time, supporting epochs

Attaches to a spool for expressive querying and patch enrichment

Learning objectives

  1. Use Patch: filter, smooth, transform and plot (notebook 01)

  2. Manage an archive with Spool: subselect regions of interest, chunk, merge, reshape and visualize contents (notebook 02)

  3. Track a deployment with Inventory: attach metadata and enrich selection and patches (notebook 03)

Bonus (time permitting)

  1. The Need for Speed: A demonstration of using JAX to speed up DASCore.

  2. Novel or Nonsense: Evaluating research created by ChatGPT Sol and Claude Fable 5.

The dataset

A complex fiber-optic network at an underground hardrock mine

Open it on its own, or see the optical paths in optical_topo.pdf

Getting started


github.com/DASDAE/galileo_2026