AusMT

AusMT is a survey-centric magnetotelluric (MT) data infrastructure for Australia. It stores curated survey packages of transfer functions with their metadata, provenance and citation records, and serves them through a public portal.

The survey is the primary object: a survey package combines transfer functions, metadata, provenance and citation information into one curated, versioned unit, rather than treating each file as an independent artifact.

It is built for researchers, students, survey custodians, data managers, research infrastructure operators, government agencies, and the archive's future maintainers.


Background

Hundreds of magnetotelluric surveys have been acquired across Australia by universities, government agencies, research infrastructure programs and industry. They support research into lithospheric architecture, crustal evolution, mineral systems, groundwater systems, geothermal resources, natural hydrogen systems and tectonic processes. Many of the resulting datasets remain scientifically valuable but sit in personal archives, institutional storage, project websites, publication supplements or legacy media.

Reuse then fails in recognisable ways:

  • Transfer functions survive while their metadata are lost.
  • Metadata survive while the data become hard to locate.
  • Processing workflows are undocumented, or the software no longer exists.
  • Several versions of a dataset circulate with no authoritative source.
  • Survey information is scattered across reports, publications and personal archives.
  • Publication practice differs between organisations.

AusMT provides one consistent framework for survey discovery, transfer-function access, metadata preservation, provenance tracking and long-term stewardship of these datasets.

Why the survey, not the file

A folder of EDI files rarely says enough. A user also needs acquisition dates, instrumentation, processing history, provenance, publications and citation information, and all of that belongs to the survey rather than to any one file. Researchers ask for the AusLAMP South Australia data or the Vulcan dataset, not for a station file. So the survey package is what AusMT publishes, versions, cites and serves, and identifiers attach to it.


Scope

AusMT curates and serves:

  • Magnetotelluric transfer functions (EDI and EMTF XML, plus a per-survey transfer-function MTH5 bundle behind a deployment flag that ships on)
  • Survey and station metadata
  • Provenance records
  • Citation information
  • Derived screening diagnostics (apparent resistivity and phase, tipper, phase-tensor parameters, dimensionality)

Further derived products (strike analysis, distortion and decomposition) are planned and are marked as such wherever they appear in this documentation.

Out of scope

AusMT does not archive raw data. Time series remain in their original repositories (national facilities such as NCI, institutional and project archives). Where a survey's time-series collection has a persistent identifier, the survey package records it, so the portal links to the time series without duplicating them. How those links are recorded is in External archives.


Design principles

  • Survey first. Stations, transfer functions and derived products exist within a survey package; identifiers, versions and citations attach to the survey.
  • Reproducible. Every published value traces to a source file, a content hash, a unique identifier and a build provenance record.
  • Interoperable. Community standards are used throughout: mt_metadata and MTH5 for parsing and storage, EDI and EMTF XML for exchange.
  • Curated. Publication happens through validation and human review, not unrestricted upload.
  • Attributable. Metadata, provenance and citation information are first-class products, and data licensing is declared per survey by its custodians.

System architecture

The framework is the ausmt repository; survey data lives in the separate ausmt-surveys repository.

submissions -> gateway -> ausmt-surveys -> engine -> portal
               (scan,      (curated        (offline   (static
               validate,    packages)       build)     site)
               curate)
  • gateway — the submission service: upload, antivirus scan, validation, curator review, and publication as a git commit to the data repository.
  • ausmt-surveys — the curated collection of published survey packages: metadata, transfer functions and provenance.
  • engine — the offline build: parses packages with mt_metadata, computes the screening diagnostics, and writes the portal's data products, canonical EMTF XML and download bundles.
  • portal — the public discovery and access interface. It consumes generated products and performs no scientific processing.

Architecture covers the components in full; Developer architecture is the maintainer's view.


Where to start

Three paths through this documentation.

Using the data. How AusMT serves data, then the data reference for every served document, then tool integration for reading the artifacts in MT software. Four hops from here to a working fetch. The Reference section documents every served document field by field.

Contributing a survey. Data lifecycle for the shape of the journey, Submission for how to submit and what the validator checks, and the survey.yaml reference for every field.

Working on the code. Developer architecture, then Build lifecycle, then Portal data files for the positional contract.

For the reasoning rather than the mechanism, read Scientific philosophy and the design rationale pages.