auraopal

🪨 AuraOpal

Offline-first field map for boulder-opal prospecting at the Opalton opal field, Central-West Queensland. Sibling of AuraGold — same offline-PWA DNA (Leaflet + service worker + bundled overlays + one-push OTA), built fresh for the opal domain: this is excavation geology, not metal detecting.

Live: https://auraopal.banksiaspringsfarm.com/

⚠️ v0.1 is a data-visualisation tool. It plots Steven’s Aug 2025 ground-truth finds against the geology, geophysics and mining-tenement layers that drive boulder-opal favourability. It does not predict where opal is — favourability scoring is deliberately deferred until the raw data layers are honest and complete.

What it shows

Anchored on Steven’s mining lease (the default view frames the ~15 ha claim).

• Base maps — Satellite (Esri, default, over-zoomed to z19 for on-claim detail), Topographic (OpenTopoMap), ⬜ Minimal (blank, offline).
• Ground truth (Aug 2025 field trip) — the claim polygon (~15 ha), 10 opal finds colour-coded by type (🟢 gem · 🔴 red · 🟤 potch · 🟠 patch/indeterminate) with depth, 3 fault markers, and 6 trench excavations.
• Terrain — 10 m contours (QLD SRTM, bundled offline — no LiDAR 1m/5m exists over remote Opalton) + Esri hillshade.
• Winton Formation — the boulder-opal host rock, from QLD state surface geology.
• Geophysics (raw signals, labelled honestly) — Radiometrics ternary (the moderate weathering signal, often more useful here than mag) + Magnetics TMI-RTP and 1VD (weak/coarse over Opalton — kept for structure). Clipped to a ~50 km buffer. None of these is a forecast.
• Mining tenements — QLD Mineral Tenement (Mining Leases / Claims / Exploration Permits; opal-mineral tenements highlighted). Doubles as the legality + historical-activity layer.
• 📋 Site info panel — stratigraphy column + Steven’s ground-truth patterns + honest notes on imagery/terrain/photo/geophysics limits.

Everything (contours, geology, tenements) works offline once loaded (service worker precaches the shell + bundled GeoJSON; map tiles and the GA geophysics WMS are cached on view).

For which active geophysical technique to actually run over the claim (GPR vs ERT vs TDEM, with AUD costs), see research/opal_favourability_data_sources.md → “Active exploration techniques”. Short version: GPR is the top pick for typical uncapped-sandstone dig areas (the basal clay sits under the ironstone target, so radar images the target and only dies beneath it); ERT is the fallback where a capping clay is present or suspected.

Structure

www/            app source — index.html, sw.js, manifest.json
data/           bundled overlays — winton_formation.geojson, tenements_opalton.geojson
icons/          PWA icons (opal-gem)
research/        opal_favourability_data_sources.md — the full QLD data landscape + verdicts
tools/          build_site.sh + data-refetch notes
.github/workflows/deploy.yml   GitHub Pages (Actions source)
qa/             verification screenshots
CLAUDE.md       project context (read this first)
NIGHT_LOG.md    long-form build log

Run locally

bash tools/build_site.sh          # assemble www + data + icons -> _site/
cd _site && python3 -m http.server 8099
# open http://localhost:8099/

Deploy (OTA)

Push to main. .github/workflows/deploy.yml rebuilds _site and deploys to GitHub Pages. Bump APP_VERSION (www/index.html) and SHELL_REV (www/sw.js) in lockstep on every deploy.

Data sources & licences

All layers cite a real, verified source (CC-BY 4.0 for the government data). See the in-app About & data sources panel and research/opal_favourability_data_sources.md. Ground-truth finds are Steven’s own field data (Google Earth KML, Aug 2025).

Advisories

• Legality shown is advisory, derived from public tenement data. Always verify tenure and permission with the Queensland Government before entering or digging on any ground.
• Aboriginal cultural heritage applies everywhere in Queensland and is not shown here (the data is legally restricted). Check the Cultural Heritage Duty of Care Guidelines separately.