An interactive 3D globe that lets you pick where Earth's North Pole moves to and shows the consequences of that crustal-displacement / pole-shift scenario — ocean redistribution and new coastlines, earthquakes, volcanism, tsunamis, glacial rebound, climate-belt shifts, and more.
Its distinguishing feature is honesty: every effect layer is graded by scientific standing — established, contested, or speculative — shown right on the control and in a "Model Caveats" panel, with named sources. It runs the fringe pole-shift scenarios (Hapgood, ECDO, Zacharias/GEOSYNC) faithfully and labels exactly where each one departs from mainstream geophysics. It is a "what-if" explorer, not a forecast.
It's a single self-contained page plus a few data files — no build step. Serve the folder over HTTP and open it:
python -m http.server 8000
# then open http://localhost:8000/ (or /geo.html)(Opening geo.html directly via file:// won't work — the browser blocks it from loading the
sibling terrain/texture files. Any static host, including the served folder above or GitHub
Pages, is fine.)
geo.html— the application. It is self-describing: a long header comment records the full build status, locked design decisions, physics constants with provenance, and known landmines. Read that header before editing.index.html— a thin redirect so the site root opens the app.earth_terrainrgb.png,plate_boundaries.json,earth_day.jpg,earth_night.jpg,earth_specular_clouds.jpg— the runtime data (real ETOPO-derived terrain, plate boundaries, and photoreal Earth textures).research/— the five research reports and premise audit behind the model, plus the full datedCHANGELOG.md.build/— data-generator scripts and their raw sources (not needed to run the app; not deployed).
Three.js + WebGL2, one custom ShaderMaterial, lil-gui — vanilla, no framework, no bundler.
Three.js loads from a pinned CDN via an importmap.
The rapid, large pole shifts this tool explores are not accepted mainstream science; the underlying real-time math is a deliberate simplification of the rigorous (non-real-time) tools it approximates (e.g. LIOUSHELL for true polar wander, SELEN4/giapy for glacial-isostatic sea level). Outputs are scenario exploration for understanding, not prediction.