Shake3D Research · Benchmark

Unofficial research results · our automatic detections vs. GeoNet’s published reports

Unofficial. These are research results from Shake3D, a volunteer-run monitor. It is not affiliated with or endorsed by GeoNet, NEMA or Civil Defence; GeoNet’s catalogue is the official record. Results refresh in the background and can be several minutes old.

🔬 Transparent Multi-Station & Detector Verification
Recorded source solutions and fusion decisions compared with the current GeoNet felt-event feed (MMI ≥3). Network readings are shown as unassociated motion, not proof that an earthquake was detected. Historical demonstrations are excluded; only recorded results appear here.
Measurement rule: A station reading is not an alert. Fusion confirmation, shadow forecasts, and delivery outcomes are reported separately. Bulletin advantage compares recorded confirmation with this monitor’s first GeoNet receipt; it is not pre-shaking warning time.
Sources Located --% Evaluating catalog...
Events with Complete Coverage -- 120 s of recorded detector readiness
Median Bulletin Advantage -- Signed time ahead of first GeoNet receipt
Median Epicenter Offset -- First accepted source vs catalog
Median Magnitude Delta -- Median absolute error of first estimate

Event maps load only when you open an event. Their base map comes from Esri's ArcGIS Online servers, so Esri receives your IP address and the map area you view. Basemap © Esri, HERE, Garmin, OpenStreetMap contributors and the GIS user community.

Loading GeoNet catalog and detector correlations...