QuakeWave
Colour = depthsurface70300700 km
Drag to orbit and tilt (tilt under the rim to see depth stalks) · pinch or scroll to zoom · two fingers (or right-drag) to pan · tap a quake for details. Azimuthal equidistant map centred on the North Pole: distances from the pole are true, the southern hemisphere is stretched, and Antarctica is the rim.
Live
Largest in this window
Magnitude over time
Each dot is one quake in the chosen window. Colour shows depth, as on the disc: shallow ~70 km ~300 km deep (700 km).
Recent earthquakes
| Place | Mag | Depth km | Time ET | Flags |
|---|
History
Where earthquakes cluster, from the USGS ComCat catalogue. Brighter cells hold more quakes (log scale).
Heat map: M5.0+ around Japan since 1973 (7,609 events), 1° cells.
Japan uses M5.0+ because the USGS API caps a query at 20,000 events and Japan has more than 28,000 M4.5+ quakes since 1973. Regional views zoom the disc surface, so the map in those views is magnified and the rim is not Antarctica.
Patterns
Not a prediction. A chain is a past M6+ quake followed, within the span, by M6+ quakes in other regions at least 1,500 km away (aftershocks removed). Most are coincidences: about a dozen M6+ quakes happen somewhere every month.
Lift compares how often a big quake in one region was followed by one in another within the span, across the whole 1976–today USGS catalogue, with the same count in shuffled data: quake times are shuffled among all M6+ mainshocks 200 times, keeping every location. 1.0× is what chance alone gives; 2.0× means twice as often.
Significant: still holds after correcting for testing thousands of region pairs (Benjamini–Hochberg q < 0.05, 5+ hits). Suggestive: beats the shuffles (p < 0.05, 3+ hits) but could be a fluke of testing so many pairs. Likely chance: no better than shuffled data. A chain is only as strong as its weakest link.
Aftershocks are removed with Gardner–Knopoff windows. Regions are broad tectonic zones. Wave rings are sped up: real surface waves cross the globe in about 1.5 hours, while the chains unfold over days to weeks, far longer than any wave takes to arrive.
Open to find chains.
Real USGS catalogue events, played back at their actual times; the rings show surface waves spreading at about 3.7 km/s. Gray dots are just quakes that happened during the window: being on screen does not make them triggered.
What controls how far shaking reaches
Short version: magnitude sets how much energy there is, depth and fault type shape where it goes, and the rock it travels through decides how fast it fades. Passing waves from a great quake can set off small quakes thousands of kilometres away, but the evidence does not show them setting off large ones, with one debated exception.
Magnitude
- Each whole step in magnitude releases about 32 times more energy; bigger ruptures are longer and radiate more long-period waves that fade slowly, so felt area grows quickly with size (USGS, magnitude and energy).
- Intensity at a site still depends on distance, depth and local ground, so magnitude alone does not set the shaking anyone feels (USGS).
Depth
- USGS classes quakes as shallow (0–70 km), intermediate (70–300 km) and deep (300–700 km). The depth stalks under each quake on the disc show this.
- Shallow quakes put much of their energy into surface waves that run along the crust, giving the strongest shaking near the epicentre and the long-distance rumble.
- Deeper quakes send energy up from below as body waves: shaking at the epicentre is weaker, but it is spread more evenly and often felt over very wide areas (USGS depth FAQ; IRIS depth animation).
Faulting style and direction
- Every quake radiates in a four-lobed pattern set by its fault orientation, which is what the beach-ball diagrams show; strike-slip, thrust and normal faults point their strong lobes in different directions (IRIS; USGS).
- Rupture directivity: when a rupture runs toward a site, shaking there is stronger and shorter; behind it, weaker and longer (Somerville et al., 1997).
- At global distances the same effect steers triggering: Denali's rupture ran southeast, aiming strong waves at Yellowstone (Husen et al., 2004), and the 2012 Indian Ocean quake's remote aftershocks lined up with its Love-wave lobes (Pollitz et al., 2012).
Tectonic setting and crust
- Old, cold, healed crust carries waves farther. The 2011 M5.8 Virginia quake was felt up to about 600 miles away; the 2014 M6.0 Napa quake, with twice the energy, only about 250 miles (USGS, East vs West Coast Earthquakes).
- This is why a New Madrid quake would shake a far larger area than a similar California quake; soft sediments add local amplification on top (USGS).
Remote and dynamic triggering
- 1992 Landers M7.3: sudden bursts of quakes up to 1,250 km away, mostly at geothermal and volcanic sites such as The Geysers and Yellowstone (Hill et al., 1993).
- 2002 Denali M7.9: triggered swarms up to at least 3,660 km away, at Mount Rainier, The Geysers, Long Valley and Coso, starting during the surface waves (Prejean et al., 2004); Yellowstone had more than 250 quakes in 24 hours and stayed elevated about 30 days (Husen et al., 2004).
- 2011 Tohoku M9.1: small quakes and tremor triggered in the US, Russia, China, Ecuador, Mexico and elsewhere, with no overall rise in global seismicity (Gonzalez-Huizar et al., 2012).
- 2012 Indian Ocean M8.6: remote M5.5+ quakes ran at nearly five times the normal rate for six days, up to M7 (Pollitz et al., 2012).
Could a big Japan quake set off California, Alaska or New Madrid?
- Small quakes, briefly, yes: passing surface waves routinely trigger small events in places like The Geysers, Long Valley and Alaska's volcanic areas.
- Large quakes, no evidence: across every M7+ quake in 30 years, Parsons and Velasco (2011) found no link between wave arrivals and M5+ quakes beyond two to three rupture lengths of the mainshock. Regional hazard rises after a big quake; global hazard does not.
- 2012 is the one strong exception, from an unusual strike-slip event; for the great Japan thrust quake of 2011 the global rate did not rise. We found no documented case of New Madrid being triggered by a distant quake.
Sources
- East vs West Coast Earthquakes USGS
- Earthquake Magnitude, Energy Release, and Shaking Intensity USGS
- At what depth do earthquakes occur? USGS
- Effect of depth on intensity of ground shaking IRIS
- Focal mechanisms explained IRIS
- Rupture directivity in ground-motion relations (Somerville et al., 1997) Seismological Research Letters
- Seismicity remotely triggered by the M7.3 Landers earthquake (Hill et al., 1993) Science
- Remotely triggered seismicity on the US West Coast after Denali (Prejean et al., 2004) BSSA
- Remotely triggered seismicity at Yellowstone by Denali (Husen et al., 2004) BSSA
- Remote triggered seismicity caused by the 2011 Tohoku-Oki earthquake (Gonzalez-Huizar et al., 2012) GRL
- The 2012 east Indian Ocean earthquake triggered large aftershocks worldwide (Pollitz et al., 2012) Nature
- Absence of remotely triggered large earthquakes beyond the mainshock region (Parsons & Velasco, 2011) Nature Geoscience
About this app
Recent quakes come live from the USGS GeoJSON feeds (past day all magnitudes, past week M2.5+, past month M4.5+), refreshed about every 60 seconds, with a bundled snapshot as a fallback. History and replays come from the USGS FDSN event catalogue (ANSS ComCat), and outlines from Natural Earth 110m land. Times are US Eastern. USGS can revise recent events as review completes. Nothing here is a forecast or prediction.
The disc is one WebGL canvas with one animation loop. The map surface is painted once into a texture and only repainted when data or a replay changes; pulses and glows are capped, pooled instances. Pixel density is capped, quality drops automatically when frames run slow, rendering stops on the Live and Learn tabs and when the app is hidden, and motion is reduced for reduced-motion settings.