Live from the USGS global catalog
It feels like there are more big earthquakes lately. This page checks that against the actual record — every recorded quake of magnitude 4.5 and above, worldwide, straight from the U.S. Geological Survey. No interpretation, no cherry-picked window. Change the filters and the charts recalculate in front of you.
Reading the catalog…
Loading global earthquake data
Fetching year-by-year from the USGS. Charts appear as each year lands.
Each bar is one year, split by how strong the quakes were. The dashed line is the average across the completed years shown. If earthquakes were genuinely increasing, the bars would climb from left to right. Watch whether they do.
The faint line is the raw monthly count — spiky, because one big quake sets off hundreds of aftershocks in the weeks after it. The bright line is the rolling 12-month average, which cancels that noise out. That bright line is the trend. A rising Earth would show a rising bright line.
Small quakes outnumber large ones roughly ten to one per step of magnitude, so on a shared axis the big ones vanish. Here each class gets its own scale, so a real rise in major quakes can't hide behind the volume of small ones. Each panel shows the rolling 12-month average.
These are the quakes that make the news. Tags come from the USGS PAGER system, which estimates likely damage and casualties. Click any row for the official event page.
Every figure behind the charts, so nothing is locked inside a tooltip.
The year in progress is not finished, so comparing its running total against completed years would always make it look low. By default this page compares 1 January → today against exactly that same calendar window in every earlier year. Switch "Compare" to whole calendar years if you'd rather see full totals, and the unfinished year is marked as partial.
The global seismic network has grown enormously. More stations mean more quakes detected — especially moderate ones in remote ocean regions that simply went unrecorded decades ago. So an upward slope over a multi-decade window can be measuring instrument coverage, not seismicity. This is why the page defaults to the last 10 years, where the network is broadly stable. If you widen the period, treat the early years as undercounts.
One magnitude 8 earthquake generates hundreds of aftershocks above magnitude 4.5 over the following months. A single great quake can therefore lift an entire year's count on its own. That's why a spike in the raw monthly line usually traces back to one event, and why the rolling 12-month average is the more honest read of whether anything is actually changing.
Significance (sig) and the tsunami flag are set for every event.
The PAGER alert level and shaking intensity are only computed for events USGS
considers significant — verified coverage is about 9% of magnitude 4.5+ events but
100% of magnitude 6+, which is where damage actually happens.
"Felt" counts come from public Did You Feel It? reports, so they reflect where
people live and have internet as much as where the ground shook. An absence of red
alerts at low magnitudes means "not assessed", not "no damage".
All figures are fetched live in your browser from the USGS FDSN event API, the public interface to the ANSS Comprehensive Catalog — the same catalog seismologists cite. Nothing is precomputed or stored on a server. Events flagged by USGS as quarry blasts, explosions or other non-earthquake sources are excluded. For the definitive reviewed record, the International Seismological Centre catalogue is the reference, though it runs roughly two years behind because it is human-reviewed.