No difference

How this verdict is computed

Every day we pull magnitude 6.0+ earthquakes since 1973 from the USGS global earthquake catalog. We deliberately ignore smaller quakes: detection networks have improved over the decades and vary by region, so small-magnitude counts are dominated by observation bias rather than actual seismicity.

Large earthquakes are followed by swarms of aftershocks. Counting them naively would let a single mainshock that happens to land near a full moon pile hundreds of events onto the "full moon" side. We therefore apply a simple declustering rule: events that occur close to a larger earlier event (same source region, within 30 days, smaller magnitude) are removed as aftershocks.

For each remaining earthquake we compute the time difference (in UTC) between the event and the nearest full moon instant. Events within ±24 hours of the full moon instant are compared against the count you would expect if earthquakes ignored the moon entirely. The same calculation is done for new moons. Whether the difference is statistically meaningful is evaluated internally and translated into the one-word verdict at the top of this page. See the methodology for the exact criteria.

How the Japan figures are counted

For Japan we use the Japan Meteorological Agency's seismic intensity database: every earthquake since 1996 whose maximum intensity (shindo) reached 3 or more, with aftershocks removed the same way as the worldwide data. We use intensity rather than magnitude because shaking is what people in Japan actually notice. The intensity scale was revised in 1996 (levels 5 and 6 split into Lower and Upper), so the verdict and yearly table start there.

Intensity 4+ alone leaves only about 1,100 quakes after removing aftershocks — too few within a day of a full moon to call it — so we widened the net to intensity 3+. That line was drawn before looking at the result, on the understanding that whatever came out would be published as is.

The list of major earthquakes includes every quake since 1919 with a maximum intensity of 6 Lower or more (6 or more before 1996), aftershocks included. Moon age is computed from each quake's origin time.

Research looking for the moon's hand in earthquakes

Whether tidal forces can trigger earthquakes is actually a serious research topic (tidal triggering). The gravitational pull of the moon and sun does exert a tiny stress on the crust, and some studies have reported small statistical correlations under specific conditions (shallow thrust faults where tidal stress aligns with fault motion, for example).

The effects reported so far are very small — not yet enough to say "full moon days are dangerous." Still, the moon's pull really does reach the ground beneath us. Our verdict answers one narrower question — are major earthquakes visibly more frequent on calendar full moon days? — and looks at a coarser scale than tidal research does.

What we can't see yet

This result doesn't mean "the moon has nothing to do with it." Here is what the current method can't see:

• Fault type and location: the hints in research come from specific settings, such as shallow thrust faults. Counting every quake worldwide together dilutes them
• The stress on a fault at that moment: the direction and size of the tidal push on a fault change with place and time. A single calendar point — "full moon" — can't capture that
• Smaller quakes: we leave out quakes below M6 to avoid detection bias, so their rise and fall isn't examined here

What we want to check next

• Split quakes by depth and fault type and run the same test
• Break the Japan intensity data down by region
• Compare only the spring tides that coincide with a close full moon (a "supermoon")

The moon's pull really does reach the ground. Whether that small force connects to earthquakes somewhere — that's what we hope to find in the data one day.

Sources

• Japan Meteorological Agency, Seismic Intensity Database — Japanese earthquakes (intensity 3+ since 1996; 6 Lower+ since 1919). Source: JMA website, processed by MoonAlibi
• USGS Earthquake Catalog — earthquake event data (public domain)
• Moon phase instants computed in-house from Jean Meeus, "Astronomical Algorithms" (UTC)