Why Japan’s earthquake early warning said “M9.9” — and what a real one would mean
At midday on October 16, the Japan Meteorological Agency issued an Earthquake Early Warning for Hokkaido through the Chubu region saying “M9.9 near the Kuril Islands, maximum seismic intensity around 7,” then canceled it eight minutes later. Posts on X claiming “it was a cyberattack” or “a failed artificial earthquake” have also spread. We check what happened and, based on materials from the Japan Meteorological Agency and the Cabinet Office, explain what would happen if an M9.9 earthquake really occurred.
Verdict: are the viral claims true?
- False“A massive M9.9 earthquake struck near the Kuril Islands”The earthquake that actually occurred was an M6.9 near the Kamchatka Peninsula, and the maximum seismic intensity in Japan was 1, in Hakodate, Hokkaido. The Japan Meteorological Agency overestimated its size and canceled the warning.
- Unverified“The Japan Meteorological Agency’s system was hit by a cyberattack”The agency denies a cyberattack, saying there was no outside access during that time. In its second report it cited causes such as applying a formula designed for domestic earthquakes to a distant one.
- False“It was not a false alarm but a failed attempt to trigger a massive artificial earthquake”What actually occurred at the same time was an M6.9 near the Kamchatka Peninsula, which the agency also announced in its earthquake information. The Seismological Society of Japan says “artificial earthquakes” are unrealistic, given the enormous cost and energy they would require.
- True“This was the first time an Earthquake Early Warning (warning level) has been canceled”The agency explains that “forecasts” have been canceled before, but this was the first cancellation of a public “warning.”
Translated and adapted from our Japanese original. Facts are as of the original publication date unless noted.
Earthquake early warnings: speed first, or accuracy first?
- Warning’s estimate (actually M6.9)
- M9.9
- From warning to cancellation
- About 8 minutes
- Energy of M9.9 (vs. M9.0 Great East Japan Earthquake)
- About 22 times
Speed first
There are only seconds to tens of seconds before shaking. Even with false alarms, warning quickly should come first.
Review the system
Repeated false warnings erode trust. The handling of distant quakes and how cancellations are communicated should be reviewed.
- Both sides agree
- The cause must be investigated so the same error is not repeated.
- What to watch next
- The results of the agency’s investigation into the cause, and measures to prevent a recurrence.
Key points
- At 12:40 p.m. on October 16, the Japan Meteorological Agency issued an Earthquake Early Warning for Hokkaido through the Chubu region, citing “M9.9 near the Kuril Islands,” and forecast a maximum seismic intensity of around 7 in Hokkaido and Tohoku. It canceled the warning eight minutes later, at 12:48.
- The actual quake was an M6.9 near the Kamchatka Peninsula at a depth of about 160 kilometers, and the maximum seismic intensity in Japan was 1, in Hakodate.
- In its second report, the agency cited as causes that it applied a magnitude formula designed for domestic earthquakes to a quake very far from Japan, and that it used different hypocenters for calculating the magnitude and for forecasting seismic intensity.
- The Tokaido Shinkansen suspended service between Shin-Yokohama and Toyohashi for about 15 minutes, and some JR Hokkaido conventional lines were also temporarily halted.
1. What is happening
Sorting out the viral claims
| Viral claim | What we could confirm |
|---|---|
| An M9.9 earthquake struck near the Kuril Islands | It was actually an M6.9 near the Kamchatka Peninsula, at a depth of about 160 km. The maximum seismic intensity in Japan was 1, in Hakodate. |
| It was a cyberattack or hacking | The agency denies it, saying there was no outside access. Its second report points to problems in the calculation system. |
| It was a failed attempt to trigger an artificial earthquake | What actually occurred was a natural M6.9 earthquake near the Kamchatka Peninsula. The Seismological Society of Japan calls the “artificial earthquake” theory unrealistic. |
| M9.9 is 16 times the Great East Japan Earthquake | It depends on the comparison. Taking the Great East Japan Earthquake as M9.0 gives about 22 times; taking it as M9.1 gives about 16 times. |
What happened
- Around 12:38 p.m.: An M6.9 earthquake occurs near the Kamchatka Peninsula
- 12:40 p.m.: The agency issues an Earthquake Early Warning for Hokkaido through the Chubu region, citing “M9.9 near the Kuril Islands,” and forecasts a maximum intensity of around 7 in Hokkaido and Tohoku
- Around 12:40 p.m.: The Tokaido Shinkansen suspends service between Shin-Yokohama and Toyohashi (resumes around 12:55)
- 12:48 p.m.: The agency cancels the warning. Some devices sounded the alarm for the cancellation notice too
- 2 p.m.: The agency holds a press conference and apologizes: “It was not a cyberattack”
- 5:30 p.m.: Second report, disclosing causes such as applying a domestic formula to a distant quake
If an M9.9 earthquake had really occurred
It would be larger than anything ever recorded. The largest earthquake recorded worldwide is the 1960 Chile earthquake (M9.5), and the largest recorded in Japan is the 2011 Great East Japan Earthquake (M9.0).
About 22 times the energy of the Great East Japan Earthquake. According to the agency, each increase of 1 in magnitude multiplies an earthquake’s energy by about 32, and an increase of 2 by about 1,000. Calculated this way, M9.9 has about 22 times the energy of M9.0 and about 4 times that of the Chile earthquake (M9.5).
The fault that slips would also be vastly larger. Each increase of 1 in magnitude makes the fault roughly 3 times longer. An M9.9 would require a fault more than twice as long as that of an M9 quake to slip all at once, and at its press conference the agency also called it “seismologically implausible.”
A useful reference is the government’s scenario for a massive Kuril Trench earthquake. The Cabinet Office projects the following if a maximum-class earthquake occurred along the Kuril Trench east of Cape Erimo. An M9.9 would be far larger even than this.
| Item | Maximum-class scenario along the Kuril Trench (Cabinet Office) |
|---|---|
| Shaking | Seismic intensity 7 in Erimo and Akkeshi, Hokkaido |
| Tsunami | Up to 28 meters high in Erimo and Kushiro town, Hokkaido |
| Deaths (winter, late night) | About 100,000. About 22,000 people who escape the tsunami would be at risk of hypothermia |
| Building damage (winter, evening) | About 84,000 buildings |
| Economic damage | About 17 trillion yen |
The scenario for the neighboring Japan Trench puts deaths at up to about 199,000 and economic damage at about 31 trillion yen.
When an Earthquake Early Warning sounds
The agency’s basic advice on seeing or hearing a warning is to “stay calm and first secure your own safety, according to your surroundings.”
- Indoors, protect your head and take shelter in a safe place, such as under a sturdy desk
- When driving, turn on your hazard lights to alert others, and if you feel strong shaking, pull over to the left side of the road (Japan drives on the left)
- Memorize the alert tone so you can protect yourself before checking the details
Sources (mostly in Japanese)
2. What each side says
We summarize each side's main arguments in our own words, based on the original statements and reporting.
Speed first
The Earthquake Early Warning does not predict earthquakes; it detects the first waves (P waves) of a quake that has occurred and warns before the strong shaking (S waves) arrives. There are only seconds to tens of seconds between the warning and strong shaking, so waiting for more data would mean the warning comes too late.
The agency says that, regarding the same sound playing for the cancellation, it is not currently considering changing the sound. The agency routinely urges people to memorize the alert tone, saying that if they know it they can protect themselves instantly before checking the details.
According to the agency, a warning is not canceled simply because the forecast is lowered afterward; it is canceled in cases such as when vibrations other than an earthquake were judged to be one. This time, it says it confirmed that there was no shaking when the seismic waves should have arrived, and canceled the warning once it judged the forecast clearly excessive.
Review the system
Because the cancellation notice played the same alarm as the original warning, many on X asked “why is it the same sound?” or said “it’s bad for my heart.” Some point out that a new strong-shaking alert and a cancellation cannot be told apart by sound, and that a different tone should be used.
The agency’s observation network extends north only to around Nemuro, so estimates for distant quakes beyond the Kuril Islands tend to be less accurate. In July 2020, an M5.8 near Torishima was also wrongly estimated as an M7.3 off the Boso Peninsula, and some are calling for the handling of distant quakes to be reviewed.
The largest earthquake ever recorded was the 1960 Chile earthquake (M9.5), and at the press conference M9.9 was described as seismologically implausible. Some argue there should be a mechanism to stop such “impossible numbers” before they are issued automatically.
Posts on X
Japan Meteorological Agency: Released its second report on the cause of the overestimate in the Earthquake Early Warning.
View the post by Japan Meteorological Agency on X ↗
Reading between the lines
This section is our editors' interpretation (speculation). It is not confirmed, but the facts it is based on are sourced.
Japan Meteorological Agency
Why did it forecast “M9.9, intensity 7”?
SpeculationAccording to the diagram in the second report, the agency’s system calculated the magnitude using a hypocenter in the ocean even farther east than the Kamchatka Peninsula, then used that value to forecast seismic intensity across Japan with a hypocenter near the Kuril Islands, closer to Japan. Applying a magnitude overestimated from a distant hypocenter to a nearby one likely produced the forecast of intensity around 7 in Hokkaido and Tohoku. The detailed mechanism is under investigation.
Facts this is based on
- • A magnitude formula designed for domestic earthquakes was applied to a quake very far from Japan.Japan Meteorological Agency (Earthquake Early Warning, second report, Japanese)(Oct 16)
- • Different hypocenters were used for calculating the magnitude and forecasting seismic intensity.Japan Meteorological Agency (Earthquake Early Warning, second report, Japanese)(Oct 16)
What people are saying
We searched Japanese posts on X (formerly Twitter) for terms such as “M9.9” and “Earthquake Early Warning cyberattack” on October 16, 2026. Alongside relief that “it was a false alarm,” posts suspecting hacking and complaints about the cancellation sound stood out.
Speed first
- “Thank goodness it was a false alarm. Even a false alarm is better than no alarm.”
- “It also matters that people can realize, the moment they see “M9.9,” that it might be a system error.”
Review the system
- “When the cancellation plays the same sound, you think it’s another earthquake, and it’s even scarier.”
- “There has been so much unauthorized access lately that I can’t help suspecting a cyberattack.”
Posts are paraphrased, not quoted, and individual accounts are not named.
3. Where things stand
Where they agree
- No M9.9 earthquake occurred near the Kuril Islands; it was actually an M6.9 near the Kamchatka Peninsula.
- The cause should be investigated so the same error is not repeated.
Where they disagree
- Whether warnings should keep prioritizing speed even at the risk of false alarms.
- Whether the sound and wording of cancellation notices should change.
What to watch
- The final results of the agency’s investigation into the cause.
- Revisions to the calculation system, such as the handling of distant quakes.
- Whether the way cancellations are communicated is revised.
No M9.9 earthquake occurred near the Kuril Islands; it was actually an M6.9 near the Kamchatka Peninsula. The Japan Meteorological Agency denies a cyberattack and cites causes such as applying a formula designed for domestic quakes to a distant one. Theories such as “a failed artificial earthquake” have no basis. Those who put speed first argue that, with so little time before shaking, warnings should be issued quickly even if some are false alarms. Those who want a review argue that, to maintain trust, the handling of distant quakes and the way cancellations are communicated should be improved.