A magnitude 6.9 earthquake struck near Russia’s Kamchatka Peninsula at 11:35 p.m. EDT Thursday, Oct. 15, according to the U.S. Geological Survey (USGS), which places it 76 km (about 47 miles) northeast of Ozernovskiy. About four and a half minutes later, at 12:40 p.m. Friday Japan time, the Japan Meteorological Agency (JMA) issued an emergency earthquake warning for a magnitude 9.9 quake that it had placed in the Kuril Islands. The warning was withdrawn about eight minutes later, according to BNO News, citing the broadcaster NHK. In a press release, JMA said the warning came from its earthquake-warning processing, which fixed the epicenter in the Kuril Islands and overestimated the magnitude as 9.9, and that it is still investigating why. USGS’s automatic PAGER model rates the real event green for shaking-related fatalities and economic losses.
This account is current as of about 6:45 a.m. EDT (10:45 UTC) Friday, Oct. 16. The USGS record was last updated at 06:13 UTC and was unchanged when read again at 10:45 UTC. JMA’s press release is dated 2 p.m. Japan time (05:00 UTC). The BNO News page was published at 04:10 UTC and last modified at 04:32 UTC. The Watchers page was published at 04:41 UTC, the PTI page is dated 10:45 India time (05:15 UTC), the Asia Business Daily page is dated 14:30 Korea time (05:30 UTC), the newscord.org page is dated 06:05 UTC, and the Türkiye Today page is dated 09:32 Turkey time (06:32 UTC). The origin and magnitude are marked “reviewed” by USGS, while the PAGER and ShakeMap estimates are automatic and may change.
What happened and when
The USGS record lists the origin time as 03:35:31 UTC on Oct. 16. That is 11:35 p.m. EDT on Thursday, 3:35 p.m. Friday in Kamchatka (UTC+12) and 12:35 p.m. Friday in Japan.
USGS places the epicenter at 51.98 degrees north, 157.27 degrees east, which it describes as 76 km northeast of Ozernovskiy, Russia. The magnitude is 6.9.
JMA’s press release describes the same earthquake as a magnitude 6.9 (preliminary value) near the Kamchatka Peninsula at around 12:38 p.m. Japan time, and JMA’s warning-history page lists 12:38:10 p.m. as the time its system first detected seismic waves. USGS’s origin time, 12:35:31 p.m. Japan time, is earlier. The pages do not explain the difference.
Russia’s Geophysical Service, the Kamchatka branch of the Russian Academy of Sciences, first estimated 6.5 and then 6.9, according to a report from India’s PTI news agency that cited the Russian news agency TASS. The same report said the Russian service put the epicenter on land, 156 km southwest of Petropavlovsk-Kamchatsky. That is a different reference point from the one USGS uses, so the two distances are not directly comparable.
How deep it was
USGS lists a depth of 135 km, about 84 miles. Early reports gave other figures. According to PTI, the initial USGS figure was 143.2 km; the raw USGS record does not contain that number. The Russian service’s figure was 138 km, according to PTI citing TASS, while another aggregator page (newscord.org) gave 127.8 km for the same service. According to Watchers, the European-Mediterranean Seismological Centre said 130 km. JMA’s own account, as relayed by Asia Business Daily, put the real earthquake at about 160 km deep.
USGS’s automated tectonic classification gives a probability of 1.0 that the event lies within the subducting slab and 0.0 that it lies on the plate interface.
Depth matters for shaking. A 2016 Associated Press explainer, carried by Fox News, said deeper earthquakes lose energy on the longer path to the surface, so they are generally less damaging for their magnitude, though they are often felt over a wider area. That is a general principle, not a finding about this event, and it is the AP’s explanation, not a USGS statement.
What the models estimate
USGS’s PAGER system estimates the likely human impact of an earthquake. Its figures are modeled, not measured, and its review status for this event is “automatic.” PAGER issued a green alert. In its words: “Green alert for shaking-related fatalities and economic losses. There is a low likelihood of casualties and damage.”
Intensity here refers to the Modified Mercalli scale, a Roman-numeral scale of how strongly the ground shakes at a given place. According to the PAGER model, about 213,660 people lived where shaking was modeled at intensity IV and about 13,999 where it was modeled at intensity V. Four people were in the zone modeled at intensity VI. None were modeled at a higher level.
PAGER also gave probabilities for deaths. It put the chance of zero or one death at 90.4%, one to 10 deaths at 9.1%, and 10 to 100 deaths at 0.44%. These are model estimates, not reports of casualties.
On the building stock, PAGER says: “Overall, the population in this region resides in structures that are resistant to earthquake shaking, though vulnerable structures exist. The predominant vulnerable building types are adobe block and unreinforced brick with mud construction.”
USGS’s ShakeMap, an automatic version that used no station readings or reported intensities, modeled a maximum intensity of 5.6 and a peak ground acceleration of 0.156 g. PAGER’s city list, which gives a population of about 181,000 for Petropavlovsk-Kamchatsky, estimates intensity 4.0 there.
Reports from the area
According to the PTI report, which cited TASS, the Russian service measured instrumental intensity in Petropavlovsk-Kamchatsky of up to 4 points and reported shaking felt as 5 in the region. The report does not name the scale, so those numbers should not be assumed to match the Mercalli figures above. The same report said eyewitnesses described about a minute of shaking, with some people leaving their apartments, light fixtures swinging and doors slamming.
USGS’s “Did You Feel It?” page had received one response, which gave an intensity of 5.3.
Anadolu’s page (its time is not known) said there were no immediate reports of casualties or damage, and it relayed that a US tsunami warning system did not expect a tsunami. That is a summary of Anadolu’s page, not a check against the warning system itself.
An aggregator site (newscord.org, 06:05 UTC) reported that Russia’s Ministry of Emergency Situations said tremors were felt in Petropavlovsk-Kamchatsky, Ust-Kamchatsky, Aleutsky, Vilyuchinsk and Yelizovsky. No ministry or regional government page was opened, so that could not be checked against a ministry statement.
On tsunami risk, Watchers (published 04:41 UTC) reported that the Pacific Tsunami Warning Center said at 03:43 UTC there was no tsunami threat. The warning center’s own message was not opened, so this is reported but not independently confirmed. The USGS record’s tsunami field is 0, which is a flag in the record, not a statement from a warning center.
Two pages addressed injuries or damage in Kamchatka: the Anadolu page above, and the newscord.org page (06:05 UTC), which listed Global Times as reporting no injuries or infrastructure damage. The other pages reviewed (BNO News, Watchers, PTI, Asia Business Daily and Türkiye Today) did not mention Kamchatka injuries or damage.
The Japan alert
JMA issued the emergency earthquake warning at 12:40 p.m. Friday Japan time for the area from Hokkaido to the Chubu region, according to its press release. JMA’s warning-history page lists two warning reports, at 12:40:02 p.m. and 12:40:41 p.m. Both described a magnitude 9.9 earthquake at a depth of 140 km at 48.2 degrees north, 153.4 degrees east, in the Kuril Islands region. They forecast shaking of seismic intensity 5-lower to 7 on JMA’s scale in parts of eastern Hokkaido, with lower intensities farther south; the second report listed intensity 1 to 5-lower for Tokyo’s 23 wards. The same page lists 12:36 p.m. as the earthquake time of that mistaken event, not the time of the warning.
BNO News, citing NHK, reported the alert at about 12:36 p.m. and the cancellation at 12:48 p.m., about 12 minutes. JMA’s records put the warning at 12:40 p.m., which makes the alert about eight minutes long, and the 12:36 figure matches the earthquake time JMA’s system assigned to the mistaken event. Fuji News Network, as relayed by Türkiye Today and Asia Business Daily, gave 12:40 p.m. for the alert and a withdrawal eight minutes later. JMA’s own press release does not give the withdrawal time.
JMA’s press release says the warning was issued in connection with this earthquake. It says the warning resulted from the agency’s earthquake-warning processing, which fixed the epicenter in the Kuril Islands and overestimated the magnitude as 9.9. The release says the cause of the overestimate is under investigation and that JMA will report it as soon as it is known. At a 2 p.m. news conference, an agency official apologized, according to Asia Business Daily. That outlet reported that an official said there were no current concerns about a cyberattack or similar incident, and Türkiye Today said officials ruled out a cyberattack.
On the effects in Japan, JMA’s release says the earthquake produced seismic intensity 1 at Hakodate, Hokkaido. NHK, as relayed by BNO News, said the earthquake produced no shaking in Japan and that no tsunami warning was issued. Asia Business Daily also reported, citing JR Central, that the Tokaido Shinkansen’s earthquake-prevention system activated automatically and suspended service on some sections, and that service resumed at around 12:55 p.m.
Why 6.9 versus 9.9 is a big gap
Magnitude is a logarithmic scale. Each whole step up releases about 31.6 times more energy (10 to the power 1.5), so a three-step gap is about 31,600 times. In a paragraph that cites the USGS, BNO News said a 9.9 would be the largest earthquake ever recorded, above the 9.5 Chile earthquake of 1960, and would release roughly 32,000 times as much energy as the 6.9. That figure is consistent with the arithmetic.
Background: the 2025 earthquake
On July 29, 2025, a magnitude 8.8 earthquake struck offshore of Kamchatka. The USGS record describes shallow reverse faulting on the plate interface, with a listed depth of 35 km (an operator-assigned value; moment-tensor solutions in the same record put it nearer 21 km), and calls it the largest earthquake worldwide since the 2011 Tohoku quake in Japan. USGS reported “Four people injured in Kamchatskiy Kray,” at least 1,400 homes damaged in the Petropavlovsk-Kamchatsky area, at least 106 buildings damaged in Severo-Kuril’sk, and tsunami wave heights up to 5 meters at the Vodopadnaya meteorological station.
In that region the Pacific plate moves west-northwest at about 80 mm a year relative to North America. USGS counted nearly 700 earthquakes of magnitude 5.0 or larger offshore of the peninsula since 1990, not counting the 2025 event and its aftershocks.
Thursday’s earthquake was much deeper than the 2025 rupture. None of the sources consulted says whether Thursday’s earthquake is linked to the 2025 event.
What is known and unknown
Known from the USGS record: the earthquake’s parameters, the green PAGER alert, and the record’s tsunami field of 0. Known from JMA’s own press release: the warning was issued at 12:40 p.m., it described a magnitude 9.9 earthquake placed in the Kuril Islands, JMA says that was an overestimate, the cause is under investigation, and Hakodate recorded intensity 1.
Reported by secondary pages: the withdrawal at 12:48 p.m. (BNO News citing NHK) and eight minutes after the alert (Fuji News Network, via Türkiye Today and Asia Business Daily); the apology, the cyberattack comments and the Shinkansen pause (Asia Business Daily and Türkiye Today).
Reported but not independently confirmed: the Pacific Tsunami Warning Center’s no-threat statement (via Watchers), the Russian service’s intensity figures and the eyewitness accounts (via PTI citing TASS), and the Ministry of Emergency Situations’ list of towns where tremors were felt (via an aggregator).
Not known: why JMA’s processing overestimated the magnitude and fixed the epicenter in the Kuril Islands; whether Russian authorities have confirmed any injuries or damage; how widely the shaking was felt beyond one “Did You Feel It?” response; and how aftershocks will behave.
USGS earlier recorded five earthquakes of magnitude 4.5 to 5.2 within 400 km since Oct. 10. The largest was a 5.2 on Oct. 11, and the most recent, a 4.9, came about 12.6 hours before the 6.9. None of the sources reviewed calls any of them a foreshock. A USGS query run at 10:45 UTC lists no other earthquake of magnitude 2.5 or larger within 400 km since the main shock, and the event record lists no aftershock-forecast product.
What comes next
JMA said it will report the cause of the overestimate as soon as it is known. USGS may update its record, and the Russian service may revise its estimates.