M6.8 earthquake offshore Japan preceded by upward cascade of foreshocks

この投稿を日本語（Google による自動翻訳）で読むには、ここをクリックしてください。

Do you rely on these posts to stay up-to-date about earthquakes and earthquake science? To make this effort sustainable, we need financial support. If you can, please consider upgrading to a paid subscription. If you use our Substack as part of a group, or for teaching, consider a teaching or institutional subscription. Every paid subscription makes a real difference — and unlocks access to the more than three hundred posts in our archive.

Free subscribers still get all of our new posts by email, and all of our posts are free to read for one month. If you have recently lost your job or are a student or researcher without ability to pay, let us know and we will upgrade your subscription to “paid” at no cost.

On November 9th, 2025, at 5:03 PM local time, a magnitude 6.8 earthquake struck east of the island of Honshu, Japan. More information is available about this earthquake from the Japan Meteorological Agency (JMA) and the USGS. More information about how JMA monitors, prepares for, and responds to earthquakes can be found here.

People paying attention to the USGS earthquake feed will have noticed that the area that produced this earthquakes also hosted several earthquakes the day prior.

Foreshocks of the M6.8 earthquake actually started about a week ago, on November 3rd, with a M5.2 event. That earthquake was followed by a sequence of small aftershocks, which progressively decayed until, on November 8th, several larger earthquakes began to occur: first, four events in the M4.5-4.9 range, then a M5.3, a M5.4, and a M5.6, one after the other. The M6.8 struck about ten hours after the M5.6. In addition to this increase over time of the largest event magnitudes, which is called a ‘cascade up’, we also see a spreading out of the earthquakes over a larger area. This is presumably due to the larger magnitude earthquakes causing more widespread stress changes, thereby triggering more events over a larger area.

Figure 1: Foreshock-mainshock-aftershock sequence of the Novermber 9th, 2025 M6.8 earthquake. Earthquakes are colored by depth. They are also projected to the right onto a timeline, and shown below the map.

During our time writing at Earthquake Insights, we have seen other notable examples of ‘cascade up’ foreshock cascades. For instance, a M7.4 earthquake offshore Kamchatka on July 20th, 2025 was preceded by M5.0 and then M6.7 foreshocks. That earthquake then followed on July 29th by a great M8.8 earthquake.

Unfortunately, there is no way to know whether this cascade up sequence has already ended with the M6.8 event, or whether another larger earthquake might eventually occur, extending the sequence further.

The M6.8 earthquake was felt across much of the islands of Honshu and Hokkaido. The maps below show intensities reported to the Japan Meteorological Agency (JMA).

Because the earthquake occurred about 100 kilometers offshore, shaking on land was only mild to moderate, with the strongest shaking reached up to JMA intensity 4. As the cartoon below shows, this level of shaking may be frightening, but it should not cause much, if any, damage.

We note that the intensity scales used by the USGS (the Modified Mercalli Intensity Scale, or MMI) and JMA (the JMA Seismic Intensity Scale) are not the same. Where an MMI of I means not felt, a JMA intensity of 1 means perceptive to some people in the upper stories of buildings. JMA intensity 4 equals MMI ~V. In other words, the JMA scale is condensed and slightly shifted relative to the MMI scale.

Curiously, it is not the locations closest to the recent M6.8 earthquake that experienced the strongest shaking: Figure 2 shows that the strongest shaking was felt in a north-south line set back from the coast (intensity 3-4), stronger than the reports of 2-3 to the east. These stronger reports correlate to the Kitakami River Basin.

Local geological conditions can dramatically affect shaking — a process known as “site effects.” Sedimentary basins commonly increase shaking, because the seismic waves slow down and pile up in soft sedimentary rocks, amplifying in the process.

Was this kind of pattern expected? Yes. The maps below show estimated horizontal site amplification factors for Japan, for two different shaking frequencies. You can actually see the river valley, where stronger shaking was reported, in the right-hand panel.

Fig. 5 Figure 4: Estimated horizontal site amplification factors for Japan. Figure 5 of Kawase et al. (2023).

Large subduction earthquakes like this recent M6.8 are capable of raising damaging tsunami waves, because they lift both the seafloor and the water column above. Recognizing this hazard, JMA issued a tsunami warning within ten minutes of the earthquake. Because a tsunami arising from the earthquake location would take about half an hour to reach shore, this provided residents about twenty minutes of advance warning. The areas in the highest-risk zone were also experiencing high tide, increasing the potential hazard.

Before those twenty minutes had elapsed, the tsunami was detected at an offshore buoy: a “tiny” wave. The waves on land ultimately were also small: 0.2 meters or less. Notably, the highest waves were recorded about fifty minutes after the first sea level rise — a reminder that a tsunami can last for a long time, and have multiple waves. It is important not to return to an evacuated area until an all-clear has been issued.

One might think that the magnitude of an earthquake is a good indicator of how large the tsunami will be. In some sense, that is correct. Larger earthquakes usually produce larger tsunamis. However, there are well known examples of earthquakes that produce unusually large tsunamis for their magnitude. Further, submarine landslides caused by the shaking can also generate large tsunamis, without adding to the measured earthquake moment. This is why it’s always a good idea to treat all tsunami warnings seriously, even if you know that the earthquake magnitude is not very large.

The eastern side of Japan is underlain by great subduction zones: to the south, the Nankai-Ryukyu subduction zone, where the Philippine Sea Plate sinks northwestward beneath Kyushu, Shikoku, and southern Honshu, and to the north, the Japan-Kuril subduction zone, where the Pacific Plate sinks westward beneath northern Honshu and Hokkaido. The definitions of the overriding plate are a bit loose here in Japan: the upper plate crust contains many active fault systems, allowing for some internal deformation. Current plate models define three different plates that move slightly relative to each other, all of them sort-of-but-not-quite part of the Eurasian Plate. On the map below, the thick black arrows show the movement of the oceanic plates relative to the Amur Plate.

Subduction zones can produce the largest earthquakes on Earth, due to the huge area of the subduction fault that separates the two plates (the “megathrust”). All recorded earthquakes above M9 have occurred on these kinds of faults. Subduction zones are also sometimes called “trenches,” because of the deep valleys in the ocean floor where the plates meet.

Figure 6: Tectonic map of Japan, with cross-sections across the Nankai-Ryukyu Trench in the south and the Japan-Kuril Trench in the north.

We last wrote about the Nankai-Ryukyu subduction zone in August 2024, when a M7.1 earthquake raised the specter of a possible great earthquake — one of the most damaging seismic events that is expected to eventually occur. Because of the possibility of earthquake triggering, a megaquake advisory was issued, alerting the public to the elevated (but still small) risk of a large rupture. That was the first time that such an advisory had actually been raised.

This recent earthquake did not occur on the Nankai-Ryukyu subduction zone, but rather within the Japan-Kuril system to the northeast. That subduction zone is also prominent in Japanese seismic awareness, because it produced the catastrophic 2011 M9.1 Tōhoku-Oki earthquake. The 2011 earthquake ruptured a part of the megathrust that was ~400 kilometers long and 220 kilometers wide, with peak slip of up to 62 meters. The resulting tsunami inundated large areas of the coastline, with waves up to 40.5 meters high. Almost 20,000 people died as a result.

The recent M6.8 earthquake was caused by breaking of that same fault, not very far away from the earlier great rupture. The focal mechanism confirms that the earthquake occurred on a very low-angle, west-northwest dipping plane: the subduction megathrust.

Where did this earthquake occur in comparison to the Tōhoku-Oki rupture? Many slip models have been calculated for that great rupture, based mostly on seismic recordings, GPS measurements of ground motion, and measurements of the ensuing tsunami waves. These models are consistent in many ways, but also show some variability. Fortunately, a recent paper has compiled thirty-two of these models for us (Wong et al., 2024). On the plot below, we have placed the epicenter of the recent M6.8 (red stars) onto each compiled map.

Figure 8: Figure 2 of Wong et al., showing 32 different slip models for the Tōhoku-Oki earthquake. Red star added to show the location of the 2025-11-09 M6.8 earthquake.

Based on these maps, we can be pretty confident that this earthquake occurred near the northern edge of the Tōhoku-Oki earthquake rupture — possibly in an area that experienced a relatively small amount of slip during that event. This isn’t unexpected. The regions of a fault surrounding a large rupture are subjected to particularly large stress changes, which is why aftershocks commonly outline the area of a large rupture.

This is actually the largest earthquake to have occurred in this particular area since the Tōhoku-Oki earthquake. While that great event was followed by a number of aftershocks above M7, mostly along the deeper edge of the rupture area closer to shore, all of the earthquakes in the northern boundary region have (until now) remained M6.7 or below.

Figure 9: Recorded M6+ earthquakes around the Tohoku-Oki rupture area since 2011. Earthquakes above M6.5 are labelled. Earthquakes are colored by depth.

A natural question is whether this M6.8 earthquake is likely, or not, to be followed by a larger earthquake. After all, the 2011 Tōhoku-Oki earthquake was preceded by a series of strong foreshocks, located close to the mainshock origin location.

However, at present we have no way to identify foreshocks before a mainshock occurs. Most of the time, magnitude 6-7 events are not foreshocks of larger earthquakes. It is much better to look at the long-term history of a region, rather than to focus on individual events.

The Japan Trench, where this earthquake (and the Tohoku-Oki earthquake) occurred, continues northward for over two thousand kilometers as the Kuril Trench. This subduction zone produced the M8.8 Kamchatka earthquake on July 29th of this year — although that event was located far, far to the north of Japan. We know that many parts of this huge subduction zone are capable of producing great earthquakes; we have seen two of them in the last two decades. What about the part between the 2011 Tōhoku-Oki and 2025 Kamchatka earthquakes?

Helpfully, a new research group has formed in Japan to understand the seismic hazard of the Kuril Trench, and explore possible response measures. In 2024, Hiroaki Takahashi, a professor at Hokkaido University, published a summary of their findings thus far.

What they found is not exactly comforting. Preserved tsunami sediments provide evidence of huge tsunamis that inundated extensive parts of the coastline along the Kuril Trench. The last major tsunami likely occurred about 400 years ago, and seems to have been raised by slip of ~25 meters on the megathrust fault. On average, earthquakes of at least M8.8 seem to re-occur every ~340-380 years. The Japanese government currently estimates a probability of 7%-40% of an earthquake of M8.8+ magnitude occurring sometime within the next thirty years, with a possible maximum magnitude of M9.3 — up to twice as large as the M9.1 Tōhoku-Oki earthquake, and comparable in size to the 2004 Sumatra-Andaman earthquake. The impacts of such a maximum magnitude earthquake would be severe, with estimates of about 100,000 fatalities. Those are some eye-widening numbers.

Fortunately, the research summary does not end there. The second part of the document explores actions that could be taken to minimize these impacts, including different approaches to communication, and tsunami evacuation plans that take into account weather conditions (with four months of wintry weather every year, it is important to consider how heavy snow or ice might require modified plans). The tsunami is actually the primary cause of expected fatalities in this earthquake, and there is a real opportunity to greatly lower the number of casualties by putting more effective evacuation plans into place.

Many magnitude ~6-7 earthquakes have struck offshore of eastern Honshu over the last century, and few of them have heralded much larger earthquakes. However, the longer record makes it clear that an event similar in scale to the Tōhoku-Oki earthquake will eventually occur.

Following the logic of the megaquake advisory system (which to be clear is only officially applicable to the Nankai region), the occurrence of this moderate-magnitude M6.8 earthquake is a good opportunity for residents and organizations to revisit their preparations and planned responses in the event of a much larger earthquake.