Tokyo Shaken: What Really Happened

Section: International
Format: Special Report
Author: Sinisa Brkic (sb)

A magnitude 5.9 earthquake beneath southern Ibaraki Prefecture shook Tokyo and large parts of eastern Japan shortly after 2 a.m. on Sunday, injuring at least 37 people and triggering emergency alerts across the capital region. The earthquake caused limited infrastructure disruption, but its effect across one of the world’s largest metropolitan areas was striking. The event also produced an apparent contradiction that quickly spread through alerts and online reports. Early calculations put the earthquake above magnitude 6, including estimates of 6.4 and even higher during the first seconds of analysis, while the Japan Meteorological Agency later reported a magnitude of 5.9. The difference is not evidence of conflicting earthquakes. It is a revealing example of how modern seismic warning systems operate when seconds matter more than initial precision.

What happened beneath Ibaraki

The earthquake struck at approximately 2 a.m. local time on August 23, with its source beneath southern Ibaraki Prefecture, northeast of central Tokyo. The Japan Meteorological Agency placed the focus at roughly 70 kilometers below the surface and reported a magnitude of 5.9, with a maximum seismic intensity of 5 lower on Japan’s Shindo scale in parts of the Kanto region.

The shaking extended well beyond the immediate epicentral area. Much of Tokyo registered intensity 4, while stronger motion was recorded at several locations around the wider metropolitan region. The United States Geological Survey placed the event at magnitude 5.8, close to a later JMA moment magnitude calculation of 5.8.

Those figures describe the energy released by the earthquake. They do not, however, directly describe how violently the ground moved at every location. That distinction is central to understanding why a magnitude 5.9 event could attract such intense attention across Tokyo.

Why 6.4 became 5.9

Japan’s Earthquake Early Warning system is designed to provide useful warning before the strongest seismic waves arrive. To accomplish that, it begins calculating an earthquake’s location, depth and likely magnitude from the first seismic signals detected by instruments close to the source.

Speed is the system’s purpose. Final precision comes later. For this earthquake, the automated calculations changed rapidly as additional stations began receiving seismic waves. The first estimates were around magnitude 5, but within seconds the system produced substantially larger values. One calculation reached 6.7 only a few seconds after detection, while later warning updates moved through estimates around 6.0, 6.1, 6.3, 6.4 and 6.6 as the system refined the source parameters.

The widely circulated figure of 6.4 therefore had a legitimate origin. It was part of the live warning process, not the final determination of the earthquake’s magnitude.

Once significantly more seismic data were available, the agency could calculate the earthquake with greater confidence. JMA reported magnitude 5.9 at a depth of about 70 kilometers, while its moment tensor analysis produced a moment magnitude of 5.8 at approximately 67 kilometers.

This is precisely why an early warning value should not be treated as a final scientific measurement. The warning system is making a time critical estimate from incomplete information while the earthquake is still unfolding across the sensor network.

Magnitude and Shindo answer different questions

Confusion also arises because Japan routinely publishes two kinds of numbers after an earthquake. Magnitude measures the size of the earthquake at its source. Shindo measures the intensity of shaking at a particular location.

One earthquake therefore has a magnitude, but it can produce many different Shindo intensities across different cities and districts. Distance from the source matters, but so do local geology, soil conditions, building characteristics and the frequency content of the seismic waves.

A Shindo intensity of 5 lower is significant. At that level, many people feel frightened and instinctively want to hold onto something stable. Objects can fall from shelves, unsecured furniture may move and some utility or transportation systems can automatically stop for safety checks.

Tokyo itself did not uniformly experience that level of shaking. Many locations in the capital recorded intensity 4, but the earthquake affected an enormous, densely populated metropolitan system where even moderate shaking can be experienced by millions of people simultaneously. That scale matters. A moderate earthquake beneath a sparsely populated region and the same size earthquake beneath the Kanto metropolitan area are entirely different public events.

The high rise effect

One of the most important measurements from Sunday’s earthquake was not the headline magnitude. It was the detection of long period ground motion across parts of the Tokyo metropolitan area. JMA recorded long period ground motion at Class 2 in southern Saitama, northwestern Chiba and Tokyo’s 23 wards. Class 1 motion was also measured across additional areas of Ibaraki, Tochigi, Gunma, Saitama, Chiba and Kanagawa.

This type of motion is particularly relevant to high rise buildings. Conventional seismic intensity is primarily designed to describe ground shaking near the surface, but tall structures can react strongly to slower seismic waves whose periods interact with the natural movement of the building.

The result can be a slow, pronounced swaying motion on upper floors even when the intensity recorded at ground level does not appear extraordinary. A person in a tower may therefore experience an earthquake very differently from someone standing at street level several blocks away.

JMA maintains a separate four level scale for long period ground motion for precisely this reason. It is intended to describe the potential effects inside tall buildings, where prolonged swaying can affect people, furniture, elevators and other building systems in ways that an ordinary Shindo value cannot fully capture. That does not mean Tokyo’s high rises suffered widespread structural damage. No such pattern has been reported. It does explain why residents in tall buildings could experience the earthquake as unusually noticeable despite a magnitude that, by Japanese standards, is not exceptional.

Injuries rise as the damage picture becomes clearer

Initial reports suggested very limited consequences, but the picture developed as local authorities collected information during the morning. At least 37 people were reported injured across five prefectures, including 15 in Tokyo, nine in Saitama, eight in Kanagawa, four in Chiba and one in Ibaraki.

Most of the injuries were reported as minor and involved people falling or striking objects after being startled awake. One elderly woman in Yokohama suffered a serious hip injury. The figures illustrate a recurring feature of nighttime earthquakes: injuries can occur not only from structural damage, but from sudden movement in dark homes when residents are disoriented.

The infrastructure effects remained limited in comparison with a major destructive earthquake. A water pipe ruptured in Tokyo’s Koto Ward and was later repaired, while power was temporarily lost to about 460 households before service was restored. Several rail services experienced delays or safety inspections, including routes serving Tokyo and Narita Airport, while Shinkansen services were not significantly affected. No tsunami warning was issued. Authorities also reported no abnormalities at nuclear facilities monitored after the earthquake, an important point in a country where seismic events immediately trigger extensive checks of energy infrastructure.

Tokyo’s warning system did what it was built to do

The changing magnitude estimates may appear to expose a weakness in Japan’s warning system. In reality, they demonstrate the difficult tradeoff at the heart of earthquake early warning.

An earthquake cannot be forecast hours in advance by the system. The warning begins only after the earthquake itself has started. Seismic instruments detect the faster initial waves, computers estimate the source and probable shaking, and alerts can then reach locations before the more damaging motion arrives. Near the epicenter, that advantage may amount to only a few seconds or may disappear entirely. Farther away, those seconds can be enough to slow trains, stop elevators at safe floors, suspend industrial processes and give people time to protect themselves.

The price of that speed is uncertainty in the first calculations. Depth, location and magnitude estimates can change as the network receives more data. A warning that initially overestimates an earthquake is therefore not necessarily a system failure. It can be the consequence of a system deliberately designed to act before the complete seismic picture exists.

For public communication, however, the distinction is essential. A live warning estimate of 6.4 should not remain in headlines after the agency has established a magnitude of 5.9. Keeping the two figures separate is not a technical detail. It is basic accuracy.

Why there was no tsunami alert

The earthquake did not produce a tsunami threat, and no warning was issued. That sharply separates Sunday’s event from major offshore earthquakes that can displace large volumes of seawater and generate destructive waves along Japan’s coastline.

The absence of a tsunami alert does not make an inland or deep earthquake irrelevant. Strong ground motion alone can disrupt transportation, damage utilities, stop elevators and cause injuries, particularly across a metropolitan area with Tokyo’s population density and vertical architecture. For residents, the correct response is therefore not to equate the absence of a tsunami with the absence of danger. The relevant risk in this case was shaking itself and the secondary effects it can produce inside buildings and across urban infrastructure.

Aftershocks remain possible, but prediction has limits

Japanese authorities have urged continued caution for further seismic activity following the earthquake. Aftershocks are a normal part of the adjustment process following many earthquakes and can occur over days or longer, although their timing and exact magnitude cannot be predicted.

That warning should not be interpreted as evidence that a much larger Tokyo earthquake is imminent. Seismology cannot use a single moderate earthquake to declare that a major metropolitan disaster is about to occur, and responsible interpretation requires a clear boundary between preparedness and prediction. Japan has strong reasons to prepare continuously for large earthquakes. Sunday’s event does not change the fundamental seismic reality of the country, but neither does it provide a scientifically credible countdown to a larger catastrophe.

A moderate quake, a megacity scale reaction

The most revealing aspect of the Ibaraki earthquake is the contrast between its magnitude and its metropolitan impact. A magnitude 5.9 event produced no widespread destruction, no tsunami and no reported abnormalities at nuclear facilities, yet it activated emergency alerts, disrupted parts of the transport and utility network, injured dozens of people and produced noticeable movement across Tokyo’s high rise landscape.

That is not a contradiction. It is what happens when a seismic event intersects with a metropolitan area of extraordinary density, complex infrastructure and thousands of tall buildings. Sunday’s earthquake also demonstrates why a single number is rarely enough to describe seismic risk. Magnitude explains the size of the earthquake. Shindo describes local shaking. Long period measurements reveal another dimension of what can happen inside high rise structures. Early warning estimates serve an entirely different purpose: giving people and automated systems whatever time can be gained before the strongest motion arrives.

For Tokyo, the earthquake was not a major disaster. It was something more instructive: a real world stress test of how a megacity experiences, measures and responds to sudden seismic motion. The correct lesson is neither complacency nor alarm. It is that understanding what the numbers mean is as important as reporting the numbers themselves.