The earthquake in Japan was one of the biggest on record and came with no warning. E&T explains why they are so hard to predict but how studying recent events could help.
The earthquake that rocked Japan on 11 March 2011 was a magnitude 9.0 – the fourth largest seismic event ever recorded. The unprecedented quantity of photography and video of the devastation on a modern society, such as Japan, were both spell-binding and horrific – cars and boats thrown about like tiny toys in the ensuing tsunami, and buildings effortlessly reduced to flotsam in moments.
The seismic setting
The beginning of this cataclysmic event can be found beneath the Pacific Ocean with the East Pacific Rise (EPR). This is a spreading zone where upwelling magma from the internal oven of the Earth's mantle rises toward the surface and spreads out, forming new oceanic crust that moves inexorably toward Central and South America to the east, and across the Pacific toward Asia to the west.
To the west of the EPR, this newly-formed tectonic crust forms the Pacific Plate – the largest of the dozen or so tectonic plates that comprise the surface of the Earth. From the EPR, the Pacific Plate moves steadily, like a conveyor belt, at a rate of about 90mm per year. After about 40 million years, the plate moves over the Hawaiian 'hot spot' – a stationary bubbling plume of magma, indicated by the active volcanoes on the south east coast of the Big Island of Hawaii. It rafts the islands of Maui, Molokai, Oahu and Kauai away to the north and west, until they sink back beneath the waves in the Leeward Islands and undersea mounts.
The Pacific Plate moves onwards until, after another 60 million years, it meets the Eurasian and North American Plates – a giant conjoined continental land mass, made of lighter felsic minerals than the heavier oceanic crust. As the plates collide, the heavier oceanic material is subducted downwards below the continental plates, forming a trench some 9,000m deep, where the Pacific Plate is finally subsumed back into the oven of the Earth. In this way, all tectonic plates are formed, move across the surface and are recycled in a process as old as the Earth itself. This 'ring of fire' surrounding the Pacific Ocean basin is one of the most active earthquake zones on the planet.
At the plate boundaries, they collide into each other or slide side-by-side, as is the case of the San Andreas Fault in California, in fits and starts. They become stuck and build up stress until finally – and sometimes violently and catastrophically – they give way in the form of an earthquake.
On 11 March at 1446 JST, a large section of the plate boundaries off the east coast of the Japanese island of Honshu shifted vertically by as much as 30-40m, the downward-trending Pacific Plate sinking into the Japan Trench relative to the continental land mass. A large section of the continental plate margin, approximately 560×280km gave way, displacing a huge volume of overlying water – picture an area of about 160,000km2 suddenly being yanked downwards and springing back – resulting in a 10m-high wall of water. The tsunami, travelling at 800km/h – as fast as a jet airliner – devastated the Honshu coast and the community of Sendai.
The energy released in an earthquake moves as waves away from the epicentre. These seismic waves (not to be confused with tsunamis or oceanic waves) can be measured in various ways by means of seismometers.
The earliest 'seismometer' was built in China in 132AD. A large globe was adorned with eight dragons, each holding a bronze ball in its mouth. Eight open-mouthed toads sat beneath the dragons and, according to design, the ball would drop into the toad's mouth in the direction of the earthquake; the other balls presumably spilling to the side of their targets. This would indicate the direction of the quake, but not its magnitude.
Around 1880, British physicist Sir James Ewing, geologist John Milne and engineer Thomas Gray developed the first horizontal pendulum seismometer. These operate on the principle that a part of the device remains stationary (suspended from the pendulum), while the machine itself moves with the earthquake, allowing the stationary mass to record the relative movement on a seismograph. The horizontal movement was first recorded on a revolving smoked-glass plate.
Later versions developed by the team added a vertical motion detector by means of attaching a spring to a bob and lever system; and combining these horizontal and vertical motion detectors into a duplex-pendulum seismometer; and using photographic recordation of the motion as opposed to etched smoked-glass or paper.
In 1887-1888, a series of ten duplex, horizontal and vertical seismometers was arrayed in northern California at the University of California at Berkeley and on nearby Mount Hamilton, which would play an important role in documenting the huge San Francisco earthquake of 1906.
Later versions of pendulum seismometers were all electromagnetic, suspending a magnetic mass within an electric coil. The magnet's motion within the coil can be converted into an electric signal, which can be used to generate light to be recorded on photographic paper. John Milne can also be credited with pushing for establishment of a global network of seismographic detection centres using standardised instruments. A version of his photographic-recording horizontal pendulum seismometer was selected by the British Association for the Advancement of Science as the standard instrument.
Modern broadband digital seismometers are capable of recording ground motion over a wide range of frequencies and seismic amplitudes. Whereas the original pendulum versions were limited by measuring the amount of movement of the mass relative to the housing of the instrument, modern seismometers measure the amount of electrical energy needed to keep the mass centred in the housing relative to the external ground motion, enabling them to record the full spectrum of seismicity, from the tiniest, imperceptible 'microseisms' to a large-magnitude earthquake.
How earthquakes work
When an earthquake occurs, the seismic energy released by the event moves out and away from the epicentre in waves. Two basic classes of seismic wave are emitted by the quake: 'body' waves, which move through the inner Earth, and 'surface' or 'long' waves (L waves), which move along or near the surface. L waves are generated by body waves. They are slower than body waves and of lower frequency, but L waves are the ones that do the most physical damage to structures due to their shaking intensity at the surface.
Body waves are characterised as 'primary' (P) and 'secondary' (S) waves. The first and fastest waves to be recorded on a seismograph are the P waves, travelling at between 1.6 and 8km/s, followed by the slower, low-frequency S waves, and finally by the L waves. P waves can travel through rock or liquid, including water and magma, whereas S waves cannot move through the molten inner Earth.
Actual rates of travel of the body waves are variable depending upon the nature of the Earth material that they are moving through, but the relative velocity between P and S waves is constant (P is about 1.7 times faster), such that the difference between the arrival times at the seismometer between the S and P waves can be used to determine the distance to the epicentre from the monitoring station.
If there were no subsequent S waves, then one knows that the earthquake was on the other side of the Earth because S waves cannot travel through the liquid core material, and there is a large 'shadow zone' of about 150˚ on the opposite side of the Earth from the earthquake that will not indicate the S waves.
In a typical seismograph, the initial motion of the P waves hits the sensor first. In the seismogram of the Honshu Earthquake, the arrival of the P waves precedes several small fore-shocks, followed by the much higher-amplitude S waves. The S waves are of lower frequency on the seismogram, but are much larger in terms of the energy associated with them.
There is a slight gap between the peak oscillations of the S waves and the onset of the devastating L waves, indicated by the intense black lines that gradually taper down over the next 15 minutes of the event. These are of still lower frequency than S waves, but represent the major portion of total energy released. Note the several high-amplitude spikes of the L waves, the largest one representing the peak momentary magnitude of the headline-making Japanese earthquake, calculated at an incredible 9.0.
Knowing the velocities of P and S waves, one can calculate the distance from the seismometer to the epicentre by subtracting the difference in time between the arrival of the first P waves from the S waves. Furthermore, by measuring the amplitude of the S waves and correlating with the distance from the epicentre, one can calculate the magnitude of the event.
Richter & Gutenberg
The magnitude of an earthquake is typically described according to the moment-magnitude scale, which has superseded the well-known Richter Scale, named for Charles Richter and Beno Gutenberg, engineers at the California Institute of Technology. Although the scales are calculated in a different way, the new scale retains the familiar continuum of magnitude values defined by the older one.
Before their development of a quantitative means of measuring magnitude, earthquakes were described according to the Mercalli Scale – a 12-point ranking based upon how buildings and people responded to the shaking. The Mercalli scale was subjective and depended very much on the distance from the epicentre and people's perceptions of the extent of shaking. A quake that shook a chandelier might be a 2 or 3, while one that knocked down buildings would be a 10.
Richter and Gutenberg based their scale, published in 1935, on the measured motion of a seismograph for a Wood-Anderson torsion seismometer. Instead of a pendulum, this machine used a small copper mass attached to a thin wire under high tension. The motion of the mass was damped by magnets – a substantial improvement over other damping methods in that the proportion of damping provided by the magnets is proportional to the motion of the object being damped.
Richter arbitrarily set the scale at "0" representing 1µm, as measured on his seismometer at 100km from the sensor; and thence upwards from 1 to 10. It is a base-10 logarithmic scale; that is, each point is 10 times the point before it, such that a magnitude 6 event is ten times greater in terms of actual measured motion than a 5 and a 9 is a thousand times more movement than a 6.
The amount of actual energy released by an earthquake, which correlates to its destructive potential, increases by 32 times for each point on the Richter Scale – or in other words, a 9 releases 32,768 times more energy than a 6.
So, one can tell the distance and magnitude of an earthquake from a seismic monitoring station, but we still have not determined exactly where the epicentre was located. With one station reporting, one can tell that the earthquake was a certain radius distance from the monitoring station by the difference between receipt of the S waves minus P waves, and if we have a second station reporting, the location is further reduced to just two possible points; but one must have at least three stations reporting to absolutely locate the epicentre of the event by 'trilateralisation' of the three distance-estimate circles from their stations.
Lessons from past events
Japan and California lead the world in their geotechnical preparations for major earthquakes, in part because both share the risks inherent in living on the edge of the 'Ring of Fire'. Lessons learned from the Loma Prieta earthquake (the 'World Series' quake – a 6.9 that killed 63 persons and injured more than 3,700 in 1989) and the Kobe quake (a 6.8 that killed 6,434 people in 1995) had provided much incentive for advancement and retrofitting of geotechnical safety measures for buildings and construction.
Most of the fatalities from the Loma Prieta quake were the result of the collapse of the vertical reinforcement bar-supported columns of the double-decker Cypress Freeway structure in Oakland. Now all new concrete construction must also have coiled rebar to prevent lateral motion failure. Base isolators were installed to isolate some buildings from lateral ground motion; and dampers – like shock absorbers – were built into the foundations of new municipal buildings.
In the aftermath of the Kobe quake, 120,000 structures were fully or partially destroyed and another 7,000 burned as a result of a subsequent fire. Damage was estimated at more than $100bn. Since then, Japan has implemented strict seismic retrofit standards, a nationwide tsunami and disaster warning system, and the world's most rigorous disaster preparedness training programme.
Yet, despite all of the best efforts for geotechnical and public preparedness, the great Tohoku earthquake, as it is called in Japan, caught everyone by surprise. Tall buildings in Tokyo and other cities fared well, but no seismic code standards could withstand the sheer power of the tsunami, wiping whole blocks of buildings to the ground. Even the multiple fail-safe measures for their nuclear power plants were obviously not enough to maintain coolant in the fuel rod containment vessels.
As I conclude this article, the full extent of the Tohoku disaster is still unfolding, with the death toll likely to exceed 20,000, radiation releases from the Fukushima nuclear power plant remain at serious levels, with evacuations ordered for all persons within 20km of the plant; and aftershocks of greater than 6-magnitude are still rattling nerves and the remains of the buildings still standing.