Figure 1: If you took away the water of the Northwestern Pacific, this is how the sea floor would look. In its move westward, the Pacific Plate slams into what looks like a wall. This is the Eastern edge of the Eurasian Plate. The yellow star denotes the location of the epicenter. (Source: NOAA)

Every textbook about Earth Science tells you that the giant lithospheric plates move past each other at a velocity of a few inches per year. The most common metaphor is that their speed compares directly to the rate at which a fingernail grows. That sounds gentle, harmless, and most of all managable. Unfortunately, these numbers are utterly misleading, because they are very long term averages measured over hundreds or even thousands of years. When you look at plate movement in the short term, you recognize a completely different picture. Most of the time, nothing happens at all. The plates are tightly locked against each other, snapping perhaps in a few small earthquakes here or there. This tectonic inactivity gives us, the residents of earthquake zones, a false sense of security. Because suddenly, within a fraction of a second and without any warning, the geologic interlocking may fail and then all hell breaks loose. The mechanical energy, caused by the constant push of the churning viscous mantle below and stored in the plates over hundreds of years, is released. In a few seconds the plates jump past each other by dozens of yards. In many respects, the plate movement is like the fate of a drag racing car. Most of time, the machine sits around idly, but once in a while, its engine is fired up and the dragster races a few hundred yards at lightning speed.

This is what happened last Friday off the east coast of Honshu. When the westward moving Pacific Plate finally unlocked itself from the Eurasian Plate, the result was the giant earthquake. Over the long term the Pacific and Eurasian Plates drift past each at an average rate of about 3.25 inches per year. But on Friday, they raced passed each other by dozens of yards in about two minutes.

	Figure 2: The rupture region (colored area) of Friday's quake may have been as large as the state of West Virginia. At the center of the area, the plates may have jumped passed each other by more than 25 meters. (Source: GFZ-Potsdam. Click to view larger image.)

Several researchers have already made models of how the interlocking between the plates actually failed, the quake's so called "rupture process". Gavin Hayes from the USGS office in Golden, CO, used the recordings of 60 seismic stations from all over the globe to compute the rupture. According to (his calculations) the rupture area underground was almost 200 miles long and 125 miles wide. Imagine, the entire state of West Virginia slipping up to 20 yards eastwards in less than two minutes.

A group of scientists from the Geoforschungszentrum in Potsdam, Germany, used the recordings of almost 500 sensitive GPS-stations in Japan to model the rupture. Their calculations yielded a 250 mile long zone with a offset of almost 28 yards (see Figure 2). Researchers from Harvard University used several hundred earthquake stations on the US mainland for their calculations and came to similar conclusions. They estimate the rupture to be 240 miles long by 150 miles across. They even simulated how the rupture spread over this area (see their animation).

In the meantime, the massive temblor has been given a name, and the magnitude of the "Great Tohoku Earthquake" has officially been upgraded to 9.0. (hra063)

The worst damage in Friday's disaster in Japan was not caused by the shaking of the seismic waves themselves, but by the tsunami. Geologic research on sand layers along the coast of northeastern Honshu has shown that the low lying areas in the prefectures of Miyagi and Fukushima have been inundated by huge tidal waves every thousand years or so. Before Friday, the last such tsunami hit the area in 869 AD. It was caused by the Jogan earthquake, which ruptured roughly the same offshore area as Friday's quake. According to historic documents, more than a thousand people perished when the tsunami washed ashore in the plains of Sendai, and the area which is now occupied by the Fukushima Daiichi nuclear plant.

Although Japan has one of the most sophisticated tsunami warning networks in the world, the coastal region around Sendai is just too close to the quake's epicenter to allow a timely warning. Even though the wave heights were forecast correctly, at more than 30 feet, the arrival of the warning was not early enough for the many inhabitants of the area to take action and flee to high ground. For the rest of Japan's Pacific coast, however, the tsunami warning was very effective.

This is also true for the warning for the whole ocean region, which was issued by the Pacific Tsunami Warning Center (PTWC) in Hawaii. Its scientists issued the first bulletin only nine minutes after the quake. It was not very specific, but stated that the earthquake was strong enough to be able to cause a tsunami. About 15 minutes later, the computers at PTWC had run the first tsunami model for the entire Pacific and the center issued a more detailed warning. It included the arrival times of the tidal wave at coastal towns in many countries and the expected wave heights. The model was updated as more data arrived at PTWC.

A tsunami travels across an ocean at about the speed of a jetliner. Thus, the wave hit the harbor town of Petropavlosk on Russia's Kamchatka Peninsula in about two hours. Five hours later, the wave arrived in Hawaii, causing minor flooding in Hilo. At around 8 am PST on Saturday morning, the tsunami reached California, causing considerable damage in the harbors of Crescent City and Santa Cruz. (See this video by a local TV station.) Finally, thirteen hours after the earthquake, the wave was registered in New Zealand. Traveling at an average speed of 495 miles per hour, it took 21 hours for the tsunami waves to reach the southern Pacific coastal region of Chile, which was devastated by an earthquake in February 2010. That event had a magnitude of 8.8 and was comparable in size to Friday's quake off the coast of Honshu (see blog March 1, 2010).

	This map shows the model calculations for Friday's tsunami. The travel times are represented by the faint blue lines. The colors show the wave heights, from more than 30 feet (dark blue and purple) to less than 1 foot (yellow and green). The model agreed very well with the actual measurements. (Source: Noaa, Click to view larger image.)

The PTWC was established in 1949 after Hawaii suffered major damage from a tsunami caused by an earthquake in Alaska. At first, it issued warnings only for Hawaii, Alaska and the US West Coast. After the giant Chile earthquake of 1960, an intergovernmental agreement extended the PTWC's responsibilities to the entire Pacific basin. During its early years PTWC relied only on seismic measurements. Later, data from tidal gauges began to be used, and after the Indian Ocean tsunami in 2004, many deep sea observatories were added. These sensors are connected by cable to buoys at the ocean's surface, from which data are sent by satellite links to the center's main building near Honolulu. PTWC is operated by NOAA. (hra062)

	USGS map showing location of Japan's magnitude 8.9 earthquake and aftershocks.

The earthquake that devastated some parts of the Japanese island of Honshu on Friday was the strongest quake ever measured in Japan. The National Earthquake Information Center (NEIC) of the USGS in Golden, CO, determined its magnitude as 8.9. NEIC scientists routinely use recordings from seismometers from all over the world to compute the strength of a quake. The Japanese Meteorological Agency, which is responsible for earthquake monitoring and tsunami warning in Japan, determined the magnitude as 8.8. Their scientists computed the value from regional seismic networks.

Even though the magnitude of Friday's quake was very large and some coastal areas in northeastern Honshu's Miyagi prefecture were devastated, the quake was by no means the worst natural disaster to hit Japan. On September 1, 1923, large parts of Tokyo and Yokohama were destroyed by the "Great Kanto Earthquake" and subsequent fires, which raged for days. More than 100,000 people lost their lives and almost 400,000 buildings were destroyed. More than 5,500 people died on January 16, 1995 in southern Honshu when the region around Kobe was hit by a quake with a magnitude of 6.9. The damage to Kobe's infrastructure was severe, as the quake toppled elevated freeways and submerged docks in the busy harbor.

On a global scale as well, Friday's quake off the coast of northern Honshu was one of the strongest temblors ever measured with seismometers. When using the USGS magnitude of 8.9, it ranks fifth on the list of most severe quakes in the last century, topped only by the 9.5 quake in Chile in 1960, the 9.2 quake in Alaska in 1964, the 9.1 Indian Ocean quake on Boxing Day 2004, and the 9.0 quake on the Russian Kamchatka Peninsula in 1952.

The seismic waves from Friday's quake were registered all over the world. It took about 12 minutes for the first seismic waves to reach the US West Coast and be recorded by the Berkeley Digital Seismic Network here in Northern California. The seismogram shown here was captured at station BKS, which is located in a tunnel just above the University Botanical Gardens.

On the other hand, it took many hours for the tsunami to cross the Pacific. Around 8 am (PST) the first effects were measured on the West Coast. In Crescent City the Tsunami reached a height of 7 feet, and along the Monterey coast the wave was three feet high, while hardly anything could be measured in the region around the Golden Gate. There the arrival of the tsunami coincided with low tide. (hra061)