|Figure 1: Thirty minute long three component recording of the recent earthquake off Japan|
During the past two weeks the Earth was seismically exceptionally quiet - until Tokyo and parts of Japan's east coast were shaken on New Year's Day. Although high rise buildings in Japanese cities swayed in a gentle fashion, no damage was reported and no tsunami was generated. The cause was an earthquake under the Western Pacific, about 300 miles southwest of Tokyo. The focus of the quake with a magnitude of 6.8 lay in the Earth's upper mantle, almost 220 miles below the ocean surface. Such deep temblors rarely cause damage because most of the waves' energy is absorbed and dissipates before the shaking reaches the Earth's surface.
Nearly 12 minutes after the quake started, the first seismic waves hit California. The recordings of these waves with the broadband seismometers of the Berkeley Digital Seismic Network allows the blogger to make good on a promise he made awhile ago, namely to further explore how to read a seismogram. Figure 1 shows a half hour long record of the seismic waves registered by a three component seismic station in the East Bay Hills. Green is the vertical ground movement, up and down. The red line represents the seismic shaking in the North-South and the blue line in the East-West direction. The most prominent onsets are the P- and S-waves, which arrived approximately 9.5 minutes apart. However, looking at the vertical component one can clearly see more onsets (red circle).
|Figure 2: Four prominent onsets on the vertical component|
When zooming into this area of the seismogram (Figure 2, seven minute window) four prominent peaks stand out. Each of them is called an onset, because it represents the arrival of a distinct type of seismic wave. Over the last decades, seismologists have learned how to distinguish between these various types and how to interpret their differences.
Take onsets 1 and 2 for instance: Onset 1 is the most direct wave between the quake's hypocenter and California, commonly labelled as "P" for primary. Eighty-three seconds later, another wave arrives. This wave travelled from the quake's focus initially to the surface of the Earth immediately above the hypocenter. There it was reflected and then followed the original P-wave on its way to California. This type of wave is referred to as "little p - big P" (pP). The sketch in figure 3 shows its idealized path.
|Figure 3: Idealized path of various seismic waves between Japan and California|
Onset 3 is generated by a wave which first travelled as a S-wave to the Earth's surface above the focus. There, upon reflection, it was converted into a P-wave, which followed its two predecessors across the Pacific. Because of the conversion from S to P, this onset is labelled "little s - big P" (sP). Finally, the fourth onset is generated by a wave, which was reflected from the Earth's surface roughly halfway between the hypocenter and California. This wave is referred to as PP and takes more than 100 seconds longer to cross the Pacific than the immediate P-wave. (hra071)
|Figure 1: This relief map of the area around New Madrid shows the low-lying areas of the Mississippi Valley in blue and purple. The higher elevations are depicted in green, yellow and brown. The white dots are the epicenters of recent earthquakes. The Reelfoot Fault is shown as a small red line. (Source: Center for Earthquake Research Information, Memphis, Tenn.)|
The earthquake sequence that struck the central Mississippi Valley two hundred years ago (see most recent blog entry) was remarkable. It occurred far from any tectonic plate boundary, the regions normally marked by high levels of seismic activity. Although temblors in these areas are much less frequent than in California or Alaska, such intra-plate earthquakes are not that unusual. The 5.8 quake that shook Virginia and the capital in August was a reminder that such quakes do happen. But what is it that causes the Earth to tremble thousands of miles from any plate boundary?
Put in simple terms, we can compare the stable interior of a continent with a pane of glass. Virtually every place in the contiguous United States east of the Rocky Mountains is part of this solid platform. Even though the Earth's crust in this area is quite solid when compared to California, it is not totally free of mechanical stress. On the one hand, the continental interior is battered by the actions at the plate boundary in the West. It is also still recovering from the huge load of the ice sheet of the last ice age. During that era, the northern part of the continent was completely blanketed with layers of ice several kilometers thick. It melted only 12,000 years ago, relieving the crust of its load. Another factor is the age of the crust far away from any active tectonic boundary. While the rocks along plate boundaries are usually young, continents can be several hundred million years old. Their age makes them extremely rigid, in contrast to the more ductile crust at the continental fringes.
And this is where the comparison with the pane of glass comes in: When battered at the edges, the mechanical stresses in the center of the pane can build up. Once under stress, a slight puncture at one point is enough to make the entire pane shatter. The area around New Madrid is such a puncture point. There the Reelfoot Fault, an ancient rupture line, crosses the Mississippi Valley (red line in Figure 1). Because of the accumulated stresses and the rigidity of the continental crust, slight movements along this fault can lead to larger ruptures, such as those of two hundred years ago.
|Figure 2: Earthquake hazard map of the United States. The large oval in the eastern half of the continent shows the hazard associated with the New Madrid Seismic Zone. Its center is purple, depicting a hazard similar to the one we have to live with here in California. (Source: USGS)|
Paleoseismic investigations in this area have shown that the earthquake sequence of 1811-1812 was not unique. Similar strong shaking occurred at least three times during prehistoric times - probably around 1450 and 900 A.D. and 2350 B.C. Because other strong temblors cannot be ruled out in the future, scientists from the USGS have determined that the seismic hazard in the area around New Madrid is as high as that along the San Andreas Fault or the Pacific coast of Alaska (see Figure 2). (hra070)
December 16, 2011
|Figure 1: Image from the archives the State Historical Society of Missouri, courtesy of the National Information Service for Earthquake Engineering, EERC, University of California, Berkeley.|
When you drive North on Kentucky State Highway 94 away from the Tennessee Border but close to the Mississippi, you may notice one of those ubiquitous solid bronze Historical Highway Markers (see figure 2). The inscription talks about the "greatest earthquake recorded in North America," which was centered in this area. What, major earthquakes in Kentucky? Aren't the really big ones supposed to happen only at the West Coast or in Alaska? Indeed the area where today's states Kentucky, Tennessee and Missouri meet was shaken by a major temblor exactly two hundred years ago today.
The earthquake struck at about 2:15 in the morning. One eyewitness, Eliza Bryan, described the unusual events: "The screams of the affrighted inhabitants running to and fro, not knowing where to go, or what to do - the cries of the fowls and beasts of every species - the cracking of trees falling, and the roaring of the Mississippi - the current of which was retrograde for a few minutes, owing as is supposed, to an irruption in its bed -- formed a scene truly horrible." Two more strong earthquakes struck the area within the next two months, one on January 23 and one on February 7, 1812.
Two hundred years ago, the Mississippi River valley was the frontier to the American West. Only eight years before the earthquake struck, the US had bought from the French the Louisiana Purchase, a huge swath of land west of the mighty river. New Madrid, a town across the river from today's historical marker, had been settled in 1789, and its 3000 residents were proud to live in the oldest city west of the river. The earthquakes, however, did not only destroy most of the ramshackle houses and log cabins in this frontier town. As the marker indicates, the tremors were also felt almost everywhere the white man had settled in America, from Boston and Maine all the way to New Orleans. The shaking was so severe that chimneys toppled in Cinncinati, Ohio, and church bells started ringing on their own in South Carolina.
Taking into account the huge area affected by the seismic waves of these three shakers, seismologists first rated the quakes with magnitudes of 8 or greater, clearly topping the magnitude 7.8 quake that shook San Francisco in 1906. When the historical marker was erected in 1964, the superlative of "greatest earthquake" mirrored the state of knowledge at the time. Now, however, more thorough investigations have led to a significant downgrade of the "New Madrid Earthquake Series." At most, the quakes had magnitudes of 7, maybe even only 6.5.
|Figure 2: The New Madrid historical marker.|
As in the case of the most recent quakes in Oklahoma, the area of seismic shaking was deceiving to the early students of the New Madrid quakes (see blog 9 Nov 11). Read more about the causes of the mid-continent earthquakes in the next blog.
When high-rises in Mexico City swayed for more than a minute on Saturday night, people feared for their lives. Many inhabitants of the Mexican capital remembered the nightmarish scenario 26 years ago, when similar movements led to one of the worst natural disasters our neighbors to the south ever suffered. On 19 September 1985 an earthquake, with a focus more than 200 miles away, wreaked havoc on many buildings in the capital city (see blog 19 September 2008). But the swaying on Saturday subsided without causing any significant damage.
Nevertheless, the latest 6.5 earthquake was a reminder that Mexico City is vulnerable to seismic waves approaching from country's distant Pacific coast. The earthquakes there are caused by the subduction of the Cocos Plate, one of the dozen or so tectonic plates which make up the outer layer of the Earth, under the North American Plate. The Cocos Plate approaches from the southwest with a speed of just over 2 inches a year. Upon colliding with the North American Plate, it bends downward and begins to sink into the Earth's hot mantle. The mechanical stresses associated with the bending and the temperature change cause the subducting Cocos Plate to break, hence the temblors.
The Cocos Plate diverges from the Pacific Plate along the East Pacific Rise. Its subduction dominates the tectonic regime along most of the western coast of Central America, from Panama all the way to the Mexican resort town of Puerto Vallarta. In the past century alone, dozens of earthquakes occurred along this subduction zone, among them the 8.1 temblor which cost thousands of lives and produced severe damage in Mexico City. The locations of each of the strong quakes which occurred along Mexico's Pacific coastline are shown as grey areas on the map, with the extremely destructive 1985 quake marked in green.
The area just to the southeast of the green region is conspicuously empty. For more than 150 years, no significant quake with a magnitude greater than 7 has occurred in the immediate vicinity of the tourist town of Acapulco in the Mexican state of Guerrero. The absence of large earthquakes in one region along a tectonic front is called a seismic gap. Such gaps are tectonic time bombs waiting to go off in a major earthquake. At most other locations along the Cocos subduction zone the tectonic stresses have been release by earthquakes over the last century. However, during that time, nothing happened in the Guerrero gap. There the stresses caused by subduction have built up. As a result, seismologists expect a really big earthquake in this region. It would not only flatten Acapulco but, like the 1985 temblor, would cause significant damage in Mexico City.
Although Saturday's quake (red star) occurred exactly in this gap, it did not release much tectonic energy. With a magnitude of 6.5 it was more than 75 times weaker than the 1985 temblor. That means that the wait for big one which will hit Acapulco is still on. (hra068)
Last Saturday's magnitude 5.6 earthquake in Oklahoma not only shook the Boone Pickens Stadium at Oklahoma State, where the home team, the Cowboys, had just beaten the Kansas State Wildcats 52 to 45. This largest quake in the Sooner State's history shook a huge area. Almost 60,000 people tuned into the "Did you feel it?" website from the US Geological Survey and reported how they felt the shaking. Computers automatically plotted each of these reports on a map, where the color depicts the intensity of the shaking felt by the person reporting. Blue represents a very light rumble, yellow stands for strong shaking and dark red and brown means that you would be lucky to survive.
The map of responses for the Oklahoma shaker covers an enormous area, ranging west to east from the Texas panhandle to Memphis, Tennessee, and north to south from Omaha, Nebraska, and Des Moines, Iowa, to Austin, Texas (see left panel in our figure). In total, the M 5.6 earthquake was felt in an area of more than 300,000 square miles, roughly three times the size of California.
Now compare that with the shake area of the largest Northern California Earthquake in the last 22 years, the October 31, 2007, temblor on the Calaveras Fault near Alum Rock northeast of San Jose. With a magnitude of 5.4, this quake was just a little smaller than Oklahoma's quake. Although it was felt as far away as Redding in the north, Carson City, Nevada, in the east and Bakersfield in the southeast, the area of more intensely felt shaking was concentrated around the greater Bay Area (see right panel in our figure). The region where shaking was felt consistently covers only about 50,000 square miles. Even if you double that number because California is a coastal state and the shaking cannot be felt on water, the shake area of the Alum Rock quake is only about a third of shake region of the Oklahoma quake - even though the two quakes were comparable in magnitude.
What causes this discrepancy? The answer is that the tectonic regime in the contiguous United States east of the Rocky Mountains differs substantially from the situation west of the mountain range. The Earth's crust here in the West is torn and jumbled by the tectonic collision of the Pacific and the North American plates. The boundary is not confined only to the San Andreas Fault, but influences the entire region from the Pacific Coast all the way to the Rockies. East of the range, however, most of the Earth's crust is old and geologically very stable. It has not seen any plate collisions for hundreds of millions of years. Seismic waves can travel through such stable regions with much less interference than in areas that have recently been torn up by tectonic forces. When an earthquake hits east the Rockies, the Midwest and East ring like a bell, to be "heard" for hundreds of miles. Here in California and the West, however, the seismic waves are more like a "thud." They are scattered, absorbed, and attenuated by the battered crust, thereby losing their punch in a much shorter distance. (hra067)