The 1975 Tangshan Earthquake

The deadliest earthquake of the 20th century

On July 28, 1976 at 3:42 am, an earthquake of magnitude 8.2 struck near the east coast of China. The epicenter was near Tangshan, an industrial city with a population of about 1 million people. (Yong et al., 1988). This quake's destruction was worsened by the fact that it struck in the middle of the night. Almost everyone in the city was asleep, and many people were probably crushed to death without even waking up. Many more who lay injured in the rubble died before they could be rescued. The quake knocked out power through the city, making rescue efforts by shocked residents of the city impossible in the dark. A smaller number of people were trapped in nearby coal mines. Many were rescued, but not until hours or days later (Yong, et al., 1988).

Officially, the Chinese government estimated between 240,000 and 250,000 people were killed (Yong et al., 1988). In the decades since the quake, the death toll is estimated closer to half a million. Either number would make this the most deadly quake in the twentieth century, and the strongest since the Alaska quake in 1964 which was of magnitude 8.4 (Bolt, 1993b).

Geologic Setting

Below the surface lies China's complex geology. Forces pushing in two different directions are squashing the Asian continent. The combination of forces has made China a very active location for earthquakes throughout history. Earthquakes have also played a significant part in Chinese science and culture. The Chinese were the first to develop functioning seismometers, and a record of major quakes in the region has been reconstructed dating back to 1831 BC (Yong et al., 1988).

Tangshan lies on a block of continental crust bounded by major faults. The faults can be hundreds of kilometers in length and have experienced large displacements in the terrain over time. The faults are locked together, grinding against each other as tectonic forces build up. Eventually, the forces will be great enough to break the fault and the the crustal blocks will slip past each other. When the blocks slip, a great amount of energy is released. The energy moves outward much like the ripples created by throwing a rock into a pond. It is this energy that shakes and deforms the earth surface and that we record on seismometers.

China is being squeezed as the Indian plate moves northeastward. As India pushes on the Asian continent, the crust wrinkles and bends, often causing cracks in the cool, rigid crust. The force of this movement was enough to raise the Himalayan mountains, the highest in the world (Sullivan, 1976).

The Pacific plate is also squeezing China, but from the west. This oceanic plate is being subducted beneath the Asian continent at a steep angle because the oceanic crust is old and dense. The plate has had time to cool since its formation. The steep angle of subduction causes a strong horizonal force which acts on the continent. The area of northern china hit by the Tangshan earthquake is recognized as being particularly prone to the westward movement of the Pacific plate. This particular quake occured on a fault in the Tancheng-Lujiang wrench fault system (or Tan-Lu) which is a fault zone, or collection of approximately parallel faults. The zone is very wide and has more than 750 km of displacement (Jiawei, 1993).

The City of Tangshan

Tangshan lies on a block of continental crust bounded by major faults. The faults can be hundreds of kilometers in length and have experienced large displacements in the terrain over time. The faults are locked together, grinding against each other as tectonic forces build up. Eventually, the forces will be great enough to break the fault and the crustal blocks will slip past each other. When the blocks slip, a great amount of energy is released. The energy moves outward much like the ripples created by throwing a rock into a pond. It is this energy that shakes and deforms the earth surface and that we record on seismometers.

The Tangshan earthquake occurred when a fault 150 km long ruptured. You can imagine the amount of energy released! The magnitude of the quake is determined by the amount of energy released during the earthquake. Larger quakes are often the result of slip on a fault that has not moved in a long time. The longer time between earthquakes allows the bonds across the fault rocks to be stronger and more energy to be stored up. The quake has to release more energy for the rock to relax (Yong, et al. 1988).

When one section of a fault moves, movement on other sections of the same fault, or movement on other faults can be triggered. This may have happened on this occasion in China. Only 16 hours after the big earthquake, an aftershock of magnitude 7.8 occurred in the same vicinity. This second quake's damage was particularly serious in Tangshan. The city was described by some as being totally destroyed, with no buildings left standing (Munro, 1976).

Actually, some buildings did fare well during at least the first quake, especially newer buildings. An eight-story framed concrete structure, only a few years old, withstood the tremor, but the older brick buildings surrounding it were completely toppled. Some other concrete buildings collapsed onto their first floors, but the upper levels remained intact.

To the north of Tangshan, where the coal mines operate, most buildings remained standing. In this area the sediment is thin above the basement rock, allowing less liquefaction, and thus less shaking and damage. Tall chimneys are often used to determine the intensity of an earthquake. Those in the Intensity XI area all collapsed. In other areas, chimneys collapsed at different heights according to the intensity at their location. The aftershock leveled everything left standing after the main tremor (Yong, et al., 1988).

The Tancheng-Lujiang Fault System

The Chinese have been monitoring earthquakes for thousands of years. They have the longest and most detailed seismological records of any country in the world (Yong et al., 1988) As the technology and methods for measuring and studying earthquakes improved, increasingly detailed studies of eastern China's geology were conducted. A giant 3600 km north-northeast trending fault system with more than 740 km of horizontal displacement across the fault zone was defined. Until the middle of the 20th century, the Tancheng-Lujiang (Tan-Lu) fault system was just thought to be several parallel, similar faults (Jiawei, 1993). Since it was recognized as one fault system, it has not just been recognized as a source of earthquakes. Several major geologic features have also developed as a result of the plate movement, including mountains in the compression zones and valleys in the extensional areas. The Himalayan mountain range to the southwest was created as a result of India's movement toward the Asian Plate, just to give an idea of what kind of features can be created in some areas. The collision of India with Asia may influence this fault zone, even from so far away. The pushing motion may help drive one side of the fault northward.

Physical and Geometric Characteristics

The Tancheng-Lujiang, or Tan-Lu, wrench fault system is a large fault system located in eastern China. It extends from the north bank of the Yangtze River all the way into the former USSR. The zone of intertwined faults is 5000 km long and up to 1000 km wide (Jiawei, 1993). All of the faults are straight or nearly straight and have steep dips, 70°-80°+ as would be expected in a strike-slip fault. All faults have been cut by newer, almost vertical faults.

There are several types of rock associated with this fault zone, the most important of which are mylonites, cataclastic breccias, and fault gouge with tectonized pebbles.

The mylonites are found in Precambrian metamorphic basement rock right next to the fault zone. Typically, they are thought to have formed at a depth of 8 to 15 km. Also in these Precambrian rocks are protomylonites, phyllonites, and ultramylonites, which are often found with the glassy, black pseudotachylites. Visible in these metamorphic rocks are distinct foliation and lineation. Other microstructures indicate sinistral horizontal movement. The cataclastic breccias are widespread along the fault zone. East of Hefei, as well as to the north, they are several hundred meters wide.

Also present is a fault gouge with tectonized pebbles. This fault rock is present along the entire length of the fault zone. There are lots of subrounded tectonized pebbles ranging from 2 to 20 cm. All of the pebbles are either polished or striated with slickensides. The slickensides are criss-crossed by shear planes or extension joints containing chlorite and or other metamorphic minerals. It is theorized that the gouge formed in many phases of compression and that the fragments were rotated, compressed, sheared, and polished during cycles of deformation. This is supported by fracture belts in which there are small folds sometimes re-deformed by later stress.

Next to gouge belts, lenses form by perpendicularly acting compression which flattens grains. These grains also have tension joints from being extended in one direction (Jiawei, 1993).

Kinematic Analysis

The fault zone is a northnortheast-southsouthwest trending set of almost parallel sinistral strike-slip faults in eastern China. The amount of displacement across the zone was not even known until fairly recently, around 1980. No one had quite realized how large the displacement was because correlative units across the fault were difficult to identify. The last estimate of displacement is greater than 740 km (Jiawei, 1993). Structures associated with the fault zone include a huge ductile shear zone. The amount of shear and deformation across the zone are what made estimates of displacement difficult. Structures associated with this fault zone have a wide range of sizes and shapes including small folds, the fore mentioned shear zone, and mountains and basins at each end. Pairs of mountains and basins formed diagonally from each other because of the stretching and compression associated with fault movement.

The areas beyond each end of the fault zone are essentially stationary relative to fault movement. As a result, the northward moving side of the zone plows into the stationary rock to its north. This movement causes a compressional are on the east side of the north end of the fault.
 
 
 
 
 
 
 
 
 
At the other end of the fault, the northward moving plate is trying to move and pulling apart from the stationary area just to its south. This causes a stretching and an area of extension on the east side of the south end of the fault.

In the areas of compression, the rocks become folded and bent. In the areas of extension, the thinned crust forms a low basin. And, in the areas where the faults are, the high grade metamorphic rocks have been smeared and stretched like silly putty. All faults are very clean cut and look like someone just cut the continental crust with a knife (Jiawei, 1993). The middle section of the fault zone cuts the upper and lower crusts, and even extends 100 km into the upper mantle, but there is no apparent displacement of the Moho (Jiawei, 1993).

Timing

Timing of faulting varies along different segments of the fault zone. The first strike-slip displacement was during the Mesozoic. This motion was sinistral strike-slip in the Archaean basement rock. The law of superposition suggests that the breccia formed during the early to late Cretaceous, after the mylonite. Through its history, the area has experienced strike-slip faulting in the opposite direction, extension, and is now a strike-slip fault again. The faults in the zone are still active today.

Regional Tectonic Significance

Earthquakes are a very important part of Chinese society, and they have learned to cope accordingly. The Chinese are among the leaders in earthquake prediction methods, and have managed to successfully predict a few major quakes. Two quakes they missed struck in July 1976 and turned out to be most deadly in the 20th century, killing around 500,000 people. One quake was located in the industrial city of Tangshan and had a magnitude 7.8, the other was just outside the city with magnitude 7.6. The two quakes struck within only 18 hours of each other (Yong et al., 1988).

The First Seismometer

The Chinese have been monitoring earthquakes for almost 2000 years. In AD 132, Zhang Heng invented the first seismometer. It consisted of a jar with ornamental dragons on its exterior. The dragons held a small ball in their mouth that would roll out if the jar was shaken in the direction parallel to the mouth opening. The jar was surrounded by frogs with their mouths open. The frogs would catch any released ball and thereby record the direction of ground shaking. However, this instrument could not measure how large the shaking was (Yong et al., 1988).

Earthquake Prediction

Obviously, being able to successfully and reliably predict the occurence of earthquakes around the world, especially powerful ones, would save innumerable lives and billions of dollars. Throughout history, people have attempted to predict earthquakes in their area by observing and cataloging everything from strange animal behavior to "earthquake weather". In the last three decades, scientists have been working toward a goal to make earthquake forecasting about as reliable as weather forecasting. They wish to be able to predict not only the date and location of a quake, but also the intensity of damage to different regions.

Why Predict Earthquakes

Besides the obvious damage to lives and private property, knowing the potential for earthquakes can be useful in other ways. Having information about the possibility of earthquakes would be very handy when choosing the site of a major structure, such as a dam or nuclear reactor. Putting such a structure in an active area could lead to many more deaths and problems than having the same structure in a stable area.

At present, we can determine which areas of the Earth have the most potential for damage, but we cannot determine within a short time when a major earthquake will occur. We know from historical records and observation how often many faults move, thus we can tell which are due to move next. However, we can only guess with an accuracy of +/- 50 to 150 years; not very useful for evacuating towns before a quake. Narrowing the window of earthquake occurence to a few hours or days would be extremely useful, but at the same time is the hardest part of earthquake prediction.

Harbingers of Doom

It has been suggested that there are certain events which occur before earthquakes. Things such as strange animal behavior and "earthquake weather" have not been studied very closely by scientists, as these signs are not considered very accurate. There are, however, signs which are more measurable and more reliable.

The strain in rocks can be measured across faults. As stress builds up, it can be measured and a maximum strain can be calculated, and a very general prediction about when the fault will rupture can be made.

An obvious and often reliable prediction tool is the measuring of foreshocks. These are small shocks, similar to aftershocks, that precede a main large quake. They grow in magnitude and are usually all in the same area. This area can be forewarned of a possible earthquake and loss of life can be avoided. The Japanese have set up an extensive network of undersea seismometers off the eastern coast of Honshu. With these seismometers, not only can foreshocks be recognized and studied, but they can also often predict tsunamis.

The most reliable sources of pre-earthquake clues are changes in P wave velocity, ground uplift or tilt, radon emission, electrical resistivity, and the number of local earthquakes.

Changes in P Wave Velocity

It was hypothesized that if the properties of rocks change before an earthquake, then the velocity at which various earthquake waves travel through them might change as well. The former Soviet Union was the first country to study this phenomenon thoroughly. They noticed that the P-wave velocity changed by about 10%-15% before earthquakes. The velocity decreases for a while, then increases back to normal just before the main shock... at least sometimes. These measurements are very easy to make because the velocity of earthquake waves are precisely what seismometers are built to measure. Seismometers and earthquakes are located all around the world, so we can measure velocity changes without adding new, expensive equipment. However, using this method alone, the precise location of the shock is difficult to predict. Of course this method does not always work, as P-wave velocities only change over a significant area before a large quake. For small to moderate quakes, other clues to forthcoming earthquakes must be found.

Ground Uplift/Tilt

As stress builds up along a fault, the ground level can change. The ground often tilts or rises in the area of a coming earthquake. For example, Palmdale California has been experiencing uplift since the 1960s. There have been no significant quakes in the area since the uplifting began, but the area is known historically for being seismically Unfortunately, many processes cause the ground surface to rise or tilt including previous earthquakes and upcoming volcanic eruptions which makes this technique difficult to use as a predicting tool. This method has been useful in Japan, however, making it worthy of further investigation.

Increased Fluctuation in Radon Gas Emmision

The next earthquake precursor is an increase or fluctuation in the amount of radon gas in an area. Deep wells have been noted to experience an increase in the release of radon gas in the area around a forthcoming earthquake. However, this data is scarce and is not known to be a clue to coming quakes or just a normal occurrence, not related to seismicity.

Changes in Electrical Conductivity

The fourth clue has been researched somewhat extensively. The electrical conductivity of rocks in the area around earthquakes has been known to change just before shocks. Worldwide studies have been done on the conductivity of water saturated rocks in the lab and in the field. The methods success is encouraging because lab results are consistent for almost everyone who performs experiments. This could be a very helpful clue to forecasting earthquakes.

Patterns of Foreshocks and Aftershocks

In terms of prediction research, the most effort has been put into studying variation in seismicity rate. But, all of the studying has not produced a definate answer. Seismologists study the pattern of aftershocks and foreshocks. In Italy, China, and California, earthquakes have been successfully predicted using a noticed increase in background seismic activity. Tiny shocks increase in number and magnitude leading up to a large shock. If these foreshocks can be recognized as such in enough time to warn people in the necessary areas, many lives can be saved.

Stages of Predictive Characteristics

These five predictors do not all happen at the same time or for all earthquakes but rather, may occur in stages. This theoretical series of events could eventually lead to a somewhat reliable way to predict the occurence of major earthquakes that could cause major death and destruction. The events are still theoretical and need much more data to make them useful enough to use (Based on Zongjin, 1990).

Stage I Elastic strain builds up along a fault due to plate movement: All parameters are at their normal state. No uplift, radon increase, etc.

Stage II Cracks begin to develop in crustal rocks in the pre-quake area. The buildup begins to be visible as an uplift of the area. The cracked rocks do not propogate P-waves as easily and their velocity slows in the area. Radon gas can escape through the newly formed cracks, and electrical resitivity decreases. The newly forming cracks and increasing stress may also result in a tiny increase in local seismicity.

Stage III Groundwater from surrounding areas can now flow into the new cracks. Because the cracks are now filled again, the P-wave velocity can increase back to normal. The ground's uplifing also ceases and radon gas emission decreases. Electrical resistivity is still decreasing.

Stage IV The Earthquake happens.

Stage V This begins as soon as the main shock stops and consists of all the aftershocks.

Bibliography

Bolt, Bruce (1993a). Earthquakes and Geologic Discovery. Scientific American, New York.

Bolt, Bruce (1993b). Earthquakes. W.H. Freeman & Co., New York.

Cipar, John Joseph (1981). Seismic Source Processes and Tectonics: Observations of Four Intracontinental Earthquakes, Ph.D Dissertation. California Institute of Technology, Pasadena California.

Coch, Nicholas (1995). Geohazards Natural and Human. Prentice Hall, New York.

Munro, Ross H (1976). "China Experiences A Powerful Quake". The New York Times, July 28, 1976.

Jiawei, Xu (1993). The Tancheng-Lujiang Wrench Fault System. John Wiley & Sons, Brisbane.

Sullivan, Walter (1976). "The Forces That Bring On Asian Earthquakes--a Long History". The New York Times, July 31, 1976.

Yong, Chen, Kam-ling Tsoi, et al. (1988). The Great Tangshan Earthquake of 1976: An Anatomy of Disaster. Pergamon Press, Beijing.

Zongjin, Ma (1990). Earthquake Prediction. Nine Major Earthquakes in China (1966-1976). Seismological Press, Beijing.