Earthquake Effects

The effects of any earthquake depend on a number of widely varying factors. These factors are all of:

  • Intrinsic to the earthquake - its magnitude, type, location, or depth;

  • Geologic conditions where effects are felt - distance from the event, path of the seismic waves, types of soil, water saturation of soil; and

  • Societal conditions reacting to the earthquake - quality of construction, preparedness of populace, or time of day (e.g., rush hour).

One can count the number of deaths caused by large earthquakes to compare the results of all these disparate factors in combination. The Oct. 17, 1989 Loma Prieta earthquake occurred in the least-populated area of the generally urban San Francisco peninsula. Construction standards in the area are relatively high, and the populace relatively prepared. However, soft, highly-saturated soils near San Francisco Bay caused some spectacular failures of large highway structures unusually far away from the event. Even though it was rush hour, many fewer cars were on the roads due to the start of the opening game of the World Series, being played locally. Thus deaths were limited to about 75.

On the other hand, the same year an earthquake of nearly identical energy struck the war-torn country of Armenia, between Russia and Turkey. It was located much closer to the major cities of the region, where poorly-engineered houses of unreinforced concrete collapsed on their occupants during the night. The number of fatalities passed 25,000.

Jan. 17, 1995 Hyogo-Ken Nanbu Earthquake


The worst earthquake catastrophe in years occurred on western Honshu Island early in 1995. More than 5000 people perished in southern Hyogo prefecture, most in the city of Kobe, Japan's most important port. The loss of so many lives, in a country where so much effort had been made to prepare for earthquakes, shocked observers worldwide. However, the magnitude of this catastrophe is probably due to a terrible coincidence of a few simple seismological and societal factors, which may become clear in the photos below.

Tectonic map of southern Japan

(from the USGS) Kobe is located farther than many other cities in Japan from the dangerous intersection of three tectonic plates: the Pacific, Eurasian, and Philippine. This triple junction is a junction of three compressive subduction zones. The red-hatched areas above are the parts of the subduction fault that had aleady broken in great earthquakes in 1944 and 1946. Kobe is also somewhat off the Median Tectonic Line, a zone of strike-slip faults.

1995 2-day Kobe aftershocks map

(from the Earthquake Research Institute, Tokyo) This map shows the epicenters of the earthquake's aftershocks within the first two days afterward. Decades of observation show that the most reliable way to locate the fault that broke in any earthquake is to observe where aftershocks are concentrated. This map shows that the earthquake fault obliquely cut the north side of Awaji Island, and crossed the bay to run along the Honshu coast directly below the city of Kobe. Probably the most important coincidence leading to the mass casualties was this ``direct hit'' of the city by such a large faulting event. There was no intervening distance to mitigate the effects. The Northridge area of Los Angeles suffered a similar coincidence on January 17, 1994.

Direct Effects


There are two classes of earthquake effects: direct, and secondary. Direct effects are solely those related to the deformation of the ground near the earthquake fault itself. Thus direct effects are limited to the area of the exposed fault rupture. Many earthquake faults (such as at Northridge) never break the surface, ruling out direct effects. In the Hyogo-Ken Nanbu event, surface rupture of the fault was observed only in a rural area of Awaji Island, with displacements of up to 3 meters. Few structures were near enough the fault to be damaged by the displacement, although underground utilities, fences, and irrigation ditches were cut. Rice paddies were thrown far out of level.

Aerial view of rupture

(taken by the Geographical Survey Institute of Japan; used by permission.)

Aerial view of the fault rupture on northern Awaji Island, taken on January 18th, the day after the event. From left to right along the rupture, a landslide from the rupture covers a road; a fault scarp across a rice paddy; a right-lateral offset in a dirt road (inset); and three more pointers to the scarp. Note how little damage there apparently is to homes even very close to the fault.

Photo of fault across a rice paddy

(from Bardet et al., 1995; used by permission)

View along the fault scarp on Awaji Island. The section of rice paddy to the right has been uplifted by more than one meter. Note the cut road in the foreground. It is often possible to measure the displacement and length of the exposed fault rupture to estimate the slip and area of the subsurface fault, providing an independent estimate of the earthquake's magnitude.

Fault scarp in rice paddy

Photo from the Japanese edition of Newsweek showing the fault scarp. Note the horizontal as well as vertical offset shown by the dike in the rice field. Well-built structures often escape major damage even so close to a seismogenic fault.

Elastic rebound, the permanent deformation of the ground due to the fault rupture, will extend many kilometers from the fault itself, and is often measurable even where the rupture itself remains buried. In the past geodesists have had to make painstaking and expensive surveys, visiting hundreds of field sites, to measure deformation of an area struck by an earthquake.

Radar interferometry concept diagram
inSAR map of Awaji Island
inSAR map of Kobe

(Analyzed by the Geographical Survey Institute of Japan and the National Space Development Agency of Japan; used by permission.)

Lately planetary geophysicists have developed a quick way to make a map image of deformation using satellite radar interferometry. They compare satellite microwave-radar images of a region taken before and after the event. Any area displaced toward or away from the spacecraft's positions will form an interference pattern, tracing out contours of equal displacement. The map image above of Awaji Island shows eight or more colored interference fringe lines approaching the fault, at 11 cm of vertical displacement per fringe contour, demonstrating almost 1 meter of uplift by the earthquake. At left, two fringes parallel the coast through the city of Kobe, showing about 20 cm of displacement across the city from the buried fault.

Secondary Effects


Most of the damage done by earthquakes is due to their secondary effects, those not directly caused by fault movement, but resulting instead from the propagation of seismic waves away from the fault rupture. Secondary effects result from the very temporary passage of seismic waves, but can occur over very large regions, causing wide-spread damage. Such effects include: seismic shaking; landslides; liquefaction; fissuring; settlement; and the triggering of aftershocks and additional earthquakes.

Secondary Effects - Seismic Shaking


PGV versus distance plot

(from the Architecture Dept. of Tokyo Metro. Univ.)

Since seismic waves spread out from their source just like ripples on a pond, they get weaker the farther you get from the earthquake. The ground acceleration versus distance plot at left shows this effect near Kobe. An acceleration of 1000 galileo for seismic waves in the ground would be equal to the acceleration of gravity; so if it acted straight up it would be able to throw any object into the air. The galileo, abbreviated Gal, is the SI unit for acceleration, cm/s/s. These accelerations are mostly from side-to-side, so even at only 50% of the acceleration of gravity are capable of toppling anything standing. Note that the recorded acceleration can vary by a factor of two or three, especially near the fault.

Kobe ground-motion measurement map

(compiled by the Earthquake Research Institute, Tokyo)

The great differences in secondary effects even between adjacent localities is shown by this map of instrumentally-recorded ground accelerations and velocities. Although the measurements do fall with distance from the epicenter, adjacent sites can vary by more than 50%. Such variation is often caused by variation in the soil conditions.

Seismograms from rock and soft-soil sites

(from the Architecture Dept. of Tokyo Metro. Univ.) The seismograms here were recoded at two different sites near Kobe. On the left are three records of the sharp pulse, lasting less than 15 s, recorded at a station founded in relatively solid rock. On the right are three records of the strong and extended shaking, lasting two or even three minutes, at sites near the coast having soft, thick, water-saturated soils. The geological conditions right at a particular site play a crucial role in the strength, and length, of seismic shaking that can be experienced there. In all earthquakes, low-lying areas having soft, water-saturated soils experience by far the most damage.

Photo of sheared house

(from Bardet et al., 1995; used by permission)

Since most seismic shaking is side-to-side, a shaken structure will undergo shear as this house front in Kobe did. Shear is the bending of right angles to other angles. As it is much more difficult to shear a triangle than a rectangle; effective seismic design requires triangular bracing for shear strength.

Photo of collapsed house

(from Kobe University) This wooden house collapsed during the seismic shaking. It is likely that its heavy roof of ceramic tile created more shear force than its wood frame was built to resist. Tile roofs are popular in Japan.

Photo of collapsed wood next to standing concrete houses

(from Bardet et al., 1995; used by permission)

Behind this completely collapsed wood-frame house is a house of reinforced concrete that suffered no structural damage. The number of wood versus masonry buildings that collapsed in Kobe astonished most observers, as wood-frame structures are usually thought to be much better at resisting shear forces. Possibly the concrete house was better-designed and stronger even for its greater weight. The proportionally heavier tile roofs on wooden houses also might have been a factor.

Photo of building collapse at 5th floor

(from Bardet et al., 1995; used by permission)

Another anomaly was the large number of about 20-year-old high rise buildings that collapsed at the fifth floor. The older version of the code they were built under allowed a weaker superstructure beginning at the fifth floor.

Photo of collapses into street

(from Kobe Univ.) This photo demonstrates the extreme danger of being in the street during seismic shaking. Signs, windows, and the entire fronts of buildings collapsed into the street. During an earthquake, it is usually much safer to Duck, Cover, and Hold On under strong furniture inside than to run out of a building. Exit carefully after the shaking stops.

Secondary Effects - Effects on Lifelines


Photo of collapsed highway support columns

Debris choking streets was just one of the coincidences that made this earthquake so deadly. Almost all utilities, roadways, railways, the port, and other lifelines to the city center suffered severe damage, greatly delaying rescue efforts. Most lifelines in Kobe were constructed 20-30 years ago, before the most modern construction standards were put into practice.

(from Bardet et al., 1995; used by permission) This elevated highway formed an inverted pendulum that the supporting columns were not able to restrain under shear during seismic shaking.

Shear failure of concrete column
Complete concrete column collapse into rubble

(from Bardet et al., 1995; used by permission)

The columns above show a failure typical of somewhat older reinforced concrete structures throughout the world. The vertical steel rods can hold the weight of the structure just fine when that weight is exerted straight down, as usual. During seismic shaking much more steel wound around the rods horizontally can keep the column from breaking apart under the shear forces. Stronger columns are more expensive to build.

Photo of toppled expressway section
Truck on toppled expressway

(from Bardet et al., 1995; used by permission, and from Japanese TV) Large sections of the main Hanshin Expressway toppled over. This was particularly likely where the road crossed areas of softer, wetter ground, where the shaking was stronger and lasted longer.

Photo of suspended rails and ties

(from Bardet et al., 1995; used by permission)

Many elevated structures were simply pulled apart by differential movements, here leaving the welded rails and ties suspended.

Photo of sinking street above collapsed subway

(from Bardet et al., 1995; used by permission)

Below one intersection a subway station collapsed, leaving the road above to sink unpredictably for months until it could be excavated.

Secondary Effects - Fire


Aerial view of the fire
Aerial view of houses burning

(from Japanese TV)

The destruction of lifelines and utilities made it impossible for firefighters to reach fires started by broken gas lines. Large sections of the city burned, greatly contributing to the loss of life.

Map showing Kobe areas devastated by fire

(from the Univ. of Texas) Most of the destruction of San Francisco from the 1906 earthquake was also due to fire. The city installed an entirely independent water system for firefighting, with its own reservoirs. The 1989 earthquake broke a firefighting water main near the Mission Street Post Office, draining the entire system in less than 15 minutes. Fortunately most damage and fires were confined to low-lying districts of the city near the Bay, and fireboats were available to pump bay water as much as one mile inland. Only a few blocks were lost.

Secondary Effects - Liquefaction


Photo of landslide headscarp

One of the reasons that areas of soft, water-saturated soil are hazardous is their potential to liquefy during strong seismic shaking. The shaking can suspend sand grains in waterlogged soil so that they lose contact and friction with other grains. Soil in a state of liquefaction has no strength and cannot bear any load.

(from Bardet et al., 1995; used by permission)

Commonly a soil layer on the side of a hill will liquefy during seismic shaking and flow as a landslide or mudflow, as here.


Photo of sandblow

(from Bardet et al., 1995; used by permission)

A liquefied sand layer can shoot to the surface through cracks, forming a sandblow or sandboil, and depositing a characteristic lens of sand on the ground with a volcano-like vent in the center. With all the material in the layer forced to the surface, the surrounding area sinks unevenly.


Dam collapse from liquefaction

(from Bardet et al., 1995; used by permission)

Entire levees, dams, and other water-saturated embankments can liquefy and flow apart during strong shaking.

Photo of leaning high-rise

(from Kobe Univ.) Buildings founded on liquefied ground will lean or topple.

Collapsed port island margin
Fallen cargo crane

(from Bardet et al., 1995; used by permission) The Kobe port, having been constructed on two artificial islands made of relatively loose fill, and always water saturated, suffered widespread liquefaction and settlement, and was incapacitated for two months. Shipping was disrupted worldwide.

Photo of uplifted pillar foundation
Photo of row of uplifted pillar foundations

(from Bardet et al., 1995; used by permission) On the port islands settlement was so pervasive that any structure built on deep pilings, like this elevated roadway, appeared to have risen a full meter. The world's longest suspension bridge, under construction but having such foundations, was hardly damaged at all.

References


  • J. P. Bardet, University of Southern California, Los Angeles, U.S.A. and F. Oka, M. Sugito and A. Yashima, Gifu University, Gifu, Japan. The Great Hanshin Earthquake Disaster (The 1995 South Hyogo Prefecture Earthquake) Preliminary Investigation Report. February 10, 1995. Incorporating:

    • 1st Report on the Great Hanshin Earthquake by the Field Investigation Team on the Great Hanshin Earthquake, Department of Civil Engineering, Faculty of Engineering, Kobe University, Japan. February 1995.

    • 2nd Report on the Great Hanshin Earthquake by the Field Investigation Team on the Great Hanshin Earthquake, Department of Civil Engineering, Faculty of Engineering, Kobe University, Japan. March 1995.

  • Okimura, T., Takada, S. & Koid, T.H. Outline of the Great Hanshin earthquake, Japan 1995. Nat Hazards 14, 39–71 (1996). https://doi.org/10.1007/BF00229911.