A surprise earthquake struck an area in Greece long considered to be "aseismic." In the aftermath, scientists and government officials are taking a new look at how earthquake risk is calculated.
by Stathis C. Stiros, Institute of Geology and Mineral Exploration, Athens, Greece
A strong earthquake with surface wave magnitude 6.6 hit the small town of Grevena in northwest Greece on May 13, 1995. This unexpected earthquake surprised the seismological community: it damaged more than 5,000 houses and caused more than half a billion dollars damage, but no deaths. This event was just one of the frequent, strong (M=>6.0) earthquakes that hit Greece but cause relatively few deaths. There are two main reasons for this phenomenon: houses in most parts of Greece are designed to withstand earthquakes, and epicenters of most shocks are in the sea. Therefore seismic waves are attenuated by the time they reach inhabited areas.
The Grevena earthquake was even more remarkable: its epicenter was inland, close to inhabited areas where most of the houses were old and poorly equipped to withstand an earthquake. Fortunately, most people were outdoors when the earthquake struck and caused buildings to collapse. Some were simply outside to enjoy a pleasant spring Saturday morning. Others left their homes after a foreshock warned of the impending earthquake.
The Grevena earthquake was also a surprise in that it hit an area that had been considered almost completely aseismic. In certain parts of Greece, no evidence of historical or 20th century seismicity exists (see figure 1). The seismicity level of some regions is thought to be far lower than that of adjacent, highly active ones. Scientists have come to regard some areas of Greece as rigid blocks surrounded by wide regions of active faulting and seismicity—oases in a desert of earthquake-prone land.
Fig. 1. Epicenters of shallow (0–50 kilometers deep) earthquakes with magnitude larger than 5 in Greece between 1965 and 1995. A star marks the epicentral area of the 1995 Grevena earthquake. Based on data compiled from various sources by D. Fountoulis at the Hellenic Antiseismic Protection Organization.
In spite of the abundance of exposed young Quaternary faults, the seismic risk in these low-seismicity areas was thought to be practically nonexistent (see figure 2). Dams in these areas were constructed without special care, and they were once thought to be ideal locations for nuclear power plants!
Fig. 2. Current seismic zoning in Greece. Zones 1 and 4 correspond to the minimum and maximum seismic risk, respectively. Seismic coastal uplifts are marked with triangles. A star denotes the epicentral area of the 1995 Grevena earthquake; solid triangles indicate the seismic coastal uplifts discussed in the text. The Grevena earthquake, as well as others deduced from coastal (Euboea, Thessaly) and archaeological data (Pella), can be identified with hitherto assumed aseismic zones.
Fossil shorelines suggest that a locale experienced either significant climate change that caused sea level to fluctuate or seismic uplift at some time in the past. The fossil shorelines at Euboea and Thessaly most likely were created by seismic uplift. First, fauna remains are present in these shorelines below the midlittoral zone, where bioerosion is intense. These remains could only be preserved if the fauna crossed this last zone rapidly and arrived at or above the supralittoral zone, where bioerosion is rather negligible. Reaching this layer at such speed requires seismic uplift. Second, there is direct evidence of tectonic effects: the fossil shorelines are internally deformed, regionally discontinuous, and they have different uplift and subsidence histories.
Archaeological excavations have also uncovered evidence of destructive paleoseismic events in Greece. Pella, the capital of ancient Macedonia, was destroyed by an earthquake at around 90 B.C. A number of buildings collapsed in a flat area several hundred meters wide, killing and burying people. No signs of fire, looting, or other anthropogenic causes for this destruction have been documented. Other effects, such as foundation instability, could not account for the widespread destruction and the subsequent abandonment of a large part of the city. An unrecorded earthquake is the only possibility. Pella is not far from the epicenter of the 1995 Grevena earthquake, and it falls within the same zone of low seismic risk (see figure 2).
Determining the possibility of future earthquakes is an extremely difficult task, not only in North and South America, where the local historical seismicity record barely covers a period of 150–200 years, but also where the historical record is at least thousands of years longer, and the pattern of seismicity is assumed to be known. Society presses for successful estimates of the seismic risk, but on which principles and methodology can these be calculated?
The Grevena earthquake proves that a lack of fatalities does not necessarily suggest weak earthquake magnitude. This is an important consideration to keep in mind when evaluating the 2500-year record of seismicity in Greece and other areas. In some parts of Greece major earthquakes strike previously seismic areas within some tens or hundreds of years and can be described by conventional seismological methods. In other areas, however, earthquakes are very infrequent, and the above methods are not useful for estimating risk. Or, even worse for calculating earthquake risk, the distribution of seismicity in time is not uniform, and periods of high seismicity alternate with quiescent periods.
One solution to this problem is to find a way to enlarge the seismicity record, and consequently to obtain better estimates of the probabilities of future earthquakes. Scientists from different countries have succeeded in this task recently and have developed various techniques for determining an area's seismic history. These are geological, geomorphological, biological, archaeological techniques that can be used to identify past earthquakes and study their parameters from their traces on the natural environment and human constructions. In spite of their progress and success, paleoseismological techniques have not been fully adopted by most agencies responsible for evaluating seismic risk in various countries, including Greece. The recent Grevena earthquake underscores this problem; hence, it may prove a benchmark for earthquake risk studies on a global scale.
Source: Eos, December 12, 1995, p. 513.
Return to Science and
Society
Return to Starting
Point
