SAN FRANCISCO - In mid-October, the people of San Francisco marked the 20th anniversary of the great earthquake of 1989. That catastrophe took the lives of 63 people in the Bay Area, and injured thousands more. Reconstruction and rehabilitation costs reached into the billions of dollars. Two decades later the residents celebrated, with music and dance performances, deejays in the neighborhoods and barbecued-food stands scattered around the city.

Beyond the festivity, however, the event was aimed at increasing public awareness of earthquakes, which are common in this area. In the legendary earthquake of 1906, approximately 3,000 people were killed in San Francisco. As recently as three weeks ago, a strong quake hit the area, but luckily only 240 homes were damaged by it. In this city, built on the San Andreas Fault, they have come to terms with the fact that the ability to predict earthquakes is not about to become a reality in the near future. Instead, current efforts are focused on minimizing damage.

"There are no two ways about it - it's impossible to predict earthquakes. All the research done on this issue over the past 30 years hasn't led to a breakthrough," says Israel Geophysical Institute director Uri Frieslander decisively. "There are always articles concerning the probabilities of strong earthquakes in various parts of the world, on the basis of quakes that happened in the past, or about small quakes that are felt only by sensitive equipment. We know that in San Francisco, Tokyo or the Dead Sea, earthquakes are to be expected, but no one can say that it's going to happen on February 27, 2010, or any other date."

In the case of Haiti, too, an American study published in 2008 stated that Port-au- Prince was liable, in the near term, to experience a major earthquake of up to 7.2 on the Richter scale. The researchers detected unusual seismic activity in the region and also based their assessment on the fact that no strong earthquakes had been recorded in Haiti since 1770.

Prof. Eric Calais of Purdue University, in Indiana, one of the authors of the study, said this week that his team had the full cooperation of the Haitian government, but its willingness to listen led nowhere. "It's too short of a time frame to really do something, particularly for a country like Haiti," said Calais, in an interview with the Associated Press.

Abysmal failures

Earthquakes occur as a result of fissures in the brittle rocks covering the earth's outermost surface and in its upper crust. The magnitude of a temblor depends on the amount of energy released when the rocks break, under tremendous pressure. Weak quakes occur frequently around the world. But every year, seismographs register about 16 earthquakes of magnitudes greater than 7.0, which have the potential to cause serious damage. Fortunately, however, most of them occur in uninhabited areas or in the oceans that cover the majority of the planet's surface.

The attempts to use such factors as unusual animal behavior to predict strong quakes in densely populated areas have thus far failed abysmally. In March 1997, the prestigious journal Science categorically stated that "earthquakes cannot be predicted."

That statement notwithstanding, in 2002, an Israeli research team investigated a possible connection between radon gas emissions and earthquakes. It reported in the journal Geology that within three days after heightened radon emissions occur in an area, there is a high statistical probability of an earthquake. Nevertheless, the team was unable to distinguish between imperceptible quakes and stronger ones capable of flattening an entire town.

In an interview with Haaretz, Dr. Gideon Steinitz, one of the head researchers of that team, explained the problem with a medical analogy: "We have found the connection between smoking and cancer, but we can't say exactly who is going to get cancer and when."

Minister without Portfolio Benny Begin, who was also part of the team and is former director of Israel's Geological Institute, said recently in a Knesset debate that, "the probability of a strong and fatal earthquake in Israel, like the one in 1927 [which registered 6.2, and killed some 500 people], is steadily increasing."

In Begin's opinion, the best way to prepare for an earthquake is to construct buildings according to proper standards. Dr. Frieslander agrees, stressing that it is possible at least to minimize damage to people and property, "if we build correctly. The building standards have been formulated on the basis of monitoring quake-prone areas. We analyze earthquakes and on this basis, we prepare maps that determine the Israeli standard and determine how to build and what to take into account."

The construction standards designed for potentially dangerous areas are aimed at making buildings capable of absorbing as much as possible of the energy released during a quake. It could have been possible, for example, to reduce the death toll in the strong earthquakes in Turkey in 1999, in which about 17,000 people were killed: According to many testimonies, the houses that collapsed were built of unsuitable materials, with sand and shells in their supporting walls.

"We could design a building with conventional techniques that would survive the largest earthquake without damage," Andre Filiatrault, of the Multidisciplinary Center for Earthquake Engineering Research at the State University of New York at Buffalo, told Scientific American, "but it would be so expensive no one would build it."

With the aim of reducing damage even further, in a number of wealthy Western countries new systems have been installed that warn of a strong earthquake a few seconds before it occurs. Though the efficacy of these systems is in doubt, it is clear that early warning enables the authorities to warn people to leave their houses, to stop subway trains and even to prevent ecological disasters by shutting down nuclear reactors. A system of this sort has already been installed in earthquake-prone areas, such as metropolitan San Francisco, and in the mass-transit systems of Tokyo and Taiwan.

A more encouraging development in the field ensued after the tsunami in the Indian Ocean in December 2004. Then, in the wake of an earthquake of between 9.1 and 9.3 on the Richter scale, giant waves battered 14 different countries and caused the deaths of approximately 230,000 people. Thereafter, an agreement was reached for the establishment of an international system that will inform states of earthquakes that are liable to lead to fatal tsunamis. Such a warning, which can be issued up to several hours before the waves hit the shoreline, will enable extensive evacuation of the population before the disaster strikes. Small signs delineating areas in danger of flooding, and public-address systems, have already been put in place on some of the islands of Thailand.

As for anticipating the earthquakes themselves, Frieslander tries to remain optimistic. "Fifty years ago no one could have predicted the Internet," he says. In the meantime, he emphasizes, the key to reducing loss of life and property lies in construction that enables buildings to better withstand earthquakes.