Protection against disasters: Can earthquakes be predicted?

Precise earthquake predictions are impossible. But there are short-term signs that can warn people in earthquake-prone areas to get to safety.

By Stephan Troendle and Lilly Zerbst, SWR

The severe earthquake in the border region between Turkey and Syria has shown that despite scientific advances, it is not yet possible to precisely predict earthquakes. Because there are usually no reliable precursor phenomena for a strong earthquake, says earthquake researcher Torsten Dahm from the German Research Center for Geosciences in Potsdam SWR.

This is how earthquake early warning systems work

There are two main types of waves that occur during an earthquake: the fast primary waves (P waves) that are barely perceptible to humans, and the somewhat slower secondary waves (S waves) that cause destruction during an earthquake. Depending on the distance from the epicenter of the quake, there can be a few seconds to a minute between the arrival of these two waves. The depth of the earthquake also plays a role.

If the sensors of an alarm system detect P waves of a certain strength, they sound the alarm fully automatically. There is often enough time to trigger sirens, stop trains and lift systems, shut off gas lines or switch traffic lights to red before the earthquake really starts. Such systems are used, for example, in Mexico, Japan and on the US west coast.

Early warning using smartphones

It is nothing new that the alarm of conventional early warning systems is now not only transmitted via siren, but also via mobile phone to those affected. But researchers at the University of California, Berkeley have also started trying to use the motion sensors on smartphones to register vibrations across the board. If many cell phones ring at the same time, this indicates an earthquake.

Animal warning systems

Researchers are also studying the behavior of animals to provide a warning within hours or even days before the earthquake – with initial success. In Italy, a notorious earthquake researcher predicted on Friday, April 4, 2009, a major earthquake in the Sulmona region of Abruzzo. The researcher had measured increased leakage of the noble gas radon, which often escapes from the earth before major earthquakes. He had also observed strange behavior by toads that withdrew from their spawning grounds five days before the earthquake.

Italy’s top civil protection chief Bertolaso ​​publicly insulted the scientist that same evening as a complete idiot. Two days later than predicted, the earth shook after all, but in L’Aquila, around 50 kilometers away. The 6.3 magnitude tremor claimed more than 300 lives.

Goats seek cover

Strange behavior also occurs in mammals, for example in goats, which always took cover in a forest a few hours before volcanic eruptions on Mount Etna in Sicily. This was observed by the behavioral biologist Martin Wikelski from the University of Konstanz. He even patented his prediction model.

In order to be able to use animals to detect earthquakes early and reliably, however, a lot of data is still required: “We need one big event after the other that they predict. Because the more often they predict it, the better it gets – like with the weather report.” , according to Wikelski in the SWR.

Earthquake forecasts remain inaccurate

Conclusion: Early warning systems do not provide predictions – they only indicate that an earthquake will come within a very short time. There have been many attempts to predict the earthquakes in the long term – none of them are accurate.

The most reliable method for earthquake prediction is still the mapping of endangered areas. “Today it is possible to set up earthquake networks and thus record all events quantitatively, including small earthquakes,” explains seismologist Dahm. The forecast models can then constantly improve as the amount of data increases and make more precise statements about the frequency of earthquakes in an area.

This allows researchers to say quite precisely how likely it is that an earthquake will occur at a specific location within a specific period of time. However, these periods are still in the order of up to 100 years, according to Dahm.

New branches of research

In the future, linking seismological and geodetic measurements could improve earthquake prediction. This is because survey data can also be used to monitor slow movements along plate boundaries, for example when an area rises over a long period of time. In this way, a build-up of tension before large earthquakes can be seen even when there is no seismic activity. In the eyes of Dahm, this data could be real game changer become. So far, such data have not been taken into account in the models.