They struck suddenly, without any warning. Two earthquakes in the southeast Turkey and north Syria they have led to an indescribable humanitarian tragedy that the world community is watching in shock.

Tens of thousands of people died, many were seriously injured or left homeless, homeless.

A week after Enceladus’ devastating impact, and while quake-hit areas are still shaking from aftershocks, USGS experts warn local residents and rescue teams that they are at serious risk from landslides and liquefaction after the strong earthquakes of last week.

“Why couldn’t we have foreseen this?”

So, at a time when the global community has mobilized to offer support and solidarity to earthquake-affected areas, the big question looms: why couldn’t the scientific community have predicted what was coming?

The East Anatolian faults on which the tremors were recorded are part of the “triple tectonic knot” in which three tectonic plates – Anatolian, Arabian and African – collide with each other.

Doing a little historical review, we find that since 1970, only three earthquakes of magnitude 6 on the Richter scale or higher have occurred in this area. For this reason, many scientists believed that a strong earthquake was only a matter of time.

Why, then, could they not have foreseen this?

How easy is the earthquake prediction process?

The actual scientific process of earthquake prediction very hardas characteristically mentioned by a BBC article attempting to answer the question of whether and to what extent an earthquake can be predicted.

And although the acquisition and processing of data after an earthquake is relatively fast and efficient, collecting data for pre-earthquake predictions is a difficult, complex and huge task..

Chris Marone, an geoscientist at Sapienaz University in Rome and the University of Pennsylvania in Pennsylvania, explains that when simulating an earthquake in the lab, scientists can observe the elements that appear first and indicate that an earthquake is coming. “However, on the contrary, there is a lot of uncertainty in the physical observation of the phenomenon,” he adds.

Geologists are trying to use modern scientific methods for earthquake prediction since the late 1960s, with little success, as the BBC article notes. Marone responds to this “deficiency” by explaining that the complexity of the seismic systems that permeate the planet, combined with dense seismic activity and human factors – that is, the impact of daily human activities on a global scale – make it difficult to pick up clear “signals” when intense seismic activity is imminent.

According to the US Geological Survey, three things are needed for a truly meaningful earthquake prediction: information about the location, time, and magnitude of the shock. Therefore, experts emphasize that none of these three parameters can be determined sufficiently and with certainty to assess and predict an earthquake.

Therefore, geologists conduct their assessments on “model maps”, on which they calculate the probability of an earthquake “strike” over several years.

Prevention possible

And while these tools can be partially used for prevention (see the need for earthquake-resistant buildings), they are not really enough to immediately predict a major earthquake and, accordingly, to take all the necessary protective measures to protect residents. area that will be hit by an earthquake. Simply put, there are currently no tools to take emergency action before an earthquake.

In particular, with regard to strong earthquakes in Turkey and Syria, Marone argues that there were many parameters that made buildings in these areas vulnerable to collapse.. He also adds that anti-seismic regulations were adopted in many Western countries, especially in the 1970s and 1980s, to protect buildings in the event of an earthquake. However, he does not lose sight of the high cost of building and maintaining buildings that meet these anti-seismic rules.

For these reasons, seismologists are trying to find ways and methods to predict earthquakes as accurately as possible. As part of these efforts, researchers rely on a wide variety of data.

Algorithms, artificial intelligence and the behavior of … animals

In particular, Chinese scientists have conducted research to detect fluctuations of electrically charged particles in the Earth’s ionosphere that occur a few days before earthquakes.

Recently, however, there has been optimism about the ability of artificial intelligence to detect those signs that the human factor ignores. In particular, there are algorithms that can, through the analysis of past earthquake data, reveal useful patterns for predicting similar events in the future.

Marone notes that the above models artificial intelligence aroused great interest. In fact, he himself, in collaboration with his colleagues, has been working over the past five years to formulate algorithms that can detect failures and potential errors in laboratory testing and earthquake simulations.

However, the application of this “prediction” ability provided by artificial intelligence in real life is riddled with a number of difficulties and problems.

There are many scientific groups that are trying to “respond” to these challenges through continuous research. A typical example is a group of scientists from China who claim to have observed “swirls” in the electrons in the atmosphere 10 days before the big earthquake in California in April 2010 over the epicenter of the earthquake that followed.

Another Israeli scientific group recently claimed to be able to predict earthquakes up to 48 hours in advance with 83% accuracy by studying changes in electrons in the ionosphere. It appears that China is also investing in this accurate methodology through related research and attempts to predict potential seismic activity through changes in electron accumulation observed in the ionosphere.

Nevertheless, according to experts, there are significant obstacles that cannot make this forecasting technique accurate and completely reliable.

Very often, after intense seismic activity, a rather archetypal argument emerges that animals “feel” an imminent “hit” from Enceladus, showing panic, horror, and violent reactions. However, the scientific use of these observations is probably very difficult, since the behavior of animals does not always give reliable conclusions.

Of course, this did not stop the scientists of the Max Planck Institute from conducting research on cows, sheep and dogs in earthquake-prone areas of Italy. These studies have demonstrated change in animal behavior depending on how close they are to the epicenter of impending earthquakes.

Other scientists are pinning their hopes on processing other data. In Japan, for example, some researchers claim to be able to predict seismic activity based on the accumulation of water vapor over the corresponding seismic zone. Still others try to analyze the microfluctuations of the earth’s gravity before an earthquake.

However, despite all these scientific approaches, no one is able to successfully and accurately predict when and where an earthquake will “hit”.

The Great Dilemma and Optimism for the Future

Marone explains that there is currently no infrastructure to reach this prediction point, emphasizing that it would take a multi-million dollar investment to put together all the tools needed to build models that would allow us to predict an earthquake with the highest possible accuracy.

There is, however, another important issue, according to Marone:Even if we can make more accurate predictions, how will we use the information we have? Until we achieve a high level of accuracy in forecasting models, evacuating entire cities or earthquake-prone buildings can be costly.in case the earthquake prediction model fails.”

Marone remains optimistic, citing similar forecasting models for other natural disasters, such as severe storms or hurricanes, that allow states and civil protection structures to take the necessary preventive measures in time. Marone hopes that something similar can be done in the future with earthquakes, but added that the scientific community is still far from that.

Leaving the question of earthquake prediction for a moment, there is another important area where artificial intelligence can be used. And this applies to the immediate determination of the losses that arose after a natural disaster. That is, a process that can indicate the needs of the affected areas, as well as the action plan of local rescue teams.

By the same logic, the algorithms used can contribute to the prediction of post-seismic activity that follows a strong earthquake. For example, researchers at Harvard study the sequence of aftershocks in the hope of predicting them in time.

Marone concludes by explaining that the scientific community is in a very good position to understand the data that comes after a major seismic event, while at the same time he emphasizes that experts are able to understand whether small enclave “shocks” can lead to a major seismic event. . . . However, he clarifies that in any case, uncertainty remains.

While in the case of Turkey, he notes that it is a rare coincidence of two really strong earthquakes, when one happened very close to the other.

Source: BBC