OPINION - Why are earthquakes unpredictable?

By Dr. Judith Hubbard

- The author is a visiting assistant professor at the Department of Earth and Atmospheric Sciences at Cornell University in the US.

ISTANBUL (AA) - Earthquakes cannot be predicted. You hear that a lot - and it is true. But it does not explain why. To understand what makes earthquake prediction currently impossible, you have to understand what an earthquake is. An earthquake occurs when a fault, a crack in the Earth’s crust, slips. Higher amounts of slip over larger areas cause bigger earthquakes. Yet, the whole fault does not slip all at once; instead, it starts at a single point - some location on the fault where the stress being applied to the fault is more than its strength. This happens a lot. Small earthquakes occur all the time; more than 20,000 earthquakes above magnitude 4 are detected globally every year.

So when people ask why earthquakes cannot be predicted, what they really want to know is why "big" earthquakes cannot be predicted. Big earthquakes begin just like small ones but keep growing. Each increment in magnitude is associated with a rupture around 5 times longer. A magnitude 5 earthquake is caused by slip on a fault around 2 kilometers long; magnitude 6 is caused by a fault around 10 kilometers long; magnitude 7 by around 50 kilometers long; magnitude 8 by around 250 kilometers long and magnitude 9 by a fault around 1250 kilometers. Ruptures spread at several kilometers per second, so a magnitude 8 earthquake might take a minute or two. As the earthquake grows in length, its total slip also grows. One side of the fault moves relative to the other side, changing the stress in the crust around it, and the stress applied to nearby faults.

- The (im)possibility of earthquake prediction

Scientists have explored whether it is possible to tell if a small earthquake will grow into a large one based on its early seismic signals. Unfortunately, the answer in a study made in 2016 was no [1]. By looking at a variety of earthquakes from around the world, the authors determined the early seismic signals of small and big earthquakes look same. One of the main things determining whether or not a small rupture will grow is the preexisting stress state along the fault. Over time, a fault will become more stressed due to slow movement of tectonic plates. The stress level depends on the history of slip on the fault - both in earthquakes, and in some cases in slower slip (creep) that does not generate shaking. This stress level can be modified by nearby earthquakes, which cause sudden shifts in the crust. When the stress on a fault is higher than its frictional strength, the fault can slip.

For a large earthquake to happen, the preexisting stress on a large fault area must be close to the frictional strength so that the dynamic stress, caused by the earthquake itself, is enough to push the fault over the edge, cascading progressively along fault for tens or hundreds of kilometers. Unfortunately, it is not possible to measure the preexisting stress state on a fault. Instead, scientists can only estimate changes in stress state - annual increments of tectonic motion, or the impact of one known earthquake on a known fault. Without information about the absolute stress level and the strength of fault, this information is not enough to tell when a fault will break.

- What can the scientists do to predict earthquakes?

We can identify and map active faults and estimate how quickly stress is accumulating. We can also estimate maximum possible magnitude of earthquakes on a given fault. A fault shorter than 200 kilometers, for instance, is too small to produce a magnitude 8.5 earthquake. This information can be supplemented and tested by long-term records of earthquakes on faults from instrumental, historical, archaeological, and geological records. Together, this information can define areas that are more likely to experience high shaking over a given period of time.

We can try to identify periods of time when hazard is higher, but these approaches need further research. Large earthquakes are preceded by smaller events, foreshocks, sometimes. This happens because even small earthquakes can cause local stress changes promoting further ruptures. However, there are many more small earthquakes than large ones - roughly ten times as many with every step down on the magnitude scale. With 1,000 magnitude 4s for every magnitude 7, in more than 99.9% of cases, a magnitude 4 is not a warning of a larger event. In addition, large earthquakes can occasionally trigger other large earthquakes on nearby faults. Modeling stress changes can highlight which faults are more likely to slip following a large event. However, it can be difficult to use this information because triggering can take place within hours, weeks, years, or never.

- How to mitigate the effects of earthquakes?

The best mitigation for earthquakes is not rapid response or earthquake prediction, but rather long-term planning and implementation of building codes. Hazard assessments should take into account local site conditions that can increase or decrease shaking. Planning should also include potential for landslides, liquefaction and tsunamis. By building structures that can withstand these hazards, communities can avoid not just deaths, but also the destruction of homes and livelihoods.

Unfortunately, some people try to take advantage of communities in earthquake-prone areas by claiming they have specialized knowledge or tools allowing them to predict earthquakes, using astrological signals or electromagnetic waves. These false statements create fear and obstruct real efforts at building resilience and preparedness. Earthquake prediction is not currently possible. However, with careful planning and continuous oversight, it is possible to build earthquake-resistant buildings that can withstand the vast majority of events - both the small ones that occur every day, and the rare ones that fundamentally challenge our society.

[1] https://agupubs.onlinelibrary.wiley.com/doi/epdf/10.1002/2016GL070081

*Opinions expressed in this article are the author's own and do not necessarily reflect the editorial policy of Anadolu