The world experiences over 13,000 earthquakes per year reaching a Richter magnitude of 4.0 or greater. But what if there was a way to predict these oft-deadly earthquakes and, through a reliable process, mitigate loss of life and damage to vital urban infrastructures?

Earthquake prediction is the "holy grail" of geophysics, says KAUST's Dr. Sigurjón Jónsson, Associate Professor of Earth Science and Engineering and Principal Investigator of the Crustal Deformation and InSAR Group. But after some initial optimism among scientists in the 1970's about the reality of predicting earthquakes, ushered in by the successful prediction within hours of a major earthquake in China in 1975, several failed predictions have since then moved the pendulum towards skepticism from the 1990's onwards.

In a study recently published in Nature Geoscience by a group of Icelandic and Swedish researchers, including Prof. Sigurjón Jónsson, an interesting correlation was established between two earthquakes greater than magnitude 5 in North Iceland, in 2012 and 2013, and the observed changing chemical composition of area groundwater prior to these tectonic events. The changes included variations in dissolved element concentrations and fluctuations in the proportion of stable isotopes of oxygen and hydrogen.

Can We Really Predict Earthquakes?

The basic common denominator guiding scientists and general observers investigating the predictability of earthquakes is the detection of these noticeable changes before seismic events. Some of these observable precursors are changes in groundwater level, radon gas sometimes coming out from the ground, smaller quakes called foreshocks, and even strange behavior by some animals before large earthquakes.

There are essentially three prevailing schools of thought in earthquake prediction among scientists. There's a first group of scientists who believe that earthquake prediction is achievable but we simply don't yet know how to do it reliably. They believe that we may, at some point in the future, be able to give short-term predictions.

Then there's another class of scientists who believe that we will never be able to predict earthquakes. Their philosophy is that the exact start of earthquakes is simply randomly occurring and that the best thing we can do is to retrofit our houses and make probability forecasts -- but no short-term warnings.

The last group, which currently represents a minority of scientists who are not often taken seriously, believes that earthquakes are indeed predictable and that we have the tools to do it.

Following the wave of optimism in the '70s and '80s, the interest and confidence of scientists in predicting earthquakes have generally subsided, along with the funding. Scientists now tend to focus mainly on understanding the physics behind earthquakes. As Prof. Jónsson summarizes:

"From geology and from earthquake occurrence today we can more or less see where in the world we have large earthquakes and where we have areas which are relatively safe. Although we cannot make short-term predictions we can make what we call forecasts. We can give probabilities. But short-term predictions are not achievable and may never be. We will see."

The Message from the Earth's Cracking Crust

Iceland was an ideal location to conduct the collaborative study undertaken by the scientists from Akureyri University, the University of Iceland, Landsvirkjun (the National Power Company of Iceland), the University of Stockholm, the University of Gothenburg and Karolinska Institutet in Stockholm, and KAUST.

"Iceland is a good testing ground because, geologically speaking, it's very active. It has erupting volcanoes and it has large earthquakes also happening relatively often compared to many other places. And these areas that are active are relatively accessible," said Prof. Jónsson.

The team of researchers monitored the chemistry, temperature and pressure in a few water wells in north Iceland for a period of five years more or less continuously. "They have been doing this to form an understanding of the variability of these chemical compounds in the wells; and then possibly associate significant changes to tectonic or major events," he adds.

Through the five-year data collection period, which began in 2008, they were able to detect perceptible changes in the aquifer system as much as four to six months prior to the two recorded earthquakes: one of a magnitude 5.6 in October 2012 and a second one of magnitude 5.5 in April 2013. Their main observation was that the proportion of young local precipitation water in the geothermal water increased – in proportion to water that fell as rain thousands of years ago (aquifer systems are typically a mix of these two). At the same time, alterations were evident in the dissolved chemicals like sodium, calcium and silicon during that precursor period. Interestingly, the proportion went back to its previous state about three months after the quakes.

While the scientists are cautioning that this is not a confirmation that earthquake predictions are now feasible, the observations are promising and worthy of further investigation involving more exhaustive monitoring in additional locations. But, statistically speaking, it would be very difficult to disassociate these changes in the groundwater chemical composition from the two earthquakes.

The reason why a change in the ratio between old and new water in the aquifer system is important is because it points to the development of small fractures from the build-up of stress on the rocks before an earthquake. So the new rainwater seeps through the newly formed cracks, or microfracturing, in the rocky soil. Prof. Sigurjón Jónsson illustrates this as follows:

"It's similar to when you take a piece of wood and you start to bend it. At some point before it snaps it starts to crack a little; and then poof it snaps. Something similar might be happening in the earth. Meaning that just before an earthquake happens, if you start to have a lot of micro-fracturing you will have water having an easier time to move around in the rocks."

The team will be presenting their findings at the American Geophysical Union (AGU) meeting in San Francisco in December 2014. "It will be interesting to see the reaction there," said Prof. Jónsson.

Related Links:

• Changes in groundwater chemistry before two consecutive earthquakes in Iceland
• Crustal Deformation and InSAR Group

By Meres J. Weche, KAUST News