Harnessing Technology For Early Warnings On Devastating Earthquakes
Earthquakes cannot be prevented. But AI-based early warning systems can save lives and infrastructure (Photo: Rajan Journalist/Wikimedia Commons)
On July 30, an earthquake measuring 8.8 on the Richter scale struck Russia’s Kamchatka Peninsula and Japan. Recorded as the sixth-strongest earthquake in history, the tremors were felt beyond Russia and Japan, and as far as Hawaii. Warnings were issued to evacuate an estimated two million people across China, Peru, Chile, Mexico, New Zealand and the Philippines.
Hours after the earthquake’s epicentre was located 120 kms from Petropavlovsk-Kamchatsky, Kamchatka’s capital, there were multiple aftershocks as strong as 6.9 in magnitude.
This was the most powerful earthquake to hit the region since 1952. The earthquake occurred in the Pacific Ocean’s ‘Ring of Fire’, a huge arc where there have been seismic activities in the past.
Even as the scale of the earthquake and the consequent tsunami rekindled memories of the devastation wrought by the 2011 and 2004 tsunamis that had hit Japan and Banda-Aceh in Indonesia, respectively, the threat passed off relatively quickly with minimal loss of human lives and infrastructural and property damages.
While US Geological Survey information showed the earthquake was shallow, Petropavlovsk-Kamchatsky residents reported that the tremors lasted as long as three minutes. Russian officials credited the minimal damage to “solid building construction and the smooth operation of alert systems”. Indeed, UNESCO’s early warning system triggered a tsunami alert in barely ten minutes.
This indicates that early warning systems played an important part in alerting officials not only in Russia but also Japan and other tsunami-affected parts of the globe, to the impending threat.
Two weeks ago, it was reported that Google harnessed motion sensors on more than two billion mobile phones for three years to detect earthquakes before sending automated messages to millions of people across 98 countries. This system led to a ten-fold increase in the number of people who received earthquake alerts.
Reducing losses
The efficacy of such an early warning system is still being examined and refined, but other related studies show that such systems are “increasingly being deployed” as a strategy to reduce losses from massive tremors. A recent study used the global Android smartphone network to develop an earthquake detection system, along with alert delivery and related user feedback framework.
Over the years, scientists’ and researchers’ understanding of locations where seismic activity can be predicted has been described as “excellent”. Besides, many developed countries have been using earthquake-resistant construction and building materials. And yet, earthquakes’ devastating effects were recently felt in Turkey (February 2023) and Morocco (September 2023), where thousands perished.
Technology is the best bet for earthquake early warning, but there are also attendant limits which were evidenced in Turkey over two years ago.
At that time, Google admitted that its system failed to send out accurate alerts that could have given at least 35 seconds lead time to people living within 98 miles of the epicentre to find safe places.
In 2016, a scientific journal posed 20 big questions about the future of humanity. One of them was whether humans will be able to predict natural disasters such as earthquakes with warning times of days or even hours. This question stands unresolved even nearly after a decade.
Research on earthquake prediction has seen its ups and downs, with hope and despair being felt with every big earthquake and with every new technical innovation to study tremors.
Over time, though, such research has become redundant.
Instead, scientists are emphasising on predicting the impact of an earthquake on built-up environments so that the damage can be mitigated by making earthquake-resistant structures and establishing an earthquake-resilient society.
As an offshoot of research on earthquakes and earthquake prediction, early warning can be communicated to places where the destructive waves are expected to reach.
The term “early warning” was coined during the Cold War to describe a military strategy to prevent a potential threat from ballistic intercontinental missiles.
Early warnings were designed to alert target areas as soon as a missile was detected by radar or a launch was discovered by a satellite system. In this context, the term “lead-time” was defined as the time lapses since the detection of a missile and the estimated time of impact on the target.
In earthquake research, the usage of these two concepts arose from the fact that the speed of the earthquake waves (3-5 km/second) is much slower than the speed of present-day communication, which is approximately 300,000 km/second.
For instance, if an earthquake occurs somewhere in the Himalayas, places which are approximately 100-200 km away from the epicentre can be warned 20-40 seconds before the arrival of the damaging waves. There are several variants in such systems, but all of them work on the same principle.
The damage to the built-up environment is more due to the slow-moving earthquake waves, known as shear and surface waves, thus providing even more lead time. Such earthquake early warning systems are in place in several countries, with Mexico and Japan among the first to install these. These systems have worked to “reduce the probability of harm or loss” before the damaging waves hit communities.
The success of such a system depends on, first, how quickly and reliably the waves can be picked up by the sensors. Second, how reliably and rapidly the system can estimate the location and magnitude of the earthquake. And finally, much depends on how people respond when an early warning is issued.
A single weak link in the system can enfeeble the entire chain or even make things worse in some cases. Such systems are not cheap and rely on the availability of communication in earthquake source regions which can be remote, such as in the case of the Himalayan belt.
Google’s early warning system
Google’s global earthquake early warning system is based on a component called accelerometer that can sense motion. And since all users invariably use Google on their phones, such motion can be analysed by it.
In a region where a large number of cellphones show a consistent motion in a short time window, earthquake wave detection algorithms can help estimating the epicentre and the magnitude of a tremor. Warning messages can then be issued immediately to nearby locations, before the tremors begin.
Soon after the Morocco earthquake, Google had announced that its early warning facility that would send out “beware” and “take action” alerts would be launched in India, which has several “severe” seismic regions. However, such a system is not available in India.
Google’s contribution to earthquake early warning is a prime example of technology matching up to nature’s vagaries in ways that can save human lives and infrastructure. Its failure in Turkey and Morocco notwithstanding, technology, with its AI-based variants, is still relatively dependable on ensuring timely earthquake alerts.
