NOAA’s tsunami map for the 2004 Indian Ocean tsunami. The map calculates the first-arrival travel times of the tsunami, following their generation at the earthquake epicenter. Image: Wikimedia Commons
On Boxing Day in 2004, I was in a small town called Medewi on Bali’s western coast. We had just left Thailand a few days before, and were spending a few days drying out from a long stint of partying. There’s a great left-hander in Medewi, and at the time, the town was a sleepy little place with no internet to speak of. At just after midnight, the Indian Plate slipped beneath the Burma Plate and spawned one of the deadliest series of tsunamis in recorded history.
An estimated 230,000 people died across fourteen countries, Thailand and Indonesia included. Medewi, and much of Bali, despite its proximity, was unaffected. I have never felt luckier and sadder in my entire life when, a few days later, we checked our emails and discovered the extent of what happened. Life is strange that way.
Although the tsunami occurred hours after the earthquake, almost no one was prepared. The devastation was immense. Due to the fact that there were no tsunami warning systems in the Indian Ocean, the death toll was incredibly high. It’s a difficult thing to set up in a poorer country like Indonesia–the infrastructure required to warn an entire country of an impending disaster is nothing to scoff at. But scientists working at the Australian National University have developed a new system that could give areas in tsunami zones an early heads up.
Right now, geologists use sensors in the ocean that detect earthquakes. They also record underwater landslides–both earthquakes and landslides are causes of tsunamis–but they aren’t an accurate prediction service. The sensors merely report an occurrence, and the prediction of a possible tsunami is so full of variables that any real accuracy is next to impossible. The warning systems are based on previous patterns in a specific area, but they aren’t able to make a real prediction about the size or strength of an incoming tsunami.
The new system is an algorithm that uses a “Time Reverse Imaging Method.” Taking information from ocean sensors already in use, it “can use it to make better predictions about what will happen once the waves reach shore.” According to Newswise, the new algorithm is just as fast as other tsunami prediction systems, but much more accurate.
“The Time Reverse Imaging Method is not based on some guess, it’s based on real-time information,” said Jan Dettmer, a seismologist at the Australian National University. “This method would improve accuracy without sacrificing speed.”
Using the plate tectonics in the Japan Trench, researchers found a way to more accurately predict not only where a tsunami might hit, but how powerful it would be. That’s not an easy thing to do, though. As the saying goes, to know the future, we must look to the past. In order to accurately predict a tsunami, one must know what it looked like when it started. Using the Japanese government’s Pacific Ocean sensors that collected information on the Tohoku-Oki earthquake and tsunami back in 2011, Dettmer and the team used math to go back to the tsunami’s point of origin. But not just where it started–they used the information to create a picture of what the tsunami looked like in its infancy. From there, they added it to the data and made a prediction of what it would look like once it made landfall. Then they compared it to actual results and created their algorithm.
Although there is still much more testing to be done, Dettmer is hopeful that the technology will be ready within the five years. “This is a step forward,” Dettmer said. “This research can be part of the next generation of tsunami warning systems that are based on real time information.”
