Environment Technology

New Time Reverse Imaging Algorithm Means Quicker Tsunami Detection

Tsunami prevention

A new algorithm has been created by seismologists that may be able to offer early tsunami warnings to coastal cities. Current tsunami warning systems use patterns based on previous data per each region and sensors within the ocean that detect irregular movements. The sensors do not help estimate how much water and how strong a specific storm will be.

Australian National University scientists have developed a Time Reverse Imaging Method that takes data in real time from ocean sensors and formulates what a tsunami would look like as it came into contact with the coast. Once scientists can pinpoint where a tsunami will begin, it is much easier to create accurate predictions as to what will occur as it makes its way towards the shore. This new method is promising, because it is not only as fast as currently used algorithms but it is far more accurate.

Seismologist Jan Dettmer from the University says the Time Reverse Imaging Method is far more useful because it is based on real-time information, versus guesses based on past data and is able to do so without sacrificing detection speed. Dettmer and his colleagues are going to discuss their new algorithm at the upcoming Acoustical Society of America meeting, which will be held through May 23rd and 27th in Salt Lake City.

Researchers observed plate tectonics in the Japan Trench in order to build the new algorithm. The crust of the Earth is separated into large plates that float atop the part of the Earth’s core known as the mantle. The plates press against one another, creating large trenches and massive mountains within the past millennia. If movements occur very quickly, an earthquake occurs, and can lead to land traveling several meters. When this happens underneath water, a tsunami is created.

Early detection of tsunamis is extremely important, as on average about 8,000 people die each and every year due to them. Dettmer says once an earthquake begins, we have a mere matter of minutes and this new algorithm offers a much more accurate prediction of the trajectory of a tsunami’s damage. To get a good idea of the course a storm will take, you must known the initial sea surface displacement, which is what a wave looked like when it first begins moving. This is extremely difficult because there are not sensors covering the entire sea floor.

Since there isn’t a complete set of sensors in the ocean at this time, Dettmer chose to view data that was gathered on March 11, 2011, the day of the Tohoku-Oki Earthquake and tsunami. The event information from that day was used to go back in time to mathematically calculate what the tsunami looked like when it first began. He then took this information and added it to sensor data which he then projected to get an idea what the tsunami would look like upon hitting land. He checked his prediction with the actual data of impact to fine-tune his algorithm.

Researchers plan to recreate Dettmer’s backtracking of data in order to prepare the new technology for implementation. The team believes it should be ready within the next five years and will bring a whole new generation to tsunami detection, saving many lives.