Tsunami Early Warning

by Mark Schrope
ISNS Contributor
March 7, 2008

The world became all too familiar with the devastating potential of tsunamis when the giant wave hit Indonesia and elsewhere in 2004, killing more than 100,000 people. Scientists had the additional discomfort of being reminded how poorly understood these rare disasters remain, because the wave's birth couldn't be forecast ahead of time, nor adequately explained afterward.

On Thursday, at the biannual Ocean Sciences Meeting in Orlando, Fla., Yuhe Song, a researcher at NASA's famed Jet Propulsion Laboratory in Pasadena, Calif., outlined his innovative theory on which types of earthquakes cause tsunamis to form. Song also described newly funded research to develop and test a prototype system aimed at exploiting his theory to provide more reliable, earlier tsunami warnings than currently available. He believes such a system might also prevent false alarms that can be economically damaging and sometimes lead to deadly panics.

For many years the prevailing theory has been that the vertical movement caused by an earthquake is the most important factor in determining if a particular earthquake will cause a tsunami. Proving the theory has been difficult because data on the question are so hard to come by, given the rarity of the events and the fact that the earthquakes that cause 95 percent of tsunamis are typically in remote areas. "The details of the link between an earthquake and a tsunami are not very well understood," says Vasily Titov, chief scientist for the National Oceanic and Atmospheric Administration's Center for Tsunami Research.

After the 2004 Sumatra tsunami, research by British scientists revealed surprisingly little vertical movement at the responsible earthquake's submerged epicenter, despite the massive wave the quake created. Their data ultimately led Song to formulate a theory about how it might instead be horizontal movement of the Earth's crust that plays the greater role in determining when and if an earthquake will spawn a tsunami.

A critical aspect of the theory is that such horizontal movement, even when it occurs far offshore, can be calculated based on movements detectable on shore. Those movements can in turn be measured using the same Global Positioning System (GPS) that enables car navigation devices. So, the Song team began exploring the use of data from an existing network of hundreds of GPS stations around the world. Unlike the typical car GPS, which is accurate to within a few meters or so, the stations are accurate to within a centimeter. Though intended for applications such as tracking airplanes, that precision is also perfect for Song's work.

Once the relevant offshore seafloor movement is calculated, that information can be used to estimate the amount of water that has been jolted around a quake zone, and whether it is sufficient to create a tsunami. The group used this measurement as a basis for developing a new tsunami scale similar to the Richter scale for earthquakes that includes a predicted threshold for tsunami creation.

To test the theory, Song and his colleagues examined three recent events. They looked at the Sumatra tsunami, one that hit Alaska in 1964, and a tsunami false alarm after an earthquake near Nias, in the Indian Ocdean, in 2005. Analysis of the 1964 event, which came well before GPS, relied on data from more primitive technologies available at the time, but followed the same theory. "We use those historical events as a laboratory to test our system," says Song.

Both the tsunamis led to values on Song's scale above the threshold, but the false alarm did not, which is highly significant, because that event caused a panic that killed over 100 people. There have been a number of other false alarms in past years.

Song's team has just received funding from NASA to conduct a study to test the prediction method. The funding will allow a few new GPS stations to be set up, work by the Song team to refine the computer models used to translate GPS data into tsunami predictions, and further work to test the models using historical events and earthquakes that may occur during the study's three-year run.

Song says the overall goal is to show whether the system is reliable enough for use by agencies such as NOAA that monitor tsunamis. "That's the stage we're at now," he says. Comprehensive GPS tsunami monitoring would require more stations than the current NASA system includes, which Song says would cost about $10 million at most. "That's very little money compared to the number of people that have died," he says.

Song believes GPS monitoring would complement other tsunami work, including a global system of buoys that detects the waves after they've been generated, because the GPS system could allow earlier warning. Anything that could provide more advance notice would be helpful, says Titov “It's very promising research, " says Titov of Song's work, "it's definitely additional data that hadn't been looked at before and it may help in many ways to get a more clear picture of the tsunami source." However, he cautions that GPS data would have to be used in conjunction with other technologies and that much work is needed before the method proves itself sufficiently to warrant public use.

Mark Schrope is a freelance science writer based in Florida.

More tsunami images and graphics in the public domain can be found at: http://nctr.pmel.noaa.gov/index.html

***This story is provided free for media use by the Inside Science News Service, which is supported by the American Institute of Physics, a not-for-profit publisher of scientific journals. Please credit ISNS. Contact: Jim Dawson, news editor, at jdawson@aip.org.