Tsunami forecast models 'too simple', say researchers

Tsunami forecasting could be significantly improved if the computer models used to predict wave behaviour were to incorporate data on the waves' complex behaviour as they come into shore and break, according to researchers.

Writing in a themed issue of Philosophical Transactions of the Royal Society A, the researchers describe the current state of tsunami dynamics and forecasting, and suggest policy changes aimed at better protecting seaside communities.

They note that tsunami warning systems have generally improved, thanks to the deployment of ocean buoys tethered to the seabed, which can detect underwater earthquakes.

"Once the initial data are known, forecasting the propagation of the tsunami across the open ocean is relatively easy," according to one of the paper's authors, Harvey Segur, a tsunami expert at the University of Colorado Boulder, United States.

Currently these data are fed into computer models that incorporate 'shallow water equations' that aim to describe how a tsunami will behave. But these equations are "too simple" to accurately model highly complex wave behaviour once it nears a shoreline, Segur told SciDev.Net.

"Those [computer] codes have been around for 20 or 30 years and [the tsunami modelling community] has not thought seriously about what happens when the waves break," he said.

He welcomed moves within the modelling community to develop equations that are capable of responding to more complex real-world situations, saying it would enable government authorities "to be a lot clearer about who needs to evacuate and who doesn't".

Adrian Constantin, professor of mathematics at the University of Vienna in Austria, says part of the problem is a lack of communication between — on the one hand — mathematicians and fluid mechanics researchers, and planners who handle practical applications of tsunami modelling.

Some South–East Asian city planners still rely on shallow water equations, he said, despite their shortcomings.

"These equations are good far out, but when you approach the beach they are not good enough, one has to do something more refined," Constantin told SciDev.Net.

Segur's co-author, Diego Arcas from the Center for Tsunami Research at the US National Oceanic and Atmospheric Administration, said despite the problems with shallow water equations, they do provide a valuable guide for authorities to respond to a tsunami threat after an earthquake.

"Even though all these errors are present in current tsunami forecasts, they are small enough that most of the time one can forecast the size of the tsunami, its arrival time and decay rate quite accurately, which provides invaluable information to emergency managers," he said.

Link to abstract in Philosophical Transactions of the Royal Society A