Geophysicists in the US have developed a new way of calculating the magnitude of an imminent earthquake by making better use of measurements of the compression waves produced early in the event. They say that the technique could be used to create a better early-warning system for earthquakes that could be used worldwide.

The majority of earthquake damage is caused by S-waves, which oscillate perpendicular to their direction of travel through the Earth, and by waves that occur on the surface. However, these are preceded by much faster-moving compression waves (called P-waves) that oscillate in the direction of travel and cause minimal damage. By making careful measurements of arriving P-waves, seismologists can get some idea of the strength of the impending earthquake. While this only gives officials tens of seconds to react, it is enough time to take some protective action such as slowing down high-speed trains, switching off gas mains and even warning the public to seek shelter.

Amplitude, period or both?

Current early-warning schemes make use of two properties of P-waves: the displacement amplitude of the P-wave's vertical component (Pd) and maximum predominant period of the P-wave (τpmax). To understand how best to use these measurements, Huseyin Serdar Kuyuk and Richard Allen of the University of California, Berkeley looked at real-life data recorded from 1992 earthquakes processed by California's real-time Earthquake Alarm Systems. They also looked at 174 earthquakes in California and Japan that have already been used in early-warning calibration studies. The earthquakes they studied varied in magnitude (M) from 0.2–8, with M = 8 being a major earthquake.

The team used these data to test five different methods for calculating Earthquake magnitude. The techniques use either Pd, τpmax or both to make their predictions. Some of the methods are already used in early-warning systems and one is a new technique developed by the researchers. The team found that its new technique – which is based on Pd measurements alone – gave the most accurate and robust estimate of earthquake magnitude. In contrast, methods that used τpmax did not do a good job of predicting the magnitudes of small earthquakes (M < 3). They were also less accurate than Pd-based techniques for larger magnitude events.

Global reach

Kuyuk and Allen's technique differs from the others tested in that it was formulated using Pd data from earthquakes that happened around the world, rather than earthquakes occurring in just one region. This knowledge could potentially now be applied to improve the accuracy of earthquake early-warning networks worldwide.

"The results of this study should further improve the performance of the earthquake early-warning system currently being developed for the west coast of the US," says Elizabeth Cochran of the US Geological Survey, who was not involved in this work. "Accurate magnitude estimates, for example knowing that an earthquake is a potentially damaging M = 6 earthquake rather than a more moderate M = 5, [are] critical for initiating appropriate actions by the public, companies and emergency personnel," she adds.

Other experts express caution about not making use of information derived from τpmax. Mark Hildyard of the University of Leeds in the UK says that while some effort has been made recently to show that amplitude-based methods can outperform their period-based counterparts, he would like to see more effort put into improving techniques that make use of τpmax.

The work is described in Geophysical Review Letters.