Frontiers Physics World  January 2018

Gravitational disturbances could predict earthquake size

Short and sweet Gravitational disturbances could offer a faster way of estimating the size of earthquakes than by using conventional seismic data such as this one from Tohoku. (CC-BY-SA 3.0 / Z22)

Disturbances in the Earth’s gravitational field caused by the 2011 Tohoku earthquake have been spotted in data recorded at the time by a network of seismometers spread throughout East Asia. The signal was identified by Martin Vallée and collaborators at Sorbonne Paris Cité, the French atomic-energy commission (CEA) and the California Institute of Technology in the US. Their analysis provides a faster and more accurate way than conventional methods of estimating the magnitude of large earthquakes.

Usually, the first physical indication of a distant earthquake is received in the form of elastic P-waves, which travel from the rupture site to a seismometer along arc-shaped paths through the crust and upper mantle. These pressure waves typically propagate at 6–10 km/s, meaning that for seismic stations more than 1000 km from the epicentre, several precious minutes can elapse between the earthquake and the arrival of the first direct seismic signal.

However, large earthquakes can rearrange the Earth’s mass in such a way that they can be detected more immediately via perturbations to the gravitational field. What happens is that as P-waves spread out from the ruptured fault, the solid medium is alternately squeezed and stretched, causing transient changes in rock density. Far beyond the primary seismic wavefront, these gravitational effects, which move at the speed of light, can trigger secondary seismic waves that can be picked up by seismometers before the direct waves arrive.

The ground accelerations measured by the gravitationally induced seismic waves in Vallée and colleagues’ data were hard to spot, being barely 1–2 nm/s2 – about 100,000 times smaller than from the P-waves. Another problem was that immediately after the fault slipped, the direct and induced effects of the gravitational perturbation cancelled out, meaning an identifiable signal became apparent only about 60 seconds after the event. The gravitational effect was most easily observable in traces from stations 1000–1500 km from the epicentre, where the P-wave delay was long enough for the signal to emerge before being overwhelmed.

The researchers also simulated the effects of earthquakes of different sizes on the data, and found that the immediate gravitational signal recorded by stations about 1300 km from the epicentre set a lower limit of magnitude 9 for the Tohoku event. In 2011, however, when the earthquake struck, the difficulty of judging magnitude based on the instrumental peak amplitudes measured at nearby stations meant that geologists underestimated its size. Measuring earthquake sizes using the current approach would therefore have allowed Tohoku’s power to have been estimated in minutes, rather than hours (Science 358 1164).

Marric Stephens