Gravity shifts could sound early earthquake alarm
As deep rock shuffles around, an area's gravitational pull changes too. Detecting these blips could provide precious minutes when it comes to tsunami warnings.
Earthquakes can shuffle around huge chunks of the deep Earth. But picking up these signs by measuring the associated transient gravity change might help provide early warnings, new research shows.
Jean-Paul Montagner from the Paris Institute of Earth Physics in France and colleagues examined data collected during the devastating 2011 Tohoku-Oki earthquake off the coast of Japan, and detected a distinct gravity signal that arose before the arrival of the seismic waves. They published their work in Nature Communications.
And while the technology to employ their system is not yet set up, they say the technique may herald new developments in early warning systems for earthquake hazards such as tsunamis.
Earthquakes are notoriously hard to predict. When a fault line ruptures, seismic waves travel through and around the Earth and these are usually the first sign that at earthquake has hit.
And even though these waves travel quickly – the fastest, P-wave or primary waves, can barrel through the Earth at 13 kilometres per second – they still mean precious seconds or minutes before the waves arrive at a seismic station.
Montagner and his crew thought there could be a way to detect an earthquake before the waves appeared.
Seismologists have known for more than a decade that there are static gravity changes following a rupture. This happens because as a fault line moves around, mass is redistributed below the surface. This means some areas suddenly become less dense while others pack on mass – and so their gravitational pull changes too.
Such changes are measured with gravimeters. The problem is there's background noise when it comes to gravity changes – the dynamic Earth constantly shifts and wriggles. Could the sudden gravity signal associated with an earthquake be teased out from the underlying noise?
To find out, the researchers needed to examine a large earthquake that happened close enough to a sensitive gravimeter, so small changes in the gravity field could be picked up, but far enough away so the P-waves didn't immediately reach seismic sensors.
They found an ideal example in the 11 March Tohoku-Oki earthquake that led to the Fukushima Nuclear Power Plant disaster.
Some 500 kilometres from the earthquake's epicentre was a gravimeter at the Kamioka Observatory. The observatory was surrounded by five seismic stations. P-waves from the earthquake took around 65 seconds to reach the stations.
Montagner and his colleagues first "calibrated" their statistical technique with 60 days of background gravity measurements – from 1 March 2011 to 5.46am on 11 March (21 seconds before the earthquake rumbled), then from 12 March to 30 April.
They compared this background with measurements taken during the earthquake and shortly thereafter, and found a distinct blip at the time of the earthquake. It was small, but strong enough to be distinguished from the background with 99% confidence.
So can this prediction technique be implemented today? Unfortunately not – it would require building a substantial network of exceptionally sensitive gravimeters which don't yet exist. But, the researchers write, they could have the potential to let seismologists estimate earthquake magnitude quickly – a process that currently takes up to several minutes.