Crustal Chemistry May Aid in Earthquake Prediction

This is Scientific American 60-Second Science. I'm Christopher Intagliata. Got a minute?
Today's early warning systems for earthquakes give you at most a few minutes to prepare for the hit.
That's cause today's systems rely on detecting the first early rumbles of an actual earthquake before sending the alarm.
Now researchers say that following the chemistry of groundwater could sound a long-term quake alarm.
They tracked samples from an artesian well in Iceland for five years and identified changes in the ratios of hydrogen isotopes and a spike in sodium levels, four to six months before two 5+ magnitude quakes.
The chemical clues suggest mixing between groundwaters,
so the researchers deduce that rocks may be fracturing, linking up underground aquifers, before the quake.
The results are in the journal Nature Geoscience.
Investigations like this one have been plodding along for 40 years, and some studies, like one following the deadly Kobe quake in 1995 have found similar correlations.
But study author Alasdair Skelton, a professor of geochemistry at Stockholm University, says the unpredictable study subject makes it tough to get funding,
"because you can in no way guarantee a result.
So I get three years of money but if there's no earthquake, there's no result."
And even if we do accumulate more results like this and researchers sound a six-month alarm, what next?
"If we're gonna ever predict earthquakes, we want something sort of intermediate term.
Not years, and not minutes or days.
So weeks or months is probably the most useful time scale,
but the sheer practicality of it, what you do about it, I'm thankful I'm not the person who has to resolve that."
Judging by the way leaders have responded to scientists' warnings about another issue (cough cough, climate change),
we might still be left ducking under the nearest table.
Thanks for the minute, for Scientific American 60-Second Science. I'm Christopher Intagliata.