Q&A
ARETHEREANY
ADVANCESIN
THE PIPELINE?
Theabilitytopredictearthquakes
hasbeenclaimedforallsortsof
things,fromoddanimalbehaviour
andchangesinrockproperties,to
electromagneticsignalsinthecrust,
andso-called‘earthquakelights’
- electricaldischargesthathave
beenobservedintheskybefore
someearthquakes.Nonehave
provedtobeuptothejob.Butthere
maybelightonthehorizon.
ScientistsmonitoringtheCascadia
Fault,offthewestcoastoftheUS
andCanada,havegivena machine-
learningalgorithmtheirdata,with
thehopethatit canfindpatternsin
theseismicrecordthatcouldhelp
predict the fault’snextbigquake.
WOULDACCURATE
PREDICTIONBEAGOOD
THING, ANYWAY?
Whatif weknewthata largemagnitudeearthquake
wasgoingtostrikea majorcity– saySanFrancisco
- inexactlythreemonths’time?Makingsucha
predictionpublicwouldcausewidespreadpanicand
a massexodusfromthecity.Businesseswouldclose
andshifttheirstockout– manyperhapsforgood.
Sharesinlocalcompanieswouldplummet.California
stateauthoritieswouldneedtosetupcampsforthe
hundredsofthousandsfleeingthecity,andfeed
themformonths.Muchbetter,saymany
seismologistsandearthquakeengineers,is toforget
prediction,andfocusinsteadonmakingbuildings
‘life-safe’.Inotherwords,constructthemwell
enoughsothattheydon’tfalldownwhena quake
strikes,andbringoldbuildingsuptoa similar
standard(retrofitting ).Thisway,damageis
minimised,thedeathtollis massivelyreduced,and
the economy and infrastructure little affected.
WILL WE EVER BE ABLE TO PREDICT EARTHQUAKES?
Someseismologistsconsidertrueearthquakepredictiontobeintrinsicallyimpossiblebecausethe
Earth’scrustis ina stateofso-called‘self-organisedcriticality’.If thisis thecase,it wouldmeanthatany
tremor– howeversmall– hassomeprobabilityofballooningintoa majorshock.Othersarguethat
predictionis difficult,butwewilleventuallyfinda way.If predictionis evertohaveanyuseatall,
however,it willhavetobeaccurate,andit willhavetoberighteverytime.Falsealarmswouldleadto
widespreadpublicangeranddisillusionment,andthey could also mean that people ignore the correct
call when it comes, leading to unnecessary deaths.
byP R O FB I L L
MCGUIRE
Billis professor
emeritusof
geophysicaland
climatehazardsat
UCL.Hisnovel,
Skyseed, willbe
publishedin
September.
Aerialviewofthe
SanAndreasFault,
runningthroughthe
CarrizoPlainin
California
Modelshowing
earthquakerisk
forecastforSan
Francisco,where
changesincolour
indicateground
displacement
SanFranciscoliesona
systemofsignificant
faultzones,including
theSanAndreasFault,
meaningthereis a high
risk of earthquakes