The mystery behind earthquakes happening in unexpected regions

The sur­face of the Earth is divided into dif­fer­ent tec­ton­ic plates, upon which the oceans and con­tin­ents rest. Over time these plates move, slide, col­lide or over­lap with on anoth­er. Where the bound­ar­ies of the plates become deformed, pres­sure often builds up to the point that is sud­denly released by a sud­den slip along the faults – this is the well under­stood ori­gin of earth­quakes¹. How­ever, towards the interi­or of tec­ton­ic plates, far from the agit­a­tion of plate bound­ar­ies, in areas that are sup­posed to be calm and free of deform­a­tion, some unex­pec­ted earth­quakes have been occur­ring. They have raised ques­tions in the field seis­mo­logy, where sci­ent­ists have pro­posed new the­or­ies to explain the ori­gin of these earth­quakes. What could be caus­ing this, if not tra­di­tion­al plate tec­ton­ics? Erosion, flu­id cir­cu­la­tion or melt­ing gla­ciers are thought to be triggers…

What exactly are we talk­ing about here? These earth­quakes – recor­ded far from the bound­ar­ies of tec­ton­ic plates – are called intra­plate earth­quakes. “Although they are known on all con­tin­ents, they are quite dis­par­ate and we have only recor­ded them in a few areas of the globe,” describes Éric Cal­ais. In 2017, an earth­quake of mag­nitude 6.5 occurred in Bot­swana, in an area with no tec­ton­ic move­ment as far as we know. In 1811–1812, four major intra­plate earth­quakes (mag­nitude 7 or more) shook the New Mad­rid region along the Mis­sis­sippi River in the United States. Intra­plate earth­quakes have also occurred in cent­ral Aus­tralia, north­ern Sweden and French Guiana… but also in main­land France, an area con­sidered as “intra­plate”². 

In 2019, an earth­quake of mag­nitude 4.9 in the Rhone val­ley sur­prised the sci­entif­ic com­munity. “Before this earth­quake, the seis­mi­city recor­ded was rel­at­ively low: 39 earth­quakes were meas­ured between 1962 and 2018 in this area, all of low mag­nitude (less than 3),” explains Jean-François Ritz. The fault at the ori­gin of this earth­quake was not iden­ti­fied as an act­ive fault either – mean­ing one with the poten­tial to gen­er­ate earth­quakes. It seemed to have been quiet for 20 mil­lion years!

“Intra­plate earth­quakes can reach sig­ni­fic­ant mag­nitudes, but they are much rarer than those loc­ated at plate bound­ar­ies,” explains Éric Cal­ais. Since the first seis­mic record­ings were made in 1974, only about twenty major intra­plate earth­quakes (of mag­nitude 6 or more) have been recor­ded through­out the world³. Older earth­quakes are known from his­tor­ic­al archives or from traces left in the geo­lo­gic­al land­scape. “Anoth­er fea­ture is their fre­quency. At the plate bound­ary, the same earth­quake tends to be repeated at a vari­able fre­quency, for example a few hun­dred years in Cali­for­nia,” con­tin­ues Éric Cal­ais. This repet­it­ive beha­viour is not observed for intra­plate earth­quakes: some seem to be ‘orphan’ earth­quakes on faults that have been inact­ive for mil­lions of years.

A bet­ter under­stand­ing of these earth­quakes is essen­tial: their mod­er­ate to high mag­nitude makes them poten­tially very destruct­ive. They affect regions that are often poorly pre­pared and can be triggered at very shal­low depths. The Le Teil earth­quake – which occurred less than 20 kilo­metres from two nuc­le­ar power plants – was triggered only 1 km under­ground⁴. It ser­i­ously dam­aged 900 build­ings, injured 3 people, and caused an eco­nom­ic loss of sev­er­al mil­lion euros. 

In France, for example, a reg­u­lat­ory zon­ing sys­tem imposes earth­quake-res­ist­ant con­struc­tion stand­ards for build­ings. But it is based on the nation­al seis­mic haz­ard map in which unex­pec­ted earth­quakes like the Le Teil earth­quake are not iden­ti­fied. “In recent years, the sci­entif­ic com­munity has become aware that intra­plate earth­quakes can occur in unex­pec­ted places and that the tec­ton­ic mod­els on which con­ven­tion­al seis­mic haz­ard cal­cu­la­tions are based do not allow this risk to be reflec­ted,” says Éric Calais.

Intra­plate earth­quakes can occur in unex­pec­ted places and clas­sic­al tec­ton­ic mod­els do not allow this risk to be reflected.

Research is act­ive in this area because the implic­a­tions are sig­ni­fic­ant, as a group of French sci­ent­ists write: “This lack of con­sensus […] leads to sig­ni­fic­ant uncer­tain­ties in the char­ac­ter­isa­tion of seis­mi­city and the asso­ci­ated seis­mic haz­ard. For a long time, the sci­entif­ic com­munity thought that plate tec­ton­ics alone explained these earth­quakes. The interi­or of the plates would deform very slightly, at such low rates that it would be impossible to actu­ally meas­ure them⁵.”

But today a com­pletely new explan­a­tion is being put for­ward. Small phe­nom­ena, very rap­id on the geo­lo­gic­al time scale (a few thou­sand years or even tens of years), are thought to be the cause of these earth­quakes⁶. These may be the cir­cu­la­tion of flu­ids: rain that seeps into the sur­face; or gases or liquids that come from the mantle, sev­er­al tens of kilo­metres down, and rise through the tec­ton­ic plate. These increase the pres­sure through the rock to the point of trig­ger­ing intra­plate earth­quakes, as sug­ges­ted for the Bot­swana earth­quake in 2017⁷. 

Oth­er trig­gers include melt­ing gla­ciers and erosion. By light­en­ing the Earth’s sur­face, these phe­nom­ena cause a slight move­ment of the tec­ton­ic plate, sim­il­ar to a rebound. For the Le Teil earth­quake, the melt­ing of gla­ciers at the end of the last major ice age (12,000 years ago) is pro­posed as one hypo­thes­is⁸. “The expert report on the Le Teil earth­quake⁹ con­cludes that it is pos­sible that the extrac­tion of rocks – also light­en­ing the earth’s crust – at a nearby quarry con­trib­uted to the trig­ger­ing of the earth­quake,” adds Jean-François Ritz. How­ever, it is clear that its mag­nitude is explained by tec­ton­ic forces.

“While these small phe­nom­ena are the key to the new the­ory, they are not the only ones at play. Although they are trig­gers of the earth­quake, it is import­ant to under­stand that it is the weak stresses accu­mu­lated – some­times over sev­er­al mil­lion years – that drive it,” adds Éric Cal­ais. Without a trig­ger, these stresses remain pre­served and no earth­quake would occur. On a scale of mil­lions or even tens of mil­lions of years, the move­ments of tec­ton­ic plates fluc­tu­ate: some plates change dir­ec­tion, oth­ers break up, or sta­bil­ise… These ancient deform­a­tions that they under­go con­sti­tute this ‘reser­voir of stresses’ dis­trib­uted across the tec­ton­ic plate. “Our obser­va­tions on the fault that gen­er­ated the Le Teil earth­quake sug­gest that melt­ing ice or erosion can also gen­er­ate reg­u­lar earth­quakes, but at very long fre­quen­cies of the order of 10,000 years¹⁰,” also points out Jean-François Ritz.

The fact remains that intra­plate earth­quakes are few, so it is dif­fi­cult to quanti­fy their beha­viour. Research­ers repro­duce them in numer­ic­al sim­u­la­tions, but it is com­plic­ated to clearly identi­fy the role of each phe­nomen­on (plate tec­ton­ics, erosion, flu­id cir­cu­la­tion, etc.). The fal­lout is sig­ni­fic­ant, as Éric Cal­ais describes, “with these spe­cial earth­quakes, it is very com­plic­ated to cal­cu­late future risks, espe­cially as they can some­times only occur once in a giv­en loc­a­tion. We lack object­ive indic­at­ors to assess future intra­plate seismicity. 

In France, work is con­tinu­ing around the Le Teil earth­quake in search of traces of past earth­quakes. “We can reas­on­ably assume that the fault where this earth­quake was triggered will remain calm for some time, but oth­er faults exist in the area,” con­cludes Jean-François Ritz. “I have no doubt that a reas­sess­ment of the region­al seis­mic haz­ard will be car­ried out in a few years’ time, we need to launch a lot more research and obser­va­tions in this area and review our ideas!”

Anaïs Maréchal