Contents lists available atScienceDirect
Earth
and
Planetary
Science
Letters
www.elsevier.com/locate/epsl
Extension
and
stress
during
continental
breakup:
Seismic
anisotropy
of
the
crust
in
Northern
Afar
Finnigan Illsley-Kemp
a,
∗
,
Martha
K. Savage
b,
Derek Keir
a,
c,
Hamish
P. Hirschberg
b,
Jonathan
M. Bull
a,
Thomas M. Gernon
a,
James
O.S. Hammond
d,
J.-Michael Kendall
e,
Atalay Ayele
f,
Berhe Goitom
eaNational Oceanography Centre Southampton, University of Southampton, Southampton, United Kingdom bInstitute of Geophysics, Victoria University of Wellington, Wellington, New Zealand
cDipartimento di Scienze della Terra, Università degli Studi di Firenze, Florence, Italy
dDepartment of Earth and Planetary Sciences, Birkbeck, University of London, London, United Kingdom eSchool of Earth Sciences, University of Bristol, Bristol, United Kingdom
fInstitute of Geophysics, Space Science and Astronomy, Addis Ababa University, Addis Ababa, Ethiopia
a
r
t
i
c
l
e
i
n
f
o
a
b
s
t
r
a
c
t
Article history:
Received19May2017
Receivedinrevisedform2August2017 Accepted9August2017
Availableonline30August2017 Editor: R.Bendick Keywords: Afar seismicanisotropy continentalrifting crustalstress
Studiesthatattempttosimulatecontinentalriftingandsubsequentbreakuprequiredetailedknowledge of crustal stresses, however observational constraints from continental rifts are lacking. In addition, a knowledgeofthe stressfield aroundactivevolcanoescan beused todetectsub-surface changesto thevolcanicsystem.Hereweuseshearwavesplittingtomeasuretheseismicanisotropyofthecrustin Northern Afar, a regionofactive, magma-richcontinental breakup. We combineshear wavesplitting tomography with modelling of gravitational and magmatic induced stresses to propose a model for crustal stressand strainacrosstherift.Results showthatattheEthiopianPlateau,seismicanisotropy is consistentlyoriented N–S. Seismic anisotropy within the rift is generally oriented NNW–SSE, with the exceptionof regionsnorthand south oftheDanakil Depressionwhere seismicanisotropy is rift-perpendicular. These results suggest that the crust at the rift axis is characterized by rift-aligned structuresandmeltinclusions,consistentwithafocusingoftectonicextensionattheriftaxis.Incontrast, we show that at regions within the rift where extension rate is minimal the seismic anisotropy is bestexplainedbythegravitationallyinducedstressfieldoriginatingfromvariationsincrustalthickness. Seismicanisotropyawayfromtheriftiscontrolledbyacombinationofinheritedcrustalstructuresand gravitationallyinducedextensionwhereasattheDabbahuregionweshowthatthestressfieldchanges orientation inresponseto magmaticintrusions.Ourproposed modelprovidesabenchmarkofcrustal stress in Northern Afar whichwill aid the monitoring ofvolcanic hazard. Inaddition weshow that gravitationalforcesplayakeyroleinmeasurementsofseismicanisotropy,and mustbeconsideredin futurestudies.Wedemonstratethatduringthefinalstagesofcontinentalriftingthestressfieldattherift axisisprimarilycontrolledbytectonicextension,butthatgravitationalforcesandmagmaticintrusions canplayakeyroleintheorientationofthestressfield.
©2017TheAuthor(s).PublishedbyElsevierB.V.ThisisanopenaccessarticleundertheCCBYlicense (http://creativecommons.org/licenses/by/4.0/).
1. Introduction
Thetransitionfromcontinentalriftingtoseafloorspreadingisa fundamentalstageofplatetectonics;yetthemechanismsbehindit haveremainedpoorlyconstrained.Manystudieswhichattemptto understand continentalrifting have used numerical models (e.g.,
Brune et al., 2016); analogue models (Corti et al., 2003) or
re-*
Correspondingauthor.E-mail address:f.illsley-kemp@soton.ac.uk(F. Illsley-Kemp).
constructionsof rifted passive margins(e.g.,Pindell et al., 2014). Suchstudiesrequireanunderstandingofhowextensionandstress areorientedacrossacontinentalrift.Inrecentyears,studieshave shown that extension migrates away from rift margins and fo-cusesatrift axes throughmagmatically accommodated extension (Ebingeretal.,2013). However,theorientation ofextension from rift margin to rift axis inthe final stagesof continentalbreakup remains poorly constrained.In addition,continentalriftstypically hostconsiderablevolcanicactivitythatthreatenslocalpopulations (e.g., Hutchison et al., 2016). The stress field that acts on a vol-canohasaprofoundeffectontheorientation offissure eruptions http://dx.doi.org/10.1016/j.epsl.2017.08.014
(Wadge et al., 2016) and any changes in the local stress field aroundthevolcanomayactasanindicatorofanimpending erup-tion (Gerst andSavage, 2004). Therefore an understandingof ex-tensionandtheassociatedstressfieldduringcontinentalriftingis crucialforvolcanichazardassessments.
An ideal location to study the stress field during continental breakupistheDanakilDepressionofNorthernAfar(Fig. 1).Inthis studyweuseshear-wavesplittingoflocalearthquakestomeasure strainandinterpretthestressfield atalate stagecontinentalrift. Thisallowsustogain adeeperunderstanding ofcontinental rift-ing,providingobservationalconstraintsforfuturestudiesofrifting andvolcanichazards.
2. Background 2.1. Tectonicbackground
The Afar Depression lies at the triple junction between the MainEthiopian Rift(MER),theRedSearift andthe GulfofAden (Tesfaye et al., 2003) (Fig. 1). Rifting in Afar first involved the separation of Arabia andAfrica and hasbeen geochronologically datedascommencing
∼
29–31Ma(Wolfendenet al.,2005), with extension oriented NE–SW and initially accommodated on large (>50 km) border faults. Between 25 and 20 Ma, extension in NorthernAfarmigrated away fromtheborder faults andfocused at rift aligned, axial volcanic segments (Wolfenden et al., 2005). Thischangeinextensionmechanismischaracterizedbyafocusing ofseismicityandvolcanicactivityattheriftaxis(Wolfendenetal., 2005).The Danakil Depression is a
∼
200 km long 50–150 km wide basin situated in the northernmost Afar Depression (Fig. 1b). It lies predominantly below sea level but is currently sub-aerial. ThecrustbeneaththeDanakilDepressionthinsdramaticallyfrom∼
27 km in Central Afar to<15
km at the Danakil Depression (Hammond et al., 2011; Makris and Ginzburg, 1987). This is in stark contrast to the Ethiopian Plateau, in the west, which has a∼
38 km thick crust (Hammond et al., 2011). Due to the low elevation of the Danakil Depression, repeated marine incursions meanthesurface geologylargelyconsistsofthick layers of evap-orites (Barberi andVaret, 1970). The NNW–SSE trendingErta-Ale volcanicrangedominatesthedepression,andisthefocusforthe majorityofQuaternarytoRecentbasaltsinAfar(BastowandKeir, 2011). Recent volcanicactivityincludes afissure eruption at Alu-Dalafilla volcanoin2008, which was orientedsub-parallelto the rift axis(Pagli et al., 2012). The Erta-Alerange actsas the mag-maticriftaxiswithintheDanakilDepression;furthertothesouth the rift axis steps en echelon to the southwest to the Dabbahu volcanicsegment.The Dabbahusegment underwentamajor dike intrusion episode from 2005–2010, which consisted of 14 large volumedikes(Belachew etal.,2011;Ebinger etal.,2010;Wright et al., 2006). To the east of the Danakil Depression, straddling the Ethiopian–Eritrean border, is the NE–SW trending Bidu vol-caniccomplex,whichconsistsoftwocalderasNabroandMallahle (Fig. 1b). In 2011 Nabro underwent a major eruption of VEI 4, which killed 7people anddisplaced over 12000 people (Goitom etal.,2015).2.2. Sourcesofcrustalseismicanisotropy
In this study we use local, crustal earthquakes and thus our measurements of seismic anisotropy are strictly limited to the crust. As extension is focused andaligned in the upper crust of the Danakil Depression, stress induced dilatancy of micro cracks will induce seismic anisotropy (Crampin, 1994); such that the anisotropydirection(φ)alignswiththemaximumhorizontalstress (S Hmax) (Hudson, 1981). In an extensional regime the minimum
Fig. 1. TheDanakil DepressionofNorthern Afar. (a) Redcirclesareearthquakes recorded between2011–2013, with horizontalerrors <5km. Blue inverted tri-anglesareseismometersdeployedbetween2011–2013.GPSvelocitiestakenfrom
McCluskyetal.(2010)showtheNE–SWregionalextension. (b) Orangetriangles areactivevolcanoes.Volcanicsegmentsshowninred(Wolfendenetal.,2005). Bot-tomleftinset:TheAfarDepression,locatedatthetriplejunctionbetweentheMain Ethiopianrift(MER),theRedSeariftandtheGulfofAdenrift.Politicalboundaries shownwithblacklines.TopographydatatakenfromNASA’sShuttleRadar Topog-raphyMission.(Forinterpretationofthereferencestocolourinthisfigurelegend, thereaderisreferredtothewebversionofthisarticle.)
horizontal compressive stress (S Hmin) corresponds to the min-imum compressive stress (S3). In contrast, S Hmax corresponds to the intermediate compressive stress (S2). In the absence of
Fig. 2. StructuralmapoftheAfarDepression,showingfaulttracesfromCentralAfar (Manighettietal.,2001)andNorthernAfar.Majorsurface,Pliocene–Recentfaults areinred.Faulttracesareavailableinthesupplementarymaterial.(For interpreta-tionofthereferencestocolourinthisfigurelegend,thereaderisreferredtothe webversionofthisarticle.)
alignedfeatures within thecrust. The alignmentof crustal struc-tures such as fault zones will induce seismic anisotropy that is parallel to these faults (Boness and Zoback, 2006). Faults in the crust of Afar generally trend perpendicular to the orientation of extension,howeverthereare localvariations inthesurface fault-ing,particularlyincentralAfar(Fig. 2).
The alignment of magmatic intrusions and melt within the crust will alsoproduce observable seismic anisotropy (Kendall et al., 2005; Dunn et al., 2005). The Afar Depression has been the focus of repeated dike intrusions and ongoing volcanic activity (Pagli et al., 2012; Wright et al., 2006). It can therefore be ex-pected that these features will affect the seismic anisotropy in
the region. Magmatic activity is focused at the rift axis and is generally aligned with regional S Hmax. Therefore it can be dif-ficult to distinguish between anisotropy induced by the exten-sional stress field and anisotropy due to aligned magmatic fea-tures.
The presence of aligned fabrics within the basement rock of theAfarDepressionmayalsoinfluencetheseismicanisotropy. Ex-posuresofbasementfoliationandProterozoicophioliteformations havebeenusedtosuggestthepresenceofNtoNEorientedsutures withintheEastAfricanbasement(Berhe,1990).Thesefeaturesare notexposedwithintheAfarDepression,andthusinterpretationof basement fabricswithin the AfarDepression must be performed withcaution. However, ifit is assumedthat the N–NEtrend ob-servedelsewhereinEastAfricaalsoexistsintheAfarDepression, then associated seismic anisotropy would align parallel to these features(e.g.,Keiretal.,2011).
Ourstudyuseslocalearthquakesthatareconfinedtothecrust, andthus anymeasurements ofseismic anisotropy originate from the crust. However, knowledge of fabrics in the mantle beneath theAfarDepressionmaybeimportantforanunderstandingof un-derlying fabrics inthe region. Hammond etal. (2014) show that anisotropyinthemantlebeneaththeAfarDepressionisstratified into two layers. In the shallow mantle,there is little anisotropy beneaththeAfarDepression,suggestingthatmelthasno preferen-tialalignment.Inthedeepermantle,anisotropyshowsaconsistent NE–SWorientation,whichisinterpretedtorepresentabroad man-tleflowtowardsthenortheast.
3. Dataandmethods 3.1. Seismicdata
The seismic network was comprised of eleven stations in Ethiopia,active fortwo yearsbetweenFebruary2011 and Febru-ary2013. Thiswas supplemented by a networkof threestations in Eritrea active between June 2011 and October 2012 (Fig. 1a,
Table 1). The resulting combined network consisted of eighteen Güralp CMG-3ESPCD instruments and four Güralp CMG-6TD in-struments that all recorded continuous seismic data at 50 Hz. Earthquakes were picked manually for both P and S waves and eventswerelocatedwithNonLinLoc(Lomaxetal.,2000).Weused a two-dimensionalvelocitymodel basedoncontrolled source ex-periments andreceiver functions(Hammond etal., 2011; Makris andGinzburg,1987).Thisresultedinatotalof4951earthquakes, in a catalogue complete above magnitude 2.0, where magni-tudes arecalculated usinga region-specificlocalmagnitudescale Table 1
Seismicanisotropyresultsforindividualstations(Fig. 4).Circularmeananisotropyorientation()givenindegreesclockwisefromnorth.Standarderror(±)isthecircular standarddeviation.Averagedelaytimeateachstation(δt) giveninseconds.
Station Latitude Longitude Elev. (m) Num. meas. (◦) ± (◦) δt (s) ABAE 13.353 39.764 1447 299 25.1 3.0 0.18 AFMEa 13.204 40.858 58 178 25.9 7.4 0.33 CAYE 14.862 39.305 2435 25 10.8 6.7 0.17 DALE 14.229 40.218 −97 176 −17.7 4.8 0.23 DOLE 15.097 39.981 88 119 50.5 4.6 0.14 ERTEa 13.446 40.497 −5 110 −80.3 2.3 0.09 GALE 13.725 40.394 −87 70 −35.6 7.6 0.27 GULE 13.694 39.589 2025 92 −1.0 3.6 0.19 HITEa 13.101 40.317 566 321 17.2 6.5 0.15 IGRE 12.253 40.461 675 42 88.0 3.2 0.09 KOZE 12.495 40.985 543 29 7.25 9.3 0.06 SAHE 12.040 40.977 365 58 42.9 5.5 0.22 TIOE 14.655 40.867 43 79 27.8 11.8 0.19 TRUE 12.481 40.315 381 8 27.8 3.7 0.14
Fig. 3. Depthdistributionoftheanalyzedearthquakesshowingapredominanceof eventsatdepths<20kmandnocleardependanceofδt with sourcedepth. Seis-micanisotropyresults(Table 1)arethereforelikelytobedominatedbyanisotropy withinthemidtouppercrust.CrustalthicknessvaluesfortheEthiopianPlateauand DanakilDepressiontakenfrom
Hammond
etal.(2011)andMakris
andGinzburg (1987).(Illsley-Kemp et al., 2017). Here we show only those earthquakes withhorizontalerrorslessthan
±
5km(Fig. 1a).3.2. Shearwavesplitting
Oncewe haveacatalogueofmanuallypickedS-waves,we ap-ply the Multiple FilterAutomatic Splitting Technique Version 2.2 (MFAST)developed bySavage etal.(2010). Thismethodisbased on the method of Silver andChan (1991), which minimizes the eigenvalueof thecovariancematrix ofhorizontalparticle motion byperformingagridsearchoverarangeoffastdirections(φ)and delaytimes(δt).
MFAST uses the method of Teanby et al. (2004) and devel-ops it furtherto allow the automatic classification ofthe quality ofthe measurement. MFAST trials aset of 14filters overa win-dowsurroundingtheS-wavepickanddeterminesthe threemost preferablefilters by evaluating signal-to-noise ratioandthe filter bandwidth. These three filters are then used to apply the Silver
and Chan (1991) splitting measurement over multiple windows.
This hasthe advantage ofavoiding the additional subjectivity of window definitions. Cluster analysis is then performed on these measurements in orderto selectthe mostreliable measurement. Furthertotheanalysisofclustering,MFAST rejectsmeasurements from waveforms that have an unclear linearity in incoming po-larization by analyzing the smallest and largest eigenvalues. The rigorousqualityanalysisofMFASTremovesanyunintentionalbias thatisassociatedwithmanualmethods.Forfurtherdetailsonthe MFASTmethod,seeSavageetal.(2010).
WeusetheincidenceanglescalculatedfromNonLinLocinorder torestrict theearthquake raypathswe analyzetothe shear-wave window. The shear-wavewindow is definedby thevertical cone beneaththe seismic station bound by sin−1
(
VS/
VP).
By limiting the incoming raypaths to this window we remove any potential effect of phase conversions at the surface (Booth and Crampin, 1985).Wefindthatduetohighvelocitygradientsnearthesurface, largelyduetosedimentation,ahighproportionoftheearthquakes fulfiltheshear-wavewindowcriteria. The resultingincidencean-Fig. 4. Rose diagrams of seismicanisotropy measurements. Onlymeasurements graded asA qualityaredisplayed. Averageanisotropyorientation()scaled for averagedelaytime(δ)plottedateachstation.
Table 1
givesthenumberof measure-mentsandtheaverageanisotropyorientationsforeachstation.(Forinterpretation ofthecoloursinthisfigure,thereaderisreferredtothewebversionofthisarticle.)gles range from 59.1◦ to 90◦ from horizontal, with an average incidence angle of 75.7◦ fromhorizontal. We choose to only in-terpret splitting measurements that were classed as A-grade by MFAST. This resulted in a total of 888 measurements across the networkfrom636individualearthquakes(Table 1,Fig. 4).Seismic anisotropy is accrued along the raypath of the earthquake, thus measurements are taken ator above theearthquake depth. Over 78%oftheanalyzedearthquakesoccurintheupper20kmofthe crust, thus seismicanisotropy results representthemidto upper crust. The depth distribution of theanalysed eventsis shown in
Fig. 3.
3.3. Shearwavesplittingtomography
To betterunderstand thespatial distribution ofseismic aniso-tropy we use the two-dimensional splitting tomography method of Johnsonet al.(2011). Thismethodassumesthat
δ
t is accrued along the entirepathlength in ahomogeneous sense. It also as-sumesthatthereisnodepthdependenceofanisotropy.Theregion isplacedinan initialgridwhichistheniteratively divided,using aquad-treemethod(TownendandZoback,2004),intoincreasingly smaller blocks basedon the ray pathdensity (Fig. 5a). We place theconstraintthateachblockmustcontainatleast10anda max-imumof65rays.In orderto determine theminimumscale offeatures that we can resolve with our dataset we performed a checkerboard test. Weinputacheckerboardwithalternatinganisotropystrength val-ues of0.01s/km and0.02 s/kmandaddingrandom noisedrawn fromastandardnormaldistribution(Johnsonetal.,2011).Weset
Fig. 5. (a) Theraypathcoveragefortheshearwavesplittingtomography,redcirclesindicateearthquakesthatyieldedshear-wavesplittingmeasurements.Theregionis dividedintoagridbasedonraydensity,withaminimumblocksizeof20 km2.Whitegridblocksdonotcontainsufficientraystoperformcalculation. (b) Checkerboard
residualsfor theshearwavesplittingtomography.Weonlyinterpretregionswitharesidual<0.005 s/kmandhaveexcellentresolutionintheDanakilDepression.(For interpretationofthereferencestocolourinthisfigurelegend,thereaderisreferredtothewebversionofthisarticle.)
theminimumcheckerboardblocksizetobe20km2andperformed
theinversion with thesame quad-tree grid described above. We thenusethe residualbetweentheinput checkerboard and inver-sionoutput toexamine theregions inwhichwe haveresolution.
Fig. 5b shows that we have a residual of less than 0.005 km/s forthemajority ofthe DanakilDepression andwestern rift mar-gin. We only interpret regions within the residual contour of 0.005 km/s.OurfinaltomographyresultsareshowninFig. 6.
3.4.Gravitationalmodelling
Northern Afar features dramatic changes in both topography andcrustal thickness. The Ethiopian Plateau has elevation above 3 kmandcrustal thicknessof
∼
38km,whereas the Danakil De-pressionlies beneathsealevelandhasa crust of∼
15km thick-ness (Hammond et al., 2011; Makris and Ginzburg, 1987). The densityofthecrust in NorthernAfaris poorly constrained, grav-ityobservations(Tiberietal.,2005)suggestthatthedensityofthe crustincreasesfromtheplateautotheriftwhichlikelyreflectsan increaseinintrusivemagmatismwithintherift.Thesevariationsin crustalthicknessanddensitywill inducegravitationalforces that contributetothestressfield.Araragietal.(2015)showedthatthe gravitationalstressaroundMt.Fujiissufficienttoalterthe orienta-tionofseismicanisotropy,itisthereforeplausiblethattheseismic anisotropy in Northern Afar may be affected in a similar man-ner.WeuseanimplementationoftheFleschetal.(2001)method (Hirschberg etal., submitted forpublication), to modelthe verti-callyaveraged, gravity-induced,deviatoricstress fieldin Northern Afar.We modelthedeviatoricstress by solving theforce balance equation,equatingstresstotheforceduetogravity,andseektheminimum stress magnitudesolution,corresponding tothe devia-toric stress due to gravity (Flesch et al., 2001; Hirschberg et al., submittedforpublication).Theforceduetogravityisrepresented bythevertically-averagedgravitationalpotentialenergy(GPE).GPE iscalculatedfromthedensitystructureoftheregionandisgiven by theformula(EnglandandHouseman,1986; Molnarand Lyon-Caen,1988): GPE
=
1 L L −h z −hρ
(
z)
gdzdz,
(1)where
ρ
is the density, g the acceleration due to gravity and htheelevation.L isthethicknessofthelayeroverwhichGPE(and, thus,thedeviatoricstressduetogravity)isaveraged;here,weuse a thicknessof 100 km.We use densities of2760kg m−3 for the
crustattheEthiopianPlateau andDanakilBlock,2890kg m−3 for
thecrust within therift, and3200kg m−3 forthemantle(Tiberi etal.,2005).Thesewere combinedwithmeasurementsofcrustal thickness(Hammondet al.,2011; MakrisandGinzburg, 1987)to determine the densitystructure used inEq. (1). We perform the calculations at 20 km spacing to provide the modelled full ten-sor of the deviatoric stress due to gravity, averaged to a depth of 100 km,including the principal axesand the S Hmax direction (Fig. 7a).
4. Resultsanddiscussion
Theaveragedelay timeof0.19s(Table 1),concentratedinthe upper 20km of the crust (Fig. 3), requires 4% anisotropy, which is consistent with previous measurements from the Afar region
Fig. 6. Shearwavesplittingtomographyresults.Redrosediagramsarenormalized, spatialaveragesofanisotropyorientations,yellowbarsshowthemeananisotropy orientation.Rosediagrams areplottedinthe centre ofeachgridblock (Fig. 5a). Shadedareasindicateregionsoutsideofresolution.(Forinterpretationofthe refer-encestocolourinthisfigurelegend,thereaderisreferredtothewebversionof thisarticle.)
(Keiretal.,2011).However, wedonotseeanyconsistentpattern in distributions ofdelay time (δt), an observation that is consis-tent with other studies (e.g., Gerst and Savage, 2004) and may be due to scattering (Aster et al., 1990). We therefore base our interpretations solely on variations in anisotropy orientation (φ). Seismic anisotropy results show a consistent N–S orientation on the western rift flank and margin (Figs. 4, 6). The orientation of anisotropy changes to a NNW–SSE trend within the Danakil De-pression.In contrast,seismicanisotropy northoftheDanakil De-pression is oriented perpendicular to the rift, NE–SW. Similarly, seismicanisotropyisriftperpendiculartothesouthoftheDanakil Depression in the Dabbahu region. In the following section we presentresults from each of theseregions and discusstheir im-plications in light of gravitational potential modelling, structural geologyandCoulombstresschangemodelling.
4.1. Thewesternriftflankandmargin
Onthe EthiopianPlateau we see aconsistent resultfrom sta-tionsGULE,CAYEandABAE,whichhave
valuesof
−
1.0◦±
3.6◦, 10.8◦±
6.7◦ and25.1◦±
3.0◦ respectively(Fig. 4,Table 1). This north–southorientationofisalsoapparentinresultsfrom split-tingtomography(Fig. 6)andgravitationalmodelling(Fig. 7);which displayaconsistentnorth–south
φ
intheregionbetween13◦ and 14◦ latitudeandhavea residualbetweenthetomographyresults andgravitationalmodellingof<20
◦.Thisorientationisparallelto theOligo-Mioceneborderfaultsandthusmayshowthat structure-induced anisotropy fromthe early stagesof rifting dominatesinthese regions, ratherthan structures associatedwith present-day extension,whichisorientedNE–SW(McCluskyetal.,2010). Simi-larobservationshavebeenmadeintheMainEthiopianrift,where anisotropyawayfromtheriftaxisalignswithinheritedstructures (Keir etal., 2005).Alternatively, the resultsof gravitational mod-ellingshow that deviatoricstressinthe EthiopianPlateauactsas an extensional force and that the associated S Hmax is oriented N–S. It may thereforebe that the anisotropy observationsat the EthiopianPlateaureflecttheextensionalgravitationalforcesinthis region. Alternatively observations may reflect a combination of bothstructuralandgravitationalstressinducedanisotropy.
At the western rift margin, we also observe a clearN–S ori-entationinsplittingtomography
φ
(Figs. 6),howevergravitational forcesinthisareaareminimal(Fig. 7).Recentstudieshaveshown thatextensionisnowfocusedatboththeriftaxisandthewestern rift margin (Ayele et al., 2007; Bastow and Keir, 2011). Detailed studies of anisotropy around active faults have shown that seis-mic anisotropy that is not recorded directly on the active fault, butisincloseproximity, willpreferentiallyalignwith S Hmax, re-gardless of any structurespresent (Boness and Zoback, 2006). If extension at therift margin hadthe sameNW–SE orientation as attheriftaxis,onewouldthereforeexpect thistobe reflectedin the anisotropy alignments. Given that anisotropy at the western rift marginisalignedN–S,thissuggeststhattheextension atthe western riftmargin isorientedE–W.Thiswouldmarka local re-orientation ofthespreadingdirectionfromtheriftaxistotherift margin. A similar rotation of extensional direction has been ob-servedintheMainEthiopianrift(Cortietal.,2013).4.2. TheDanakildepression
Observational results within the rift display a greater varia-tioninboththesinglestationandtomographyresults(Figs. 4,6). Thissuggeststhat morelocalfeatures areaffectingtheanisotropy in thisregion. Generallywithin the DanakilDepression we see a NNW–SSE orientation inboth thesingle station,DALE(
−
17.7◦±
4.8◦)andGALE(−
35.6◦±
7.6◦),andtomographyresults.This ori-entationisparalleltothepredictedS Hmaxfromtectonicextension intheregionandwithsurfacefaulting (Fig. 2)butisin disagree-mentwithgravitationalforces,whichpredictcompression(Fig. 7). The observed,rift-alignedanisotropy occursinthevicinity ofthe Erta-Alevolcanicsegment,whichhasbeenshowntohostan abun-danceofalignedmeltandmagmaticstructures(Paglietal.,2012). Thealignmentofmeltwithinthecrustisaneffectivewayof gen-eratingseismicanisotropy(Dunnetal.,2005),wethereforesuggest thatalignedmeltistheoriginoftheNNW–SSEorientedanisotropy in the Danakil Depression. Thisadds further supportto previous studieswhichsuggestthatextensionwithinAfarisfocused atthe riftaxis(BastowandKeir,2011;McCluskyetal.,2010;Wolfenden etal.,2005).We do, however, see some complexity inthe anisotropy field within the Danakil Depression. At approximately 13.8◦ latitude, thetomographyresultsshowanisotropy measurementsthattrend E–W, whichishighly obliquetothetrendoftherift.These mea-surementslieincloseproximitytotheAlu-Dalafillavolcanic com-plex (Fig. 1); which has experienced significant recent volcanic activity (Pagli etal., 2012). Volcanicprocesses havebeen repeat-edlyshowntoinfluenceandchangetheanisotropyintheirvicinity (e.g., Gerst and Savage, 2004; Savage et al., 2010). We lack the spatial andtemporal resolution to fully investigate whether Alu-Dalafilla is affecting thelocal anisotropy field but suggest that it maybethecauseoftheseanomalousmeasurements.
Weseeabimodalpatterninthestationresultsatthesouthern endoftheDanakilDepression(ERTE,AFME)(Fig. 4).Bothstations haveriftalignedanisotropyandanapproximatelyrift perpendicu-laranisotropy.Inthetomographyresultswe seethatinthe
Giuli-Fig. 7. (a) ThegravitationallyinduceddeviatoricstressfieldintheDanakilregion.ThehightopographyoftheEthiopianPlateaupromotesanE–Worientedextensionand compressionwithintheDanakilDepressionthatrotatesfromE–WattheErta-AlesegmenttoNE–SWatthenorthernDanakilDepression. (b) Residualbetweenthemodelled gravitational
S H
maxandobservedtomographyresults.WeseeexcellentagreementontheEthiopianPlateauandnorthernDanakilDepression(40◦E,15◦N).(Forinterpretationofthecoloursinthisfigure,thereaderisreferredtothewebversionofthisarticle.)
etti Plain, which lies between the Tat-Ale and Erta-Ale volcanic segments (Fig. 1), there is a clear rift perpendicular anisotropy which is accountable for the bimodal distribution at ERTE and AFME (Fig. 6). Surface faulting here shows complex orientations and suggest an interaction between the two volcanic segments (Barberi and Varet, 1970; Barberi etal., 1974) and thismay ac-count for the change in anisotropy orientation. At station TIOE, whichissituatedwithintheDanakilBlock,weseealargeamount ofvariationintheorientationofanisotropywithlittleobvious co-herency.The DanakilBlockhasacomplex andpoorly understood tectonicandstructuralhistory(Fig. 2).Wesuspectthatthis struc-turalcomplexity isthe originofthevariableanisotropy observed atTIOE,howeverfurtherresearchintothestructuresoftheDanakil Blockwouldberequiredtoverifythis.
Further to the north, at a latitude of 15◦, the tomography and
at DOLE (50.5◦
±
4.6◦) show a clear clockwise rotation ofanisotropy towardsa riftperpendicular orientation (Figs. 4, 6). This area hosts Alid volcano, which is poorly studied but is un-derstood tobe active (Duffield etal., 1997).The surface faults in thisregionfollowasimilarNNW–SSEtrendtotheDanakil Depres-sion (Duffield et al., 1997). We thus conclude that the observed anisotropyisnot causedbysuch structures.Extensionrateinthe DanakilDepressiondecreasestotheNorth,suchthatinthe north-ernmostDanakilDepressiontherateofextensionisapproximately zero(McClusky et al., 2010). If the stress due to tectonic exten-sion is minimal, it could be expectedthat the anisotropy would becontrolled bygravitationalstresses.Themodelledgravitational stress in the northern Danakil Depression (Fig. 7a) shows com-pression and an S Hmax oriented approximately perpendicular to therift.Themodelled S Hmaxorientationisinexcellentagreement (residual<20
◦) withtheobservedφ
(Fig. 7b).Thisthereforesug-geststhat inthenorthernDanakilDepression,gravitationalstress, andnottectonicextension,controlsthestressfieldandactsto in-ducerift-perpendicularanisotropyinthisregion.Previousworkby
Araragietal.(2015) hasshownthattheseismicanisotropy atMt. Fuji is influenced by the gravitational stress of the volcano. Our studyisthefirsttofindthatseismicanisotropycanbeinducedby gravitationalforceswhicharisefromcrustalthicknessvariations.
Further to the south (<13◦ Latitude), anisotropy appears to align NNE–SSW atstations TRUE (27.8◦
±
3.7◦) andKOZE (7.25◦±
9.3◦).Inaddition,measurementsatstationHITEthathavecome fromthe southalign NNE–SSW, asshownin thetomography re-sults (Fig. 6). This orientation of anisotropy is sub-parallel with the Alaytavolcaniccomplexandaligns withsurface faulting that isassociatedwithobliqueextension betweentheDanakil Depres-sionandDabbahu volcanic segment.Thissuggeststhat extension inthisareaisnotfocusedNE–SW,asthemajorityofextension in Northern Afar is,but is oriented oblique to the spreading direc-tion,SEE–NWW.Inthissenseextensionistransferredbetweenthe DabbahuvolcanicsegmentandDanakilDepression.4.3. TheDabbahuregion
Furthersouth,intheDabbahu region(Fig. 1),boththe tomog-raphy(Fig. 6)andstationresultsatIGRE(88.0◦
±
3.2◦)andSAHE (42.9◦±
5.5◦)(Fig. 4)suggestanisotropyisorientedobliquetothe rift,NE–SW.Inthisregionwecancompareourresultstoa previ-ous studyof seismic anisotropy by Keir et al. (2011), who used seismicity datafrom2008. Wesee anexcellent agreement atthe Ethiopian Plateau, where Keir et al. (2011) also show anisotropy orientedN–S,howeverintherift,neartheDabbahu volcanic seg-ment,weseeastarkdiscrepancybetweenthetwosetsofresults.Fig. 8. AcomparisonbetweentheanisotropyresultsatstationSAHEin2011(thisstudy)andin2008(Keiretal.,2011).Weobserveaclearchangeintheorientationof anisotropybetweenthetwotimeperiods.
Keiretal.(2011) showmanymeasurements thatarealigned par-alleltothe rift(NNW–SSE),howeverweseeno evidencefor rift-parallel anisotropy in either the single station results (Fig. 4) or the tomography(Fig. 6). The discrepancy inresults ismost clear atstationSAHE, whichwas occupiedduring both studies.Keir et al.(2011) obtainedan anisotropyorientationof
−
12◦,whereas in ourstudyweobtainanaverageorientationof42.9◦.Toensurethat thedifference inanisotropyorientation isnot duetothemethod werecalculatethe2008dataatstationSAHE(Fig. 8).Wecalculate an anisotropy orientation of−
39.0◦ which agrees with the rift-parallelmeasurementsthatKeiretal.(2011)foundintheDabbahu region.TheseismicdatausedbyKeir etal.(2011) wasobtainedfrom 2005–2008, during this time period the Dabbahu volcanic seg-mentunderwentaseriesofdikeintrusions(Belachewetal.,2011;
Ebingeretal.,2010;Wrightetal.,2006).Theinitiationofvolcanic activityatDabbahuwas markedby seismicswarmsandafissure eruptionon26September2005whichwasassociatedwiththe in-trusion of a
∼
60km long dike whichinduced openingof up to 8 m (Wright etal., 2006). From 2005 through to 2010a further 13dikesintruded inthe riftsegment, threeofwhich were asso-ciated withbasaltic fissure eruptions(Ebingeret al., 2010; Ham-lingetal., 2009). Volcanicintrusionshavebeenshowntochange the localseismic anisotropy (Gerst andSavage, 2004; Johnson et al., 2011),so it is possiblethat the Dabbahu dikeintrusionswill have influenced the seismic anisotropy measured by Keir et al. (2011).In order to test this hypothesis we perform Coulomb stress changemodelling usingthe package of LinandStein (2004) and
Todaetal.(2005).Weusethistocalculatetheinducedmaximum horizontalstress(S Hmax)fromtheDabbahudikesequenceandtest whetherthisinfluencedtheanisotropymeasurementsofKeiretal.
(2011). We use constraints on dike geometry and opening from
geodeticandseismicity observations(Belachewetal.,2011; Ham-lingetal.,2009;Wright etal., 2006).Theresults oftheCoulomb
stresschangemodellingshowthattheresultsofKeiretal.(2011), and the recalculated resultat SAHE, can be accurately explained by themodelled,inducedstress field,withan averageresidualin
φ
of27◦±
17◦(Fig. 9).Wethereforesuggestthatthepreviousrift parallel anisotropy orientations fromthe Dabbahu region are not necessarily representativeof thelong termanisotropy,rather the anisotropy results of Keir et al. (2011) represent the stress field around the actively intruding Dabbahu volcanic segment. If this is the case, we can further suggest that the resultspresented in thisstudyrepresentthebackgroundstressfieldintheDabbahu re-gion.Inthissense, thestress fieldintheDabbahuregionchanges orientation episodicallywiththeoccurrenceofmagmaintrusions. Thislendsobservational supporttonumericalmodels which sug-gestthatthestressfieldaroundmagmaticsegmentsincontinental rifts changesbetweenmagmaticandamagmaticperiodsof exten-sion(Beuteletal.,2010).The reason behind the NEE–SWW orientation of anisotropy (IGRE,SAHE)intheDabbahuregionisunclear.Thesouthernextent of theDabbahu volcanic segment extendsinto centralAfar, a re-gionthatispostulatedtobethelocationoftheAfartriplejunction (Tesfayeet al.,2003). The stress field isthereforeexpected tobe extremelycomplexhere.ThesurfacefaultingincentralAfaris gen-erallyorientedNW–SE,consistentwiththeRedSearift(Fig. 2).As the anisotropy measurements do not align withthesestructures, onecaninferthattheanisotropyiseitheroriginatingfromdeeper structural fabrics, orit isrepresentative of thestress field in the region.TheNE–SWorientationoftheanisotropyissub-parallelto thetrendoftheMainEthiopianrift,itmaybethattheanisotropy isthereforeoriginatingfromeitherdeepstructuresincentralAfar related tothe Main Ethiopianrift, orthat thestress field related to Main Ethiopian rift extension iscurrently dominantin central Afar. Alternativelytherift perpendicularorientation ofanisotropy ispredictedbythegravitationalmodelling(Fig. 7),thereforeitmay bethat theanisotropy intheDabbahuregionisinfluencedbythe gravitationalstressduetocrustalthicknessvariations.
Fig. 9. Coulomb stress change modelling of the Dabbahu dikeintrusions from 2005–2008. Small black lines are the modelled S Hmax, thick black lines are
anisotropymeasurementsfrom
Keir
et al.(2011)measuredbetween2005–2008, andtherecalculatedresultatstationSAHE.(Forinterpretationofthecoloursinthis figure,thereaderisreferredtothewebversionofthisarticle.)4.4.Summary
Our results present the first detailed observations of crustal strain in Northern Afar. We use theseobservations to propose a modelfor the stress field in the crust during the final stages of continentalbreakup.Seismicanisotropyattheriftaxisisoriented perpendiculartotectonicextension,suggestingthattheriftaxisis thecurrentlocusforextension. We show that extension changes orientation fromNE–SW at the rift axisto a W–E orientation at the rift margin. We also show that the stress field can change dramatically within the rift, such that north of the Danakil De-pression,gravitationalforcesduetogradients incrustalthickness aresufficienttoinfluencetheanisotropywithinacontinentalrift. Further,throughCoulombstresschangemodellingoftheDabbahu dikeintrusionsweshowthatthestressfieldinthisregionchanges periodicallythroughtimeandthatduringtimesofnoactive intru-sions, the extensional stress field is not observed in the seismic anisotropy.
The observation of rift-aligned seismic anisotropy at the axis ofcontinentalrifts isinagreement withprevious studiesof con-tinentalrifts (Keir et al., 2011) andocean ridge segments (Dunn etal., 2005). We suggest that the anisotropy within the Danakil Depressionoriginatesfromrift-alignedmeltwithinthecrust.This adds further support to the suggestion that magmatic extension focusesattheriftaxisduringcontinentalriftingandthatthis
per-sists through to seafloorspreading (e.g., Wolfenden etal., 2005). However, we demonstratethat the orientation of the stress field away fromtherift axisismorecomplexthan previously thought andisheavilyinfluencedbygravitationalforcesandmagmatic ac-tivity. Our results are amongst the first observations of gravity induced seismic anisotropy and shows that variations in crustal thicknessisan importantconsideration instudiesofcrustal seis-micanisotropyincontinentalrifts.Wealsonotethatgravitational forcesact to inducecompressionwithin the riftsuch that S Hmax isorientedperpendiculartotherift(Fig. 7).Severalstudiesof sur-facegeologyinNorthernAfarhavenotedthepresenceofcross-rift structures(e.g.,Barberietal.,1974)andAfarhostsdikeintrusions andvolcanicchains that donot alignwith S Hmax induced by re-gionalextension (Wadgeetal.,2016).We showthat gravitational forcesduetovariationsincrustalthicknessact toinduceastress fieldthatishighly-obliquetothestressfieldduetotectonic exten-sion.Itmaybethatthisinducedgravitationalforceistheoriginof theobservedcross-riftstructuresandvolcanism.
Inadditiontogravitationalinfluenceonthestressfield,wealso demonstrate that magmatic intrusions can dramatically alter the orientationofthestressfield.WeuseCoulombstresschange mod-ellingofdikeintrusionsattheDabbahuvolcanicsegmenttoshow that previous measurements of rift aligned anisotropy are best explained by the induced stress field from the actively intruding dikes(Keiretal.,2011)(Fig. 9).Seismicanisotropymeasurements presentedinthisstudyarenotinfluencedbyactiveintrusionsand displayarift-perpendicularorientationintheDabbahuregion.This suggeststhat thecurrentstressfieldintheDabbahuregionisnot characterizedbytheregionalNE–SWextensionoftheDanakil De-pression.
5. Conclusions
Wepresentthefirstdetailedstudyoftheseismicanisotropyof theDanakilDepressionofNorthernAfarthroughautomatedshear wavesplitting,shearwavesplittingtomographyandforward mod-elling of both the gravitational and magmatically induced stress fields. This enables us to propose a model forthe crustal stress fieldintheregionandwedrawthefollowingconclusions:
1. Seismic anisotropy on the Ethiopian Plateau is controlled by gravitationallyinducedextensionand/orpreexistingstructures inheritedfrominitialMiocenerifting.
2. Seismic anisotropy within the DanakilDepression, at therift axis,iscontrolledbytheNW–SEorientedextensioninthe re-gion,suggestingthatextension iscurrentlyfocusedattherift axis throughrift-aligned magmaticintrusions andstructures. Thereisalso anincrease invariabilityofanisotropydirection withintherift,implyingthatlocaleffectssuchasvolcanic ac-tivityareaffectingthelocalstressfield.
3. At the actively extending western rift margin seismic aniso-tropyisorientedN–S,obliquetotheinferredextension direc-tionattheriftaxis.Thissuggeststhatthereisarotationinthe extension directionfromthe riftaxistotherift margin,such thatextensionatthewesternriftmarginisorientedE–W. 4. AtthenorthernterminationoftheDanakilDepression,
aniso-tropyrotatestowardsarift-perpendicularorientationwhichis coincidentwithadecreaseinextension rate.Weproposethat thestressfieldduetotheregionalextensionbecomesminimal inthisareaandthatstressinducedbygravitationalforcesnow control anisotropy.Thesegravitationalforces causean S Hmax orientation that is perpendicular to the rift. This is amongst the first observation of gravitational forces inducing seismic anisotropy.
5. Through Coulomb stress change modelling of the Dabbahu dike intrusions we show that the anisotropy south of the
Danakil Depression, in the vicinity of the Dabbahu segment, changes temporally. Previous measurements of anisotropy record the stress field around an actively intrudingsegment, howeversincethe cessationofintrusionsthe stress field has re-orientedtoabackground,riftperpendicularorientation. We envisage that our results will have great importance for numerical and analogue models of continental rifts and for re-constructionsofriftedmargins.Weshowthatthedistributionand orientationofcrustalstressandstrainduringcontinentalriftingis morecomplexthanpreviously thoughtandthat thisisan impor-tant consideration forfuturestudies. In addition,our resultswill aid the monitoring ofvolcanic hazards in Northern Afarby pro-vidingabenchmarkofanisotropymeasurements,againstwhichto testanytemporalchanges.
Acknowledgements
We thankSEIS-UK foruse ofthe instruments andtheir com-puting facilities. The facilities of SEIS-UK are supported by the Natural Environment Research Council (NERC) under agreement R8/H10/64.FIKisfundedthroughNERCstudentshipNE/L002531/1 and a grant to GSNOCS from Roy Franklin O.B.E. BG is funded through a PhD scholarship by the University of Bristol and En-gineering and Physical Sciences Research Council (EPSRC: Grant Number DTG EP/L504919/1). DK is supported by NERC grant NE/L013932. Fundingfor fieldworkis fromBHP Billiton.We also acknowledge assistance from Addis Ababa University and the Afar National Regional State Government. The MFAST package is available fordownload athttp://mfast-package.geo.vuw.ac.nz. The TESSA package is available for download at https://sites.google. com/site/jessicahelenjohnson/tessa.Codesforcalculatingthe gravi-tational deviatoric stress are available at http://github.com/ hamishhirschberg/stress-modelling.
Appendix A. Supplementarymaterial
Supplementarymaterialrelatedtothisarticlecanbefound on-lineathttp://dx.doi.org/10.1016/j.epsl.2017.08.014.
References
Araragi,K.R.,Savage,M.K.,Ohminato,T.,Aoki,Y.,2015.Seismicanisotropyofthe uppercrustaroundMountFuji,Japan.J.Geophys.Res. 120(4),2739–2751. Aster,R.C.,Shearer,P.M.,Berger,J.,1990.Quantitativemeasurementsofshearwave
polarizations at the Anzaseismic network, southern California:implications forshearwavesplittingandearthquake prediction.J. Geophys.Res. 95(B8), 12449–12473.
Ayele,A.,Stuart,G.,Bastow,I.,Keir,D.,2007.TheAugust2002earthquakesequence innorthAfar:insightsinto theneotectonicsoftheDanakilmicroplate.J.Afr. EarthSci. 40,70–79.
Barberi,F.,Bonatti,E.,Marinelli,G.,Varet,J.,1974.Transversetectonicsduringthe splitofacontinent:datafromtheAfarrift.Tectonophysics 23(1–2),17–29. Barberi,F.,Varet,J.,1970.TheErtaAlevolcanicrange(Danakildepression,northern
Afar,Ethiopia).Bull.Volcanol. 34(4),848–917.
Bastow,I.,Keir,D.,2011.Theprotracteddevelopmentofthecontinent–ocean tran-sitioninAfar.Nat.Geosci. 4,248–250.
Belachew,M., Ebinger,C., Coté,D., Keir,D., Rowland, J., Hammond,J.,Ayele, A., 2011.Comparisonofdikeintrusionsinanincipientseafloor-spreadingsegment inAfar,Ethiopia:seismicityperspectives.J.Geophys.Res. 116(B6).
Berhe,S.,1990.OphiolitesinnortheastandEastAfrica:implicationsforProterozoic crustalgrowth.J.Geol.Soc. 147(1),41–57.
Beutel,E.,vanWijk,J.,Ebinger,C.,Keir,D.,Agostini,A.,2010.Formationandstability ofmagmaticsegmentsintheMainEthiopianandAfarrifts.EarthPlanet.Sci. Lett. 293(3),225–235.
Boness,N.L.,Zoback,M.D.,2006.Mappingstressandstructurallycontrolledcrustal shearvelocityanisotropyinCalifornia.Geology 34(10),825–828.
Booth,D.,Crampin,S.,1985.Shear-wavepolarizationsonacurvedwavefrontatan isotropicfreesurface.Geophys.J.R.Astron.Soc. 83(1),31–45.
Brune,S.,Williams,S.E.,Butterworth,N.P.,Müller,R.D.,2016.Abruptplate accelera-tionsshaperiftedcontinentalmargins.Nature 536(7615),201–204.
Corti,G.,Bonini,M.,Conticelli,S.,Innocenti,F.,Manetti,P.,Sokoutis,D.,2003. Ana-logue modellingofcontinental extension:areviewfocused onthe relations betweenthe patternsofdeformationand thepresenceofmagma.Earth-Sci. Rev. 63(3),169–247.
Corti,G.,Philippon,M.,Sani,F.,Keir,D.,Kidane,T.,2013.Re-orientationofthe exten-siondirectionandpureextensionalfaultingatobliqueriftmargins:comparison betweentheMainEthiopianRiftandlaboratoryexperiments.TerraNova 25(5), 396–404.
Crampin, S., 1994. The fracturecriticality ofcrustalrocks.Geophys. J. Int. 118, 428–438.
Duffield,W.A.,Bullen,T.,Clynne,M.,Fournier,R.,Janik,C.,Lanphere,M.,Lowenstern, J.,Smith,J.G.,Giorgis,L.,Kahsai,G.,etal.,1997.GeothermalPotentialoftheAlid VolcanicCenter,DanakilDepression,Eritrea.Tech.Rep.,USGeologicalSurvey. Dunn, R.A.,Leki ´c,V.,Detrick,R.S.,Toomey,D.R.,2005. Three-dimensionalseismic
structureofthe Mid-AtlanticRidge(35N):evidenceforfocusedmeltsupply andlowercrustaldikeinjection.J.Geophys.Res. 110(B9).
Ebinger,C.,Ayele,A.,Keir,D.,Rowland,J.,Yirgu,G.,Wright,T.,Belachew,M., Ham-ling, I., 2010. Length and timescales of riftfaulting and magma intrusion: the Afarriftingcyclefrom2005topresent. Annu.Rev.EarthPlanet.Sci. 38, 439–466.
Ebinger,C.J.,vanWijk,J.,Keir,D.,2013.Thetimescalesofcontinentalrifting: im-plicationsforglobalprocesses.Spec.Pap.,Geol.Soc.Am. 500,371–396. England,P.,Houseman,G.,1986.Finitestraincalculationsofcontinental
deforma-tion,2:comparisonwiththeIndia–Asiacollisionzone.J.Geophys.Res. 91(B3), 3664–3676.
Flesch,L.M.,Haines,A.J.,Holt,W.E.,2001.DynamicsoftheIndia–Eurasiacollision zone.J.Geophys.Res. 106(B8),16435–16460.
Gerst,A.,Savage,M.K.,2004.SeismicanisotropybeneathRuapehuvolcano:a possi-bleeruptionforecastingtool.Science 306(5701),1543–1547.
Goitom,B.,Oppenheimer,C.,Hammond,J.O.S.,Grandin,R.,Barnie,T.,Donovan,A., Ogubazghi,G.,Yohannes,E.,Kibrom,G.,Kendall,J.-M.,Carn,S.,Fee,D., Seal-ing,C.,Keir,D.,Ayele,A.,Blundy,J.,Hamlyn,J.,Wright,T.,Berhe,S.,2015.First recordederuptionofNabrovolcano,Eritrea,2011.Bull.Volcanol. 77(10),1–21. Hamling,I.,Ayele,A.,Bennati,L.,Calais,E.,Ebinger,C.,Keir,D.,Lewi,E.,Wright,T., Yirgu,G.,2009.GeodeticobservationsoftheongoingDabbahuriftingepisode: newDyke intrusionsin2006and2007.Geophys.J.Int. 178(2),989–1003. Hammond,J.O.S.,Kendall,J.-M.,Stuart,G.,Keir,D.,Ebinger,C.,Ayele,A.,Belachew,
M.,2011.ThenatureofthecrustbeneaththeAfartriplejunction:evidencefrom receiverfunctions.Geochem.Geophys.Geosyst. 12(12).
Hammond,J.,Kendall,J.-M.,Wookey,J.,Stuart,G.,Keir,D.,Ayele,A.,2014. Differ-entiatingflow,melt,orfossilseismicanisotropybeneathEthiopia.Geochem. Geophys.Geosyst. 15(5),1878–1894.
Hirschberg,H.P.,Lamb,S.,Savage,M.K,submittedforpublication.Calculationofthe lithosphericand crustalstressandviscosity fieldsinthe NewZealandplate boundaryzone.J.Geophys.Res.
Hudson,J.A.,1981.Wavespeedsandattenuationofelasticwavesinmaterial con-tainingcracks.Geophys.J.Int. 64(1),133–150.
Hutchison,W.,Pyle,D.M.,Mather,T.A.,Yirgu,G.,Biggs,J.,Cohen,B.E.,Barfod,D.N., Lewi,E.,2016.TheeruptivehistoryandmagmaticevolutionofAlutovolcano: newinsightsintosilicicperalkalinevolcanismintheEthiopianrift.J.Volcanol. Geotherm.Res. 328,9–33.
Illsley-Kemp,F.,Keir,D.,Bull,J.M.,Ayele,A.,Hammond,J.O.,Kendall,J.-M.,Gallacher, R.J.,Gernon,T.,Goitom,B.,2017.Localearthquakemagnitudescaleandb-value fortheDanakilregionofnorthernAfar.Bull.Seismol.Soc.Am. 107(2),521–531.
Johnson,J.H.,Savage,M.K.,Townend,J.,2011.Distinguishingbetweenstress-induced andstructuralanisotropyatMountRuapehuVolcano,NewZealand.J.Geophys. Res. 116,B12303.
http://dx.doi.org/10.1029/2011JB008308
.Keir, D., Belachew, M., Ebinger,C., Kendall, J.-M.,Hammond, J.O.S.,Stuart,G.W., Ayele,A.,Rowland,J.,2011.Mappingtheevolvingstrainfieldduringcontinental breakupfromcrustalanisotropyintheAfarDepression.Nat.Commun. 2,285.
Keir, D., Kendall, J.-M., Ebinger, C.J., Stuart, G.W., 2005. Variations in late syn-rift melt alignment inferred from shear-wave splitting in crustal earthquakes beneath the Ethiopian rift. Geophys. Res. Lett. 32 (23).
http://dx.doi.org/10.1029/2005GL024150.
Kendall,J.-M.,Stuart,G.,Ebinger,C.,Bastow,I.,Keir,D.,2005.Magmaassistedrifting inEthiopia.Nature 433,146–148.
Lin,J.,Stein,R.S.,2004.Stresstriggeringinthrustandsubductionearthquakesand stress interactionbetweenthe southern SanAndreasandnearby thrustand strike-slipfaults.J.Geophys.Res. 109(B2).
Lomax, A.,Virieux, J.,Volant, P.,Berge-Thierry,C., 2000.Probabilistic earthquake location in3D and layeredmodels. In: AdvancesinSeismic EventLocation. Springer,pp. 101–134.
Makris,J.,Ginzburg,A.,1987.TheAfarDepression:transitionbetweencontinental riftingandseafloorspreading.Tectonophysics 141,199–214.
Manighetti,I.,Tapponnier,P.,Courtillot,V.,Gallet,Y.,Jacques,E.,Gillot,P.-Y.,2001. Strain transfer between disconnected, propagating rifts in Afar. J. Geophys. Res. 106(B7),13613–13665.
McClusky,S.,Reilinger,R.,Ogubazghi,G.,Amleson,A.,Healeb,B.,Vernant,P.,Sholan, J.,Fisseha,S.,Asfaw,L.,Bendick,R.,Kogan,L.,2010.Kinematicsofthesouthern RedSea–AfarTripleJunctionandimplicationsforplatedynamics.Geophys.Res. Lett. 37(5).
Molnar,P.,Lyon-Caen,H.,1988.Somesimplephysicalaspectsofthesupport, struc-ture,andevolutionofmountainbelts.Spec.Pap.,Geol.Soc.Am. 218,179–208. Pagli,C.,Wright,T.J.,Ebinger,C.J.,Yun,S.-H.,Cann,J.R.,Barnie,T.,Ayele,A.,2012.
Shallow axial magma chamber at the slow-spreading Erta Ale Ridge. Nat. Geosci. 5(4),284–288.
Pindell,J.,Graham,R.,Horn,B.,2014.Rapidoutermarginalcollapseattheriftto drifttransitionofpassivemarginevolution,withaGulfofMexicocasestudy. BasinRes. 26(6),701–725.
Savage, M.K., Wessel, A., Teanby, N.A., Hurst, A.W., 2010. Automatic measure-ment of shear wave splitting and applications to time varying anisotropy at Mount Ruapehu volcano, New Zealand. J. Geophys. Res. 115, B12321.
http://dx.doi.org/10.1029/2010JB007722.
Silver,P.,Chan,G.,1991.Shearwavesplittingandsubcontinentalmantle deforma-tion.J.Geophys.Res. 96(B10),16429–16454.
Teanby, N., Kendall, J.-M., Van der Baan, M., 2004. Automation of shear-wave splittingmeasurementsusingclusteranalysis.Bull.Seismol. Soc.Am. 94(2), 453–463.
Tesfaye,S.,Harding,D.,Kusky, T.,2003.Earlycontinental breakupboundaryand migrationofthe Afar triplejunction,Ethiopia. Geol.Soc. Am. Bull. 115(9), 1053–1067.
Tiberi,C.,Ebinger,C.,Ballu,V.,Stuart,G.,Oluma,B.,2005.Inversemodelsofgravity datafromtheRedSea–Aden–EastAfricanriftstriplejunctionzone.Geophys.J. Int. 163(2),775–787.
Toda,S.,Stein,R.S.,Richards-Dinger,K.,Bozkurt,S.B.,2005.Forecastingthe evolu-tionofseismicityinsouthernCalifornia:animationsbuiltonearthquakestress transfer.J.Geophys.Res. 110(B5).
Townend,J.,Zoback,M.,2004.RegionaltectonicstressneartheSanAndreasfaultin centralandsouthernCalifornia.Geophys.Res.Lett. 31(15).
Wadge,G.,Biggs,J.,Lloyd,R.,Kendall,J.-M.,2016.Historicalvolcanismandthestate ofstressintheEastAfricanRiftSystem.Front.EarthSci. 4.
Wolfenden,E.,Ebinger,C.,Yirgu,G.,Renne,P.,Kelley,S.,2005.Evolutionofa vol-canicriftedmargin:southernRedSea,Ethiopia.Bull.Geol.Soc.Am. 117(7–8), 846–864.
Wright,T.,Ebinger,C.,Biggs,J.,Ayele,A.,Yirgu,G.,Keir,D.,Stork,A.,2006. Magma-maintainedriftsegmentationatcontinentalrupture inthe 2005Afardyking episode.Nature 442(7100),291–294.
