Project D escription
Synopsis. We propose a portable broad band seism ic d eploym ent that extend s from the
Paleozoic terranes of Maine to the Archean-age Superior craton of central Québec and
that focuses especially upon three prom inent tectonic terrane bound aries. Our project
has both a science com ponent and a p ublic service com ponent. The science com ponent
w ill test three hypotheses related to continental evolution: 1) that the m ajor terrane
bound aries have a d eep seism ically-im ageable expression that cuts across the pattern of
preexisting structures (e.g. truncation of layering, steps in lithospheric thickness); 2) that
the lithosphere-asthenosphere bound ary (LAB) d eepens tow ard s the center of the
craton and that com m only-em ployed seism ic m ethod s give m utually-consistent
estim ates of its d epth; and 3) that the pattern of azim uthal anisotropy in northeastern
N orth Am erica is prim arily an asthenospheric signal associated w ith present -d ay
m antle flow . The p u blic service com ponent extend s EarthScope coverage into an
im portant and und er-sam pled region of N orth Am erica, enabling it to better fulfill its
Science Plan.
Motivational Introduction. Our project is d esigned to ad d ress three of the prim ary scientific
targets id entified in the Earth Scope Science Plan: a) im aging of the N orth Am erican crust and
lithosphere; b) continental evolution through geologic time, and c) d eep Earth structure and
d ynam ics.
N orth Am erica spans a large range of tectonic features, includ ing the strike slip and subd uction
plate bound aries of the w est coast, the rifting w ithin the Basin and Range, the arc volcanism of
the Cascad es, the hotspot volcanism of Yellow stone, the orogenic belts of the Appalachians,
Laram id es and Ouachitas, and the central cratonic core of the continent (w hich itself is a
collection of Archean and Proterozoic terranes w ith a com plicated history). The Earth Scope
Transportable Array (TA, see Figure 1) covers all of the features to som e degree, but som e m uch
better than the others. Arguably, one of the poorest covered is the cratonic core, ow ing to m ost
of it being centered in Canad a, beyond the northern bord er of the USArray footprint. Such a
state of affairs is ironic, for as the old est part of the continent and the part w ith the thickest
lithosphere, the cratonic core is fund am ental to a com plete und erstanding of N orth Am erica.
Figure 1. Shear wave seismic velocity at
100 km depth in the surface-wave
tomography model of N ettles &
Dziewonski (2008) illustrating key
aspects of the lithosphere of the N orth
A merican continent: seismically slow
western tectonically active part, faster
values
and
small-scale
regional
variations within the northern and
eastern stable regions. The USA rray TA
grid (circles, status on June 22, 2011)
will largely miss the fastest regions of the
continental lithosphere. The proposed
array (white line) crosses a region of
progressive change in wave speed, from
C-1
very high to moderate, sampling structures associated with major continent-forming episodes. It will avoid a region
of proposed lithospheric indentation beneath the A tlantic coast and southeast Canada (low values left of “V s” label
on the plot) that is more complex and possibly related to the M esozoic impact of a hot spot.
By saying this, w e d o not w ant to im ply that no inform ation about cratons can be gained from
stud ying d ata from the TA as it is presently configured . Quite to the contrary, the TA covers
the Proterozoic section of the craton, w hich extend s d ow n from southern Canad a through the
central US as far south as Arkansas, quite w ell. And its coverage of the Archean -age Wyoming
craton, w hich is centered beneath Wyom ing and southern Alberta, is pretty good , too. Many
im portant insights w ill no d oubt be gained by stud ying each. But neither of these regions is
id eal for investigating the suite of questions that are im portant for und erstanding the
fund am entals of cratonic structure and evolution. The Proterozoic region of the US is too young
and lacks the extrem ely thick lithosphere of the central craton. The Wyom ing craton, w hile
Archean, has been heavily m od ified by the Laramid e d eform ation and the Yellow stone hotspot,
and probably lost m uch of its original thick lithosphere. In contrast, the cratonic core of N orth
Am erica centered in Québec and northern Ontario has som e of the thickest lithosphere globally,
and has been m ore-or-less tectonically quiet since the Proterozoic Grenville orogeny, at ~1.0 Ga.
We believe und erstanding it is essential for understand ing the events that subsequently shaped
the passive continental m argin of eastern N orth Am erica, w here accretion of m aterial onto the
core of the craton d uring the Grenville and Appalachian orogenies w ere central events.
The TA now w ill extend into southern Canad a, to fill in the ind entation caused by the
southw ard excursion of the US bord er in the Great Lakes region (Figure 1). This extension w ill,
in our opinion, be vital to m aking northeastern TA d ata as fantastically useful as the w estern
d ata has alread y proven to be. N evertheless, it w ill not really ad d ress the problem of im aging
the core of the continental craton, because it w ill cover m ostly Proterozoic-age terranes. The
central Archean-age core of N orth Am erica is further to the north (Figure 2). Thus, in ad d ition
to ad d ressing issues specifically id entified in this prop osal, our seism ic array w ill extend the TA
into the Archean-age region of central Québec, improving its overall configuration and ability to
ad d ress the broad goals set out in its Science Plan.
Our proposed array consists of tw o parts: 1) a sparse linear array, w ith station spacing
com parable to the TA, that extend s from the Paleozoic Appalachian orogen on the N ova Scotia
coast to the Archean cratonic core south of H ud son Bay; and 2) a set of three d ense subarrays,
each strad d ling a m ajor terrane bound ary (Superior-Grenville, Grenville-Appalachian and
intra-Appalachian), w ith station spacing on the ord er of 10 km . Our array w ill be supplem ented
by existing Canad ian broad band stations that provid e sparser tw o -d im ensional coverage of
Québec. The d ata from these Canad ian stations is public, though not archived by the DMC, and
w e w ill, as part of this project, obtain, analyze and archive it. We are proposing, in collaboration
w ith the University of Québec at Montreal (UQAM), to operate this array for three years, a tim e
period that w ill overlap w ith the easternm ost d eploym ent of the TA.
Background.
1. Lithosphere/A sthenosphere System . The notion that lithosphere beneath continents is distinct
from that of the oceans, in both its properties and its longevity, fo rm s the basis of the
tectosphere concept proposed by Jord an over three d ecad es ago (Jordan, 1978). Subsequent
stud ies of continental lithosphere have show n that, though its vertical extent varies greatly, it
C-2
d oes not seem to extend anyw here as d eep as the transition zone (as in the original tectosphere
proposal) (Fischer and van d er H ilst, 1999; Gung et al., 2003; Eaton et al., 2009; Fischer et al.,
2010). Its d istinct properties are d ue both to thermal and chem ical effects (e.g., Perry et al., 2003;
Artem ieva, 2006). Clear correlations relate ages of the large continental terranes and properties
of their lithosphere (e.g., Perry et al., 2003; Sim ons et al. 2002). Extensive literature d ocum ents
lateral variation of properties w ithin the continental lithosphe re. Lithosphere also has internal
layering on a broad range of scales (e.g., in Eurasia, Morozova et al., 2000; N orthern N orth
Am erica, Snyd er et al., 2004; Yuan and Rom anow icz, 2010; Arabia, Levin and Park, 2000).
Our present-d ay notion of the continen tal lithosphere is that of a prod uct of “tectonic”
processes, sim ilar to those that shape the rocks exposed on the surface of our planet. To ad vance
our und erstand ing of how continents have form ed in the past, and how they evolved over the
course of the Earth’s history, w e need to d evelop links betw een tectonic history and resulting
lithospheric structures. In regions of active tectonism (e.g., the H im alayas), w e can stud y a
snapshot of the ongoing continental evolution. Stable continental interiors, on the other hand,
offer a record of this evolution, w hich need s to be unraveled . In this view , the d eep structure of
the continental lithosphere is the longest preserved record of events on Earth.
The interaction betw een the continental lithosphere and the rest of the Earth’s mantle, and
especially the asthenosphere, is com plex. Processes that influence the state of the lithosphere
includ e: lateral asthenospheric d rag (e.g., Bokelm ann, 2002; Artem ieva & Mooney, 2002; Eaton
and Fred eriksen, 2007); plum es and other types of upw elling (e.g. Sleep, 1990; Rond enay et al.
2000; Moucha et al., 2008; Schutt et al., 2008), d ow n -w ellings (e.g. Forte et al., 2007; 2010),
und erplating, etc. In turn, stiff continental lithosphere likely im pacts the d ynam ics of the
m antle, e.g. by creating cond itions for sm all-scale convection (King & And erson, 1998) or by
restricting asthenospheric flow by lateral or vertical constriction (Alvarez, 1982; Buck et al.,
2009). Large continental areas m ay act as therm al blankets (Gurnis, 1988), w arm ing the
asthenosphere, though this notion has recently been questioned (Lenard ic et al., 2005; Trubitsyn
et al., 2008). The processes in the lithosphere and asthenosphere are especially interconnected in
cases w here gravity-d riven instabilities d evelop (e.g., H ousem an & Molnar, 1997; Pysklyw ec et
al., 2000; 2010; Zand t et al. 2004).
Vertical and lateral changes in properties of the
asthenosphere, and how they relate to the
configuration of the lithospheric “lid ” above, are
the key constraints on the m od els w e have for the
d ynam ics of the joint lithosphere/ asthenosphere
system . Both num erical sim ulations and the
grow ing num ber of regional studies of lithosphereasthenosphere interaction suggest that this system
has a broad range of behaviors (Pysklyw ec et al.,
2010). H ow different m od es of interaction develop,
w hat controls them , and how they changed over
the course of Earth’s evolution are key questions.
Figure 2. A schematic map of major tectonic divisions in
the study region. Thick lines denote primary targets of
the proposed study – the Grenville Front (red), the St.
Lawrence Rift / A ppalachian Front (blue) and the
C-3
N orumbega Fault Z one (green).
Eastern N orth Am erica is the prod uct of a long continent -build ing process that is, for the tim e
being at least, “finished”. Unlike m any other parts of the N orth American continent, it is
presently not being m odified , sim plifying the task of relating past causes and effects. Despite
the tectonically “passive” state, there is am ple evid ence for com plex and d iverse structure both
w ithin the bod y of the continent, and in the asthenosphere beneath it (e.g., Moucha et al., 2008,
also see follow ing section). An investigation of the proposed region w ill m ake a significant
contribution to the bod y of know led ge about lithosphere/ asthenosphere system behavior.
2. Eastern N orth America. Accreted over the last 3 Ga, the N orth Am erican continent consists of a
set of Archean cratons, Paleoproterozoic orogenic belts and a sequence of younger terranes
(Whitm eyer and Karlstrom , 2007; H ibbard et al., 2007; see Figure 2).
The old est of the post-Archean terranes is the Grenville province, form ed at ~1 Ga and
associated w ith the closure of a now −d efunct ocean basin d uring the Proterozoic (Moore, 1986;
Whitmeyer and Karlstrom , 2007; H ynes & Rivers, 2010). Its origin has yet to be com pletely
established , but it contains som e accreted sections as w ell as rew orked material of the Superior,
and of Paleoproterozoic-age orogens. It separates the cratonic core of the continent from the
younger (0.3−0.4 Ga) Appalachian terranes that w ere accreted d uring the closure of the
also−d efunct Iapetus ocean d uring the Paleozoic (Taylor, 1989; H ibbard et al, 2007). These three
teranes have very d ifferent ages of form ation and are separated by bound aries that are w ell
d efined structurally. H ow ever, the deep lithospheric expression of these bound aries is presently
unknow n (and w ill be investigated by our proposed effort). The d eep bound aries d o not
necessarily coincid e w ith those on the surface, because surficial structures are d om inated by
shallow -d ipping thrust faults, as w as revealed by the LITH OPROBE program in Canad a
(Clow es et al.; 2010).
Seism ic tom ography stud ies consistently id entify the region around H ud son Bay as the
locus of extrem ely fast m antle that extend s to d epths of about 300 km (e.g., N ettles and
Dziew onski, 2008; Grand , 1994; van d er Lee 2001; van d er Lee and Fred eriksen, 2005). At a
d epth of 150 km this region is one of tw o cold est regions in the Earth’s m antle (Forte and Perry,
2000; Artem ieva, 2006, Figure 3).
Measurem ents of surface w ave
d ispersion betw een sites in this
area yield very high values of shear
w ave speed d ow n to d epths of 150250 km (Darbyshire et. al 2007;
Darbyshire & Eaton, 2010).
Figure 3. A ge and thickness of the N orth
A merican
continent
derived using
geological information (the ages of crustforming events) and a global heat flow
database. A dapted from A rtemieva (2006)
and www.lithosphere.info .
The bulk of the old and thick lithosphere of N orth A merica lies north of the region to be sampled by the USA rray.
C-4
Lithospheric properties change from the core of the continent tow ard s the m argin, although not
in a sim ple w ay. Surface w ave tom ography (van d er Lee, 2002) sugges ts relatively thin
lithosphere (80 km ) near the Atlantic coast of N ew England . This estimate agrees to a d egree
w ith find ings of sharp seism ic im ped ance contrasts at d epths of 90-110 km using converted m od e bod y w aves (Rychert et al., 2005; 2007; Abt et al., 2010). Due to a very gentle grad ient of
speed w ith d epth , the surface w ave tom ography of van d er Lee (2002) fails to d efine the LAB
beneath the Appalachians. Sim ilarly, a surface w ave d ispersion stud y in the Ontario section of
the Superior Province (Darbyshire et al., 2007) is not able to d efine the greatest d epth of the
lithosphere there. At the junction of the Grenville and Appalachian terranes , a significant
change in the d egree of lateral heterogeneity of the upper m antle at ~300 km d epth w as
interpreted by Levin et al., (1995) as a transition from the lithosphere to the asthenosphere.
Beneath the Grenville and Superior Provinces this finding is supported by recent w ork by a
UQAM stud ent Melanie Villem aire (thesis supervised by Darbyshire), but is con trad icted by
Aktas and Eaton (2006) w ho find sim ilar scales of lateral heterogeneity extend ing throughout
the upper m antle. An upper lim it on the thickness of the lithosphere is placed by the fact that
the 410-km discontinuity is seen to be essentially fla t beneath eastern North America (Li et al.,
1998), w ith an im plication that the cooling effect of the continental m antle d oes not reach that
d eep.
N um erous tom ographic im aging stud ies d ocum ent strong lateral heterogeneity of the
lithosphere along the Atlantic seaboard . The heterogeneity appears in both continent -scale and
regional-scale im ages, suggesting m ultiple scales of velocity variation (Levin et al., 1995; 2000;
Li et al., 2003; N ettles & Dziew onski 2008, van d er Lee & Fred eriksen 2005; Villema ire, 2011).
Most notable is the low -speed em baym ent at d epths ~100-150 km som etim es referred to as the
“d ivot” (Fouch et al., 2000; area shad ed yellow at the low er ed ge of the m ap in figure 4). In
m any stud ies this low -velocity anom aly extend s further to the n orthw est (e.g. van d er Lee &
Fred eriksen, 2005; N ettles & Dziew onski, 2008, Rond enay et al., 2000; Fred eriksen et al., 2007). It
follow s a zone of Cretaceous volcanism , a seism ically active zone in w estern Québec and a set
of kimberlite d eposits (e.g. H eam an & Kjarsgaard, 2000) in northeastern Ontario, and is thought
to be correlated w ith the interaction betw een the N orth Am erican continent and the Great
Meteor hotspot (e.g. Sleep, 1990).
Figure 4.
Geophysical constraints on the
structure and thickness of the
lithosphere in the study region.
Colors show shear wave speed at
100 km depth in the model of
N ettles and Dziewonski (2008).
W hite contours (labeled in km)
are estimates of the depth to the base of the lithosphere from
the 1°x1° global model based on heat flow and regional
geology constraints (A rtemieva, 2006). Faster speed
generally corresponds to thicker lithosphere, but not always.
(inset ) A vertical profile through the shear velocity model
along the proposed array deployment (purple line on the
map). N ote the very low level of detail in the image.
C-5
Fouch and Rond enay (2006) review evid ence for significant levels of seism ic anisotropy
throughout the region. H ow ever, the relations of the observed anisotropic texture to the present
plate m otion, asthenospheric processes and the history of tectonic events in the region are not
fully w orked out. Various authors explained the observations in term s of texture rem nant from
the tim e of continental accretion (e.g., Barruol et al. 1997), asthenospheric flow m od ulated by
lithosphere shape (Fouch et al., 2000), and a com bination of both resulting in at least tw o layers
of texture (Levin et al., 1999; 2000ab; Yuan and Rom anow icz, 2010). At least one stud y (Levin et
al., 2000b) interprets the anisotropic signature as evid ence of a past d elam ination (or “d rip ”)
event at the tim e of Appalachian orogen form ation.
3. LITHOPROBE’s view of Continental A ccretion. Across Canad a, the LITH OPROBE project
(Clow es et al., 2010) provid ed a w ealth of inform ation about crust al structure and the tectonic
processes that shaped the Canadian Shield and its peripherals. The Abitibi-Grenville transect
(Canadian Journal of Earth Sciences, vol. 37) w as an in -d epth stud y of the structure and
tectonics of the Grenville Province, using active source seism ology near the Ontario-Québec
bord er and also in eastern Québec near Lake Manicouagan . The m ain focus of the geophysical
part of these transects w as crustal architecture. Refraction and reflection stud ies (e.g. Lud d en &
H ynes, 2000; H ynes et al., 2000; Martignole et al., 2000; Mereu et al., 2000; White et al., 2000)
show a com plex crustal structure in w hich terrane bound aries d ip southw ard s in the Grenville
Province and northw ards in the Superior craton. In particular, the Grenville Front is interpreted
as the surface expression of a southw ard -dipping tectonic front. Of particular note in the
Superior is a strong reflector beneath the Moho, d ipping north to a d epth of 60-70 km ,
interpreted as the expression of a fossil subd uction zone asso ciated w ith the accretion of the
Abitibi subprovince to the Opatica subprovince (see Figures 2, 5). Crustal thickness and the
character of the Moho vary along the transect, from ~35 km to over 50 km .
Figure 5: M ap of the seismic refraction-reflection studies
from the LITHOPROBE A bitibi-Grenville transect
(www.LITHOPROBE.ca).
The
LITHOPROBE
teleseismic transect discussed in the text lies close to the
north-south-trending set of lines just east of the OntarioQuébec border. A s shown by the thick dashed line, our
proposed teleseismic transect lies in a region not studied
by the LITHOPROBE project.
The LITH OPROBE Abitibi-Grenville stud y
also includ ed a north -south transect of
broad band seism ographs along the OntarioQuébec bord er (Rond enay et al., 2000a, 2000b).
The d ata collected w ere used to provid e
inform ation on crustal/ upperm ost-m antle
structure from receiver functions, on m antle anisotropy from SKS splitting analysis and on local
m antle heterogeneity from P-w ave travel tim e tom ography. Although the array aperture w as
very lim ited , tom ographic m od els show ed a local linear WN W -ESE low -velocity anom aly that
w as interpreted as lithospheric m od ification arising from the passage of the Great Meteor
hotspot beneath the Canad ian Shield .
C-6
LITH OPROBE provid ed a w ealth of inform ation about the crustal expression of terrane
bound aries (although not in the region of our proposed stud y, w hich lies alm ost exactly
betw een the tw o LITH OPROBE transects; Figure 5).
H ow ever, except to the extent that one line im ages the
shallow m antle expressions of w hat m ight be a fossil
subd uction zone, w hich reinforces the im portance of
subd uction in the continental assem bly process,
LITH OPROBE has provid ed no d irect inform ation
about either the expression of sutures in the d eep
lithosphere or of the depth to the LAB. Thus, our
project can be thought of as extending LITH OPROBE
both laterally – and into a region unaffected by hotspots
– and to a m uch greater d epth.
Figure 6. P wave tomography slice at depth of 150 km
illustrates the presently available resolution of seismic
properties in central Québec (V illemarie, 2011). Red line
shows proposed array, white lines show the Grenville Front and the A ppalachian Front.
4. Recent regional studies. Prelim inary crustal m od els have been obtained through receiver
function analysis (Darbyshire & H obbs, 2011) for a set of recently -deployed seism ograph
stations in central and w estern Québec, operated by the University of Québec at Montreal
(UQAM). Four of the stations analyzed lie along our proposed transect and w ill form part of the
project d ata set. The prelim inary m od els suggest crustal thicknesses in the range 35-40 km , and
an intriguing d egree of crustal com plexity. LAB d epth and signature w ere probed (als o by
receiver function analysis, Menke et al., 2010) at a set of sites in Maine and Québec. We found
cand idate LAB features beneath the Appalachians and the Grenville Province, but had a hard
tim e id entifying one beneath the Superior. A recent P-w ave regional tom ography stud y by
Villem aire (2011) m od els upper m antle structure from 100 km to 900 km d epth in the Superior Grenville-Appalachians area. As show n in figure 6, our proposed transect lies on the ed ge of the
station coverage used in the tom ographic m od el and , as such, should provid e com plem entary
inform ation and im prove constraint on the geom etries of the m antle heterogeneities imaged
beneath the northeastern part of the m od el.
Scientific Themes We Will Pursue
1. The nature of the Lithophere-Asthenosphere Boundary
Id entified as one of ten “Grand Challenges” in seism ology (Lay, 2009), the LAB is the subject of
very active research. Long-period surface w ave tom ography indicates that at a global scale the
LAB has significant topography, being d eepest b eneath the cratons (Gung et al, 2003;
Rom anow icz, 2009; Fischer et al., 2010). Unfortunately, the d epth resolution of the tom ography
is too poor to resolve d etails such as w hether the LAB is a d istinct interface or a broad transition
zone (e.g. figure 4; Eaton et al., 2009). This issue has been ad d ressed through a com plem entary
technique, receiver function (RF) analysis, w hich id entifies sharp interfaces through the short period seism ic w aves converted from them (Eaton et al., 2009 and Fischer et al., 2010 provid e
com prehensive review s). RF analysis, w hen applied at the global scale (e.g., Rychert and
Shearer, 2009), ind icates that a d istinct interface is present at d epths of 80-120 km . This interface
C-7
has a negative velocity jum p w ith d epth, consistent w ith a seismically-faster lithosphere above a
slow er asthenosphere, and hence has been id entified as the “LAB interface”. The m ean d epth of
the “LAB interface” varies from region to region, being (as expected) d eepest und er the cratons
and shallow est und er the oceans. H ow ever, the overall m agnitud e of its topographic variation
(about 25 km ) is less than the am ount expected from global surface w ave tom ography (at least
100 km ). An ad d itional com plication is the fact that in som e places this feature clearly falls
w ithin the lithosphere (as constrained by xenoliths and tom ography, Fischer et al, 2010). Abt et
al. (2010) interpret such observations as evid ence for a “m id -lithospheric d iscontinuity” beneath
N orth Am erica.
Whether the “LAB interface” is a ubiquitous feature w hen view ed at regional scales is less
certain, since global com pilations m ight not necessarily pick up on sm all-scale variability.
Continent-w id e surveys in N orth Am erica (Abt et al., 2010) and Australia (Fischer et al., 2010)
d ocum ent the w id esp read presence of the “LAB interface” beneath Proterozoic and
Phanerozoic regions. Stud ies in the Archean Kaapvaal craton are m ore equivocal. Savage and
Silver (2008) interpret a 150 km d eep interface as an intra -lithospheric bound ary, in contrast to
the LAB, citing the deeper extent of the lithosphere d eterm ined by tom ographic m ethod s. On
the other hand , H ansen at al., 2009 interpret a sim ilar feature as the LAB on the basis of it being
the only feature w ith a velocity d rop w ith d epth. Kum ar et al. 2007 rep ort a sim ilar but d eeper
(250-300 km ) feature beneath Kaapvaal, w hile Wittlinger and Farra (2007) argue for m ultiple
converters betw een 150 and 300 km d epth. Thus, w hile d eep (>100 km) interfaces occur beneath
the Kaapvaal craton, the id entification of any one of them as “the” LAB is contested .
Our proposed array w ill provid e the ability to perform a key test of the d epth-behavior of the
LAB interface – w hether it is continuous across regions of d ifferent tectonic history and w hether
it plunges d eeply d ow n beneath cratons. We w ill also establish the d eep -lithospheric
expression of the terrane bound aries and search for steps in LAB d epth across them , as m ight be
expected if the lithosphere retains some of its pre-assem bly configuration.
The planned array crosses a region w here the lithospheric thickness, as inferred from both
surface w ave tom ography and heat flow , changes d ram atically, d oubling from about 100 km at
the Atlantic coast to over 200 km near H ud son Bay (Artemieva, 2006; N ettles and Dziew onski,
2008; Darbyshire and Eaton, 2010; Figure 4). It also avoid s areas, such as the “d ivot” (see
above), w here hotspot activity m ay have caused lithospheric thinning. We can exam ine w hether
the d epth to the LAB interface, as inferred from RF analysis of the arr ay d ata, closely tracks this
w ell-d ocum ented thickening. Such a d oubling, w ere it to be observed , w ould give further
cred ence to the notion that the LAB interface identified by RF analysis really d em arcates the
bottom of the lithosphere. If the relief of the interface is d eterm ined to be less - and especially if
it is found to have very subd ued relief - then its interpretation as a true delim iter of the bottom
of the lithosphere w ould be in d oubt. In this case, a key issue w ill be w hether the d etailed
prop erties of the interface and the lithosphere above it (as d eterm ined through analysis of the
array d ata) give any hint to its actual significance. Factors such as velocity contrast across the
interface, the presence or absence of anisotropy, its cross-cutting relationships w ith any other
observed features; all w ould be im portant elem ents of a reinterpretation.
2. Lithospheric expression of terrane boundaries
Tod ay's continents w ere built up over a period of nearly four billion years through the plate
C-8
tectonic processes that includ e the rew orking of locally -d erive m aterial and the accretion of
sm aller, genetically d istinct terranes. In m any instances, the original m antle lithosphere of these
terranes has been preserved , as is evid enced by the age correlation observed betw een m antle
sam ples brough t up in xenoliths and d iam ond inclusions and surficial rocks (Sleep, 2005;
Carlson et al., 2004). H ow ever, the lithosp heric m antle m ay subsequently have been m odified
by processes such as the und erplating of new m a terial as the lithosphere cooled further and
incorporated form er asthenosphere; d elamination or "d rips" as the lithosphere becam e
gravitationally unstable; and therm o-m echanical erosion via plum es, subd uction and rifting.
The relative abundances of "pristine" original lithosphere, heavily-m od ified lithosphere and
new lithosphere are m ostly unknow n.
Our proposed array focuses on three com plex terrane bound aries that d iffer in age of their
form ation (from Meso-proterozoic to Paleozoic), and in their subsequent tectonic history. While
having been recognized for a long tim e as m ajor features of the surface geology, all three
bound aries have in comm on a lack of clear understand ing of w hether, and how , they extend
into the lithosphere.
Superior-Grenville Boundary (Grenv ille Front ). The continent-scale Grenville Front (GF)
separates the exposure of the Archean Superior province from the Grenville Orogen (Irving et
al., 1972; Moore, 1986). Initially believed to be the locus of the continent -continent collision, and
thus a quintessential “continental suture” (Dew ey and Burke, 1973), the GF has been later
interpreted as a m ajor contractional fault system (e.g., Rivers et al., 1989) acting on the form er
passive m argin of the Archean -age continent. Extensive (10 km ) and laterally varying d egree of
uplift w as accomm od ated by the GF (e.g., Rivers et al., 1989). Since the end of Mezoproterozoic
(~1 Ga, H ynes and Rivers, 2010) there w as little tectonic activity on the GF. Seism ic stud ies (e.g.
Green et al., 1988; Martignole et al., 2000; White et al., 2000) show ed it to extend through the
m id d le crust, and in m any reconstructions (Rivers et al., 1989; Lud d en and H ynes, 2000; H ynes
and Rivers, 2010) it is show n to cut the crust and sole into the Moho. Associations of the GF and
lithospheric-scale structure have been proposed by Aktas and Eaton (2006) (seism ic velocity
change) and Fred eriksen et al. (2006) (local variation in m antle fabric).
Grenville-Appalachian Boundary (The Apallachian Front ). The bound ary betw een the
Appalachians and the Grenville Province (historically referred to as Logan’s Line, Alcock, 1945,
Thom as, 2006) runs SW, crossing into St. Law rence River near Quebec City (e.g., Trem blay et
al., 2003). It is another case of tectonic inheritance (Thom as, 2006), as tw o episod es of rifting
(one successful and one not), and a contractional tectonic front all took place broad ly along this
bound ary. A locus of faulting associated w ith the opening of the Iapetus ocean in earliest
Paleozoic tim e (Kum arapeli, 1985), and subsequently a northw estern -m ost reach of the Taconic
nappes, the Appalachian Front (AF) is a good exam ple of the m ism atch betw een surface and
d eep bound aries – Grenville-age rocks are know n to extend east beneath it. N early coincid ent
w ith the AF is the Mezozoic-age St. Law rence rift, a site of failed continental separation, and one
of the m ost seismically active areas of Eastern N orth Am erica (e.g. Lam ontagne et al., 2003).
Int ra-Appalachian Boundary (Norumbega Fault Zone). The N orum bega Fault Zone (N FZ) of
coastal Maine is a 40 km w id e and over 400 km long d extral shear zone erod ed d ow n to m id crustal d epths (Lud m an and West, 1999), w ith evid ence of m otion from m id -Paleozoic to
Cretaceous tim e (Wang and Lud m an, 2004; West and Rod en -Tice, 2003). In earlier
reconstructions of Appalachian terrane m osaic (e.g. William s and H atcher (1982)) the boundary
C-9
betw een tw o m ajor terranes, Gand er(ia) and Avalon(ia), w as traced along this fault zone,
although m ost recent com pilations (e.g. van Staal et al., 2009; H ibba rd et al., 2007) d raw this
bound ary offshore. As argued by Lud m an (1986) the N FZ started as a suture betw een elem ents
of the future Gand er terrane, but subsequently acted as a transcurrent bound ary, w ith possible
m od ern analogs being the San And reas, Anatolian or Denali faults.
Tw o key questions are the d egree to w hich the lithosphere preserves a seism ically d etectable
structural signature of the original terranes, and w hether the assem bly process creates new ,
recognizable signatures that are localized to the im m ed iate vicinity of the bound aries. These
questions can be ad d ressed by exam ining the cross-cutting relationships, and is sim plified by
the approxim ate locations of the bound aries being know n on the basis of surface geology.
In a contractional environm ent, the surface geology at a terrane bound ary is d om inated by
low angle faults, m any of w hich are im bricated , and w hich cause allocthonous slivers of
m aterial being transported far from their point of origin. The d eeper part of the lithosphere is in
a d uctile, as contrasted to brittle, d omain, and so m ay have a com pletely d ifferent character.
Measurem ents of bulk seism ic velocity, anisotropy and lithospheric thickness across the
bound aries are key to d eterm ining w hether they narrow or w id en w ith d epth.
Of these properties, anisotropy has been arguably the best stud ied to d ate. H ow ever, few
cases have been id entified in w hich the shear w ave fast d irection sharply and unequivocally
changes across a tectonic front. For exam ple, Rond enay et al. (2000) observ e only a m od est
rotation of fast direction across the GF, not a sharp jum p. One possibility is that the lithospheric
signal is being obscured by a stronger one associated w ith asthenospheric flow (see Section 3,
below ). Analysis for m ultiple anisotropic la yers (e.g. Levin et al., 1999, 2008) m ay provid e a
clearer sense of variation in the lithosphere.
An intriguing possibility is that several “LAB” interfaces m ight be present in som e areas
(e.g., Wittlinger and Farra, 2007), a d iscontinuous set of shallow er and old er interfaces that
pred ate the assembly, and a d eeper, younger and m ore continuous one that has d eveloped
afterw ard . The subd uction zone interface observed at ~100 km d epth betw een tw o terranes
w ithin the Slave Craton (Chen et al., 2009) might cou nt as one of these, since presum ably the
present-d ay LAB, w hile not yet im aged , is now d eeper. H aving, as in northeastern N orth
Am erica, a suite of sutures of d ifferent ages m ight prove critical to the recognition of this
situation.
3. Exploring the lithosphere-asthenosphere transition w ith mantle-flow induced seismic
anisotropy
The d epth-depend ent transition in m antle flow from the lithosphere to the asthenosphere and
the correspond ing variations in lattice-preferred orientation (LPO) of mantle m inerals can be
explored through direct com parisons of seismic anisotropy w ith predictions of m antle flow
obtained on the basis of seism ically-constrained geod ynam ic m od els. These com parisons w ill
provid e unique insights on the im pact of rheological stratification betw een the lithosphere and
asthenosphere, and hence the d egree to w hich the d eform ation in these tw o layers is coupled.
The geod ynamic im portance of such com parisons betw een m antle flow and anisotropy has
been previously illustrated in tom ography-based m antle convection pred ictions und er the
w estern US by Becker et al. (2006) and und er the eastern half of the US by Forte et al. (2010).
The possibility of correlating m antle flow to seism ic anisotropy has been assum ed since the
connection w as first proposed by H ess (1964). The m ain d ifficulty w ith this assum ption is that a
sim ple connection betw een flow d irections and LPO inferred from seism ic anisotropy is not
C-10
w ell established on theoretical ground s (e.g. Kam inski & Ribe 2002). The relationship betw een
seism ic anisotropy and m antle flow has been show n to d epend on the geom etry of the flow
field (e.g. 2-D plane strain versus 3-D d eform ation) and especially on the com bined spatial and
tem poral variation of the d eform ation history in the m ineral grains (Kam insk i & Ribe 2002). In
previous w ork Forte’s group has consid ered a num ber of possible proxies for LPO and their
correlation to seism ic anisotropy, includ ing the pred icted m antle flow d irections and the
orientation of axes along w hich the rate of extension is at m axim um (e.g. Gaboret et al. 2003,
Forte et al. 2010). These flow -related proxies for seism ic anisotropy are sensitive to d etailed
geom etry of the m antle flow field and to the assum ed m antle viscosity structure. The latter
sensitivity is especially useful for exploring the d ifferences in d eform ation in the asthenosphere
relative to that in the lithosphere.
Figure 7. Relating previous
inferences
of
azimuthal
seismic anisotropy below
eastern N orth A merica with
mantle flow predictions at
130 km depth, from figure 6
of Forte et al. (2010). The
thick brown line is a portion
of the proposed dense-array
transect shown above in
figure 3. The magenta bars
represent
inferences
of
anisotropy from shear-wave
splitting analysis carried out
by Barruol et al. (1997) and
Fouch et al. (2000). The green bars represent the horizontal component of maximum flow-induced
stretching. The blue arrows represent the horizontal component of mantle flow. The inputs employed to
calculate these mantle flow predictions are detailed in Forte et al. (2010).
In Figure 7 w e com pare m antle-flow proxies for m ineral preferred orientation und er the eastern
US w ith inferences of azim uthal anisotropy obtained by Barruol et al. (1997) (w ho thought the
signal in in the lithosphere) and Fouch et al. (2000) (w ho posited asthenospheric origin of the
signal). The inferred d irections of anisotropy are based on relatively short observation period s,
and assum e a single layer of anisotropic m aterial. Thus they represent an average of
lithospheric and asthenospheric values, as illum inated from a few d irections (m ostly in Western
Pacific). N evertheless, w ith few exceptions, w e observe a rem arkably good correlation betw een
splitting observations and the pred icted present-d ay, asthenospheric flow d irections. One
possible interpretation of this overall agreem ent is that the main cause of seism ic anisotropy in
this region is d ue to m antle-flow ind uced LPO in the asthenosphere. There are, how ever,
regions of notable d isagreem ent, includ ing one in the vicinity of ou r proposed array (upper
right-hand corner of Figure 7) that m ust be und erstood before this hypothesis can be accepted .
Significant com plexity is seen in a m ore recent SKS splitting com pilation of Liu (2009), and
Levin et al. (1999; 2000a) argued for at least tw o d istinct vertically separated layers of
anisotropy in N ew England Appalachians.
C-11
The long deploym ent d uration (3 years) of our array w ill provid e d ata w ith the broad azim uthal
coverage need ed for precise d eterm ination of anisotropic d irections for both single-layer and
m ultiple-layer analyses.
The hypothesized d oubling of lithospheric thickness from the Atlantic coast to cratonic core
along the array should result in coherent variation in anisotropic layer thicknesses in a tw o layer interpretation. Deviations betw een the pred icted asthenospheric flow and the observed
seism ic anisotropy w ill provid e im portant clues about the w ay that anisotropic fabric is created
and m aintained .
PROPOSED WORK.
Observatory Array D esign.
Our array consists of a sp arse transect supplem ented w ith three higher-density subarrays. The
transect extend s from the coast of N ova Scotia, through Maine, to the coast of Jam es Bay in
Québec (Figure 8). The three sub arrays, one in Maine and tw o in Québec, strad d le the tectonic
bound aries that are our prim ary targets. In Maine w e w ill locate array nod es in consultation
w ith the state’s Geological Survey. In Québec, w e w ill be guid ed by Fiona Darbyshire w ho has
consid erable experience operating in central Québec, and by And rew H yn es of McGill
University w ho is an expert on the regional tectonics, and especially on the Grenville province .
Figure 8. A map of broadband seismic
observatories to be used in the proposed
research project. Blue and green
triangles show existing stations,
Canadian and US, respectively; open
circles show projected TA grid
locations; red triangles show the
proposed portable deployment (larger
symbols – 3 years, smaller – 2 years).
GF – Grenville Front, StLR – S.
Lawrence Rift, N FZ – N orumbega
Fault Zone.
D ense sub-arrays
GF
StLR
The geom etry of the array in central
Québec follow s the road system of
the region. All sites of the array w ill
NFZ
thus be vehicle-accessible and w ill
use, w henever possible, line pow er.
Darbyshire alread y operates a set of stations in the region, includ in g four along the transect line
that w ill anchor the sparse array. Other Canadian and US stations w ith open d ata form a sparse
2D netw ork, m ainly to the SW of the proposed transect, that w ill provide supplem entary d ata
capable of resolving lateral changes in properties on the ~200 km scale.
C-12
We plan to use FlexArray or PASSCAL equipm ent (d epend ing on availability). Based on
d iscussions w ith Dr. Greg And erson it is our und erstand ing that the use of USArray flexible
equipm ent in Canad a is w ithin the guid elines of the EarthScope w ork plan. Levin w ill
contribute three sets of broad band equipm ent (Taurus-Trillium 120 systems) ow ned by Rutgers.
Thus for the 3-year sparse array w e w ill need 5 ad d itional broad band sensors, and for the 2year d ense arrays – 26 sensors, m any of w hich m ay be of the “interm ed iate” variety (e.g.
Trillium 40 sensors) that are in less d em and .
The orientation of the array is chosen to optim ize m ultiple consid erations, as follow s: The array
sam ples the thickest part of the craton in central Québec, w hile avoid ing areas influenced by
hotspot m agm atism (e.g. the “d ivot” of Fouch et al., 2000); it is orthogonal to the three
bound aries, sim plifying the tectonic interpretation; it is approxim ately aligned w ith the
teleseismic arrival paths from northern and w estern Pacific subd uction zones (see Figure 9); and
it crosses the Charlevoix seism ic zone, an area of consid erable tectonic interest, w hich w ill
provid e regional seismic sources that w ill com plem ent the teleseismic d ata.
Suitability of the Proposed Array for Various Analysis Techniques.
Body wave tomography. The sparse array w ith an elem ent spacing of ~100 km is d esigned to
support tom ography of the upper m antle along a slice crossing Québec and Maine from the
Archean-age Superior province to the northw est, through the Proterozoic-age Grenville
province, to the Paleozoic-age Appalachian province in the southeast. The prim ary goal is to
im age along the line, but low er-resolution off-strike velocity control w ill be provid ed by the
existing 2D netw ork and TA d ata. We estim ate the num ber of usable events w e are likely to
collect by com parison w ith global seismicity d urin g years 2006-09 (Figure 9). During those 3
years, 528 m b≥6.0 earthquakes occurred , of w hich 193 had northw esterly backazim uths w ithin
30° of the axis of the d ense line, and 29 had correspond ing southeasterly backazim uths.
This suite of events provid es ad equate ray coverage for teleseism ic tom ography. Also, the array
w ill likely record about ten regional events of m b≥3.5 (e.g. in the Charlevoix seism ic zone,
through w hich the array passes, the Western Québec Seism ic Zone, etc). These regional events
w ill be especially useful in constraining the compressional and shear velocity in the crust and
upper 50 km of the mantle. We m ay even get lucky and record another M~5 event like the one
in June, 2010.
Imaging interfaces with converted phases. The array configuration and expected d istribution of
sources w ill allow us to use converted phases to d etect interfaces in the upper m antle. In
ad d ition to the Moho and the LAB, w e may find other interfaces that are reported either w ithin
the lithosphere (e.g., H ales) or at its low er bound ary (Lehm ann). The array w ill offer an
excellent view of the transition zone d iscontinuities as
w ell. A sw ath centered beneath the array w ill be extrem ely
w ell-sam pled , w hich w ill allow m igration -based analysis
m ethod s (e.g., Levand er et al., 2006) to be em ployed there.
Constraints on seism ic w ave speed from tom ography
along the array w ill allow proper m igration of converted
w ave energy. Changes in interface d epth (if any) across
geological province bound aries should be especially w ellresolved in this sw ath.
C-13
Figure 9. 3 years of seismic activity, 1/1/2006-1/1/2009, M >6.0.
Long operation of the array w ill be essential for resolving d etailed structure of id entified
interfaces, such as the magnitud e of change across them , vertical exten t, etc. (e.g. Rychert et al.,
2007).
Shear W ave Splitting. N ortheastern N orth Am erica is extrem ely w ell situated , in term s of the
azim uthal d istribution of global seismic source regions, for using core phases to observe m antle
anisotropy. Am ong events show n in Figure 9, 351 events w ith m b>=6.0 are betw een 90° and
180° from the array, a range in w hich SKS or PKS is typically w ell-observed . Areas of d ense
sam pling along the array w ill facilitate estim ates of the d epth w here anisotropy is located using
Fresnel zone argum ents (e.g., Salim beni et al., 2008), w hile the long d uration of observations
w ill ensure ad equate directional coverage need ed to em ploy m ulti-event inversions (Menke and
Levin, 2003). Constraints on vertical and lateral variations in anisotro py w ill be especially useful
for com parison w ith mantle flow calculations (e.g., Forte et al., 2010).
Surface wave tomography. Data from the array w ill be used to construct a d etailed image of shear
w ave speed in the lithosphere and asthenosphere along th e array using sources in the w estern
Pacific. This image w ill serve as the primary basis for estimating d epths to subhorizontal
interfaces id entified in converted -w ave analysis. The supplem entary 2D array, w ith likely
station spacing of ~200 km , w ill provid e coverage of otherw ise unsam pled parts of the region.
Jud ging from the three-year-long record of m b>6.0 seism icity, the azimuthal coverage w ill be
excellent: The three largest azim uthal gaps are only 8°, 9° and 17° w id e. We w ill be able to
resolve both lateral heterogeneities and azimuthal anisotropy using the 2D array (e.g.
Darbyshire & Lebed ev, 2009).
How the D ata Address the Scientific Hypotheses
Hypothesis 1: M ajor terrane boundaries have a deep seismically-imageable expression that cuts across the
pattern of pre-accretion lithospheric structures.
The d ense sub-arrays w ill allow the three bound aries to be probed w ith receiver functions, body
w ave tom ography and shear w ave splitting. The high station d ensity in the subarrays w ill
allow receiver functions to be m igrated (e.g., Levand er et al., 2006), a process w hich increases
the fid elity of resulting im ages and , especially, helps in the d etection of lateral variations. The
receiver functions w ill potentially im age both internal interfaces and the LAB, and so w ill be
able to d etect truncations and / or steps in their depth. Bod y w ave tom ography w ill be able to
d etect bulk changes in seism ic velocity and in the “texture” of sm all-scale heterogeneities,
across the bound ary. Sim ilarly, shear w ave splitt ing m ay reveal changes in the orientation of
anisotropic fabric, as w ould be expected from the juxtaposition of separate blocks.
Conversely, the hypothesis m ay be falsified by the d etection of features that are continuous
across the terrane bound aries and w hich, by the principle of superposition, post-d ate them .
Such observations w ould argue for substantial post -suture rew orking of the lithosphere. The
d ata w ill also be able to d istinguish a sub-vertical extension of a tectonic bound ary from a m ore
gently inclined one, and w ill enable com parisons betw een near -Moho and near-LAB structures.
Hypothesis 2: The lithosphere-asthenosphere boundary (LA B) deepens towards the center of the craton;
C-14
and commonly-employed seismic methods give mutually-consistent estimates of its depth.
Regional surface w ave tom ography w ill provid e a m ore-d etailed picture of the thickening of the
lithosphere beneath the craton than is available through continental-scale stud ies (e.g. N ettles
and Dziew onski, 2008; van der Lee 2001). Because of its poor d epth resolution, it is not
expected to provid e a specific d epth to the LAB, but w ill serve as a backdrop against w hich the
RF, bod y w ave tom ography and shear w ave splitting results are interpreted . The key
observation in support of the hypothesis w ould be an RF interface of the right sign for a fast over-slow interface, w hich d eepens system atically tow ard s the craton in a m anner consistent
w ith the surface w ave tom ography.
Alternatively, the hypothesis w ould be refuted if no RF interface deepens system atically across
the profile. This result w ould indicate that no RF reflector is consistently a true LAB, perhaps
because the bound ary is, in certain tectonic regim es, a transition zone too broad to be im aged
w ith high frequency w aves. Bod y w ave tom ography and splitting may provid e ancillary
evid ence for or against the hypothesis, if they have features that systematically vary across the
profile.
Hypothesis 3: That the pattern of azimuthal anisotropy in northeastern N orth A merica is prim arily an
asthenospheric signal associated with present-day mantle flow.
It is likely that the asthenosphere thins tow ards the center of the craton, as its top portion is
d isplaced by the lithosphere. Thus, to the extent that shear w ave splitting is the com bined effect
of an anisotropic lithosphere and an anisotropic asthenosphere w ith d ifferent textures, any
system atic variation in lithospheric thickness w ill cause a correspond ing variation in splitting.
The prim ary direction of N orth Am erican splitting is parallel to its absolute plate m otion (Liu,
2009) and hence likely to be asthenospheric, though a second , lithospheric layer has also been
ind entified through splitting stud ies (Levin et al. 1999) and inversion of surface w ave and
splitting d ata (Yuan and Rom anow icz, 2010).
The hypothesis w ould be refuted if the anisotropy, interpreted in the context of a tw o layer
m od el, d oes not evolve as expected across the profile.
Education and training. The m ain ed ucational goal of the proposed effort is to introd uce
und ergrad uates to scientific research through a sum m er internship involving 3-5 of them per
year in the field effort. Field trips w ill be d esigned to ed ucate the participants in the science
m otivating this proposal, as w ell as to m eet technical go als. Ed ucational elem ents w ill includ e
d aily group science discussions, stops to exam ine the geology, and visits to UQAM and the
Maine Geological Survey. Som e fund ing for und ergrad uates is built into this proposal, but w e
w ill also tap into fund ing available through the Colum ba University Earth Intern Program and
the IRIS internship program . The proposed project also has a grad uate ed ucation elem ent: It
w ill likely form the core of the d issertation w ork of Ms. Ayd a Shokoohi Razi (Rutgers) and w ill
also partially support a grad uate stud ent at Colum bia. In ad dition, und ergrad uate and grad uate
stud ent training at UQAM w ill likely be facilitated through the d ata collected by this project,
though w ith ind epend ent fund ing.
Project Management and Work Plan. The PI’s w ill w ork closely together and be jointly
responsible for the overall successful and tim ely com pletion of the project. Levin w ill take the
lead on array planning and data archiving; supervision of the RU grad uate stud ent and
und ergrad uates; analysis of the d ata using receiver functions and splitting m easurem ents.
Menke w ill advise the LDEO grad uate stud ent, and take the lead on field w ork logistics;
supervision of sum m er interns, and analysis of the d ata using body w ave tom ography.
C-15
Canad ian collaborator Darbyshire w ill take the lead on analysis of the d ata using surface w ave
tom ography. Geod ynamic m od eling by Forte w ill utilize our findings, provid ing interpretations
in term s of m antle flow . Seism ological find ings w ill be put into tectonic context w ith the help of
And rew H ynes of McGill University. Both PI’s and Darbyshire w ill be actively involved in the
fieldw ork to install, service and d ecom m ission the array, w ill lead field trips, and participate in
the integration of the results and their presenta tion at m eetings and in journal articles.
Operational plan for the field array. The array w ill be d eployed in tw o stages and w ill collect
d ata for 2-3 years. The sparse linear array w ill be installed in the sum m er of 2012, d uring w hich
tim e site surveys for the three d ense subarrays also w ill be perform ed . At this tim e w e w ill take
ad vantage of TA site survey activities in the southern regions of our stud y area. The three
subarrays w ill be installed d uring the sum m er of Year 2013. All instruments w ill be rem oved
d uring the sum m er of 2015. The TA is sched uled to operate in this region from 2013 to 2015.
Details of field operations sched ules and costs are presented in LDEO budget narrative.
To d eploy the array, w e w ill set up tw o staging areas, at LDEO an d at the University of Québec
at Montreal (UQAM). Both institutions have ad equate space for short -term storage of
equipm ent, and for perform ing necessary pre-d eploym ent tests. The same facilities w ill be used
at the end of the experim ent to d ecom m ission, p ack and ship the equipm ent. Most w ork w ill be
perform ed d uring summ er field trips, how ever w e w ill also visit the sites in fall of 2012 and
2013 for quality-control purposes (inspect site cond itions, d ow nload a subset of d ata for
evaluation purposes, etc). We w ill d ow nload d ata from the instrum ents annually, starting in
sum m er of 2013. Data storage capacity of the FlexArray d ata loggers is more than ad equate for a
year-long record ing period . We w ill seek to establish local oversight of our equipm ent (e.g., by
contracting w ith land ow ners), so that w e could be alerted about equipm ent em ergencies
(w eather d am age, vand alism etc.). Much of the array is reasonably close to the places w here PIs
live and w ork, thus w e have an option of accessing stations south of Lac Saint-Jean in Québec
any time.
D ata Management Plan. Data archival at IRIS DMC in Seattle w ill be perform ed in a speed y
m anner. Our sum m er em ployees w ill help w ith routine operations of d ata upload and transfer
to the DMC. Data from sparse array stations w ithin the TA footprint w ill be m ad e public upon
collection, to prom ote the goals of the Earthscope-w id e stud ies. We w ill release all other d ata
accord ing to the FlexArray/ PASSCAL rules, 24 m onths after com pleting the field w ork.
Prior Support.
Lev in: EAR 0545698 Seism ic anisotropy and rock texture w ithin the Cascad ia m egathrust zone,
02/ 06/ 2006 - 02/ 05/ 2010 (w ith a no cost-extension), $129,552.00
This project supported analysis of d ata from the TA d eploym ent in Cascad ia. We used receiver
function m ethod ology to characterize the crustal and upper m antle structure of the Cascadian
forearc. Key find ings based on the survey of perm anent observatories (N ikulin et al., G3, 2009)
are the clear signature of the top of the slab in most locations, frequent evi d ence for subd ucted
crust, and presence of SH -polarized m od e-converted phases best explained by seismic
anisotropy in a layer betw een the N orth Am erican and Juan d a Fuca plates, or w ithin the
subd ucting crust. The layer betw een the plates is restricted to the d epth range of the slab ~40
km , and has seism ic properties (low velocity, high anisotropy, high Vp/ Vs ratio) suggestive of
serpentinite. Analysis of d ata from the entire TA in Cascad ia confirm ed presence of
C-16
serpentinized layer along m ost of Cascad ia. This grant supported grad uate stud ent Alex
N ikulin and an und ergrad uate research assistant Ben Marshall.
Publication: N ikulin, A ., V . Levin, and J. Park, Receiver function study of the Cascadia megathrust:
evidence for localized serpentinization, G3, 10, Q07004, d oi:10.1029/ 2009GC002376, 2009.
Menke: Crustal Accretion and Mantle Processes Along the Subd uction -Influenced Eastern Lau
Spread ing Center (w ith Spahr Webb), OCE 0426369, 09/ 01/ 2008-08/ 31/ 2011. $320,795.00.
This is a com bined active and passive seism ic experim ent along the Eastern Lau Spread ing
Center to test the follow ing hypotheses: 1. Circulation in the m antle w ed ge is d om inated by slab
d riven flow . 2. Interaction of the arc and backarc m agm a prod uction controls the character of
the rid ge by influencing m elt flux, petrology, and geochemistry. 3. Variations in the m antle m elt
supply control rid ge crest features such as m orphology, therm al structure, and hyd rotherm al
venting. The array of 55 broad band ocean bottom seism ographs and five land seism ographs
w as successfully d eployed , operated for ~1 yr, and successfully recovered . We are currently in
the m id st of the d ata analysis process, and expect that d ata w ill enable us to im age the larger scale structure of the melt prod uction region and the m antle flow pattern.
C-17