Embed Size (px)
Citation preview
Tectonophysics 663 (2015) 339–353
Contents lists available at ScienceDirect
Tectonophysics
j ourna l homepage: www.e lsev ie r .com/ locate / tecto
Crustal structure of an intraplate thrust belt: The Iberian Chain revealedby wide-angle seismic, magnetotelluric soundings and gravity data
Hoël Seillé a,e, Ramon Salas b, Jaume Pous a, Joan Guimerà a, Josep Gallart c, Montserrat Torne c,Ivan Romero-Ruiz a, Jordi Diaz c, Mario Ruiz c, Ramon Carbonell c, Ramón Mas d
a Departament de Geodinàmica i Geofísica, Universitat de Barcelona, Barcelona, Spainb Departament de Geoquímica, Petrologia i Prospecció Geològica, Universitat de Barcelona, Barcelona, Spainc Institute of Earth Sciences Jaume Almera ICTJA-CSIC, Barcelona, Spaind Departamento de Estratigrafia, Universidad Complutense, Ciudad Universitaria, Madrid, Spaine Schlumberger, IEM CoE, Milan, Italy
E-mail addresses: [email protected] (H. Seillé),[email protected] (J. Pous), [email protected] (J. Guimerà(J. Gallart), [email protected] (M. Torne), ivan(I. Romero-Ruiz), [email protected] (J. Diaz), [email protected]@csic.es (R. Carbonell), [email protected]
http://dx.doi.org/10.1016/j.tecto.2015.08.0270040-1951/© 2015 Elsevier B.V. All rights reserved.
a b s t r a c t
a r t i c l e i n f oArticle history:Received 30 January 2015Received in revised form 10 August 2015Accepted 20 August 2015Available online 2 September 2015
Keywords:Magnetotellurics3D MT inversionSeismic refraction/wide-angle reflection profile2D seismic velocity–depth model2D forward gravity modelingIberian Chain
The IberianChain is a Cenozoic intraplate thrust belt locatedwithin the Iberianplate. Unlike other belts in the Ibe-ria Peninsula, the scarcity of geophysical studies in this area results in a number of unknowns about its crustalstructure. The Iberian Chain crustwas investigated bymeans of a NE-SW refraction/wide-angle reflection seismictransect and twomagnetotelluric profiles across the chain, oriented along the same direction. The seismic profilewas designed to sample the crust by means of three shots designed to obtain a reversed profile. The resultingvelocity–depth model shows a moderate thickening of the crust toward the central part of the profile, wherecrustal thickness reaches values above 40 km, thinning toward de SW Tajo and NE Ebro foreland basins. Thecrustal thickening is concentrated in the upper crust. The seismic results are in overall agreement with regionaltrends of Bouguer gravity anomaly and the main features of the seismic model were reproduced by gravitymodeling. The magnetotelluric data consist of 39 sites grouped into two profiles, with periods ranging from0.01 s to 1000 s. Dimensionality analyses show significant 3D effects in the resistivity structure and thereforewe carried out a joint 3D inversion of the full impedance tensor and magnetic transfer functions. The Mesozoicand Cenozoic basins along the Chain are well characterized by shallow high conductive zones and low velocities.Elongated conductors reaching mid-crustal depths evidence the presence of major faults dominating the crustalstructure. The results from the interpretation of these complementary geophysical data sets provided the firstimages of the crustal structure of the Iberian Chain. They are consistent with a Cenozoic shortening responsibleof the upper crust thickening as well as of the uplift of the Iberian Chain and the generation of its present daytopography.
© 2015 Elsevier B.V. All rights reserved.
1. Introduction
The Iberian Peninsula is constituted by a continental crust formedmostly during the Variscan Orogeny. While the center and westernparts the Central Variscan Massif has remained tectonically stable forthe last 300 Ma, the eastern part of Iberia is mostly formed by thickMesozoic sedimentary basins inverted during the Alpine orogeny asthe result of the convergence between the Euroasiatic and Africanplates. This has resulted in the development of the Pyrenees and theBetic Ranges at its northern and southern margins, but also in the for-mation of intraplate thrust belts. The Iberian Chain, the major of those
[email protected] (R. Salas),), [email protected]@ub.educsic.es (M. Ruiz),cm.es (R. Mas).
belts, developed during the Cenozoic because of the contractive inver-sion of the Iberian Mesozoic rift basins (Salas et al., 2001).
The Iberian Peninsula has been extensively explored using seismicdata since the late 1970s to determine its crustal structure. The crustalimbrication beneath the Pyrenees, reaching thicknesses of 45–50 km,has been documented from wide-angle (Daignières et al., 1981; Gallartet al., 1981) and deep multichannel seismic profiles (Choukroune andEcors-Pyrenees Team, 1989; E.C.O.R.S. Pyrenean Team, 1988) andconfirmed later on by Pedreira et al. (2003). The seismic explorationof the Valencia Trough using marine multichannel profiles and wide-angle profiles (Dañobeitia et al., 1992; Gallart et al., 1995; Torne et al.,1992; Watts et al., 1990) has revealed the strong variation in crustalthickness as a result of the rifting process that affected the zone, withMoho depths around 35 km in NE Iberia, thinning to 15–18 km beneaththe centre of the Valencia Trough and thickening again toward theBalearic promontory. The crustal structure beneath central IberianMassifwas already explored by early profiles (Banda et al., 1981), but the recent
340 H. Seillé et al. / Tectonophysics 663 (2015) 339–353
IBERSEIS and ALCUDIA experiments have allowed to define preciselyits geometry and velocity structure, characterized by a subhorizontalMoho located close to 32 km (Carbonell et al., 2004; Ehsan et al., 2014,2015; Simancas et al., 2003). Although scarce seismic information isavailable for the Iberian Chain, Zeyen et al. (1985) showed an averagecrustal thickness of 30–32 km, with a local thickening beneath thecentral northern part of the chain. Díaz and Gallart (2009) compiledthe results from deep seismic sounding profiles beneath Iberia and itssurrounding waters, providing a crustal thickness map of the area.Recently,Mancilla andDíaz (2015) have presented another crustal thick-ness map obtained in this case from the analysis of teleseismic receiverfunctions. Although some differences can be detected between bothmaps, the results from the two independent data and methods areremarkably consistent, hence confirming themain features of the crustalstructure beneath Iberia.
Magnetotelluric profiles have provided crustal resistivity images ofthe main orogens and related foreland basins in the Iberian Peninsula.The Alpine orogens are the Pyrenees (e.g., Campanya et al., 2012; Pouset al., 1995), the Cantabrian range (e.g., Pous et al., 2001), and theBetic Mountains (e.g., Pous et al., 1999; Ruiz-Constán et al., 2012),while Monteiro Santos et al. (1999), Almeida et al. (2005), Pous et al.(2004), Muñoz et al. (2008), and Pous et al. (2011) have investigatedthe Iberian Variscan Massif. As the electrical resistivity is a physical pa-rameter independent of the elastic seismic parameters, a combination
Fig. 1. Simplified geological map of the Iberian Chain and surrounding Tertiary basins (modifieFig. 2. magnetotelluric sites (red dots), seismic shots (blue stars), and the major geological uni
of both geophysical methods has contributed to clarify ambiguities inthe interpretations (e.g., Carbonell et al., 2004).
Nevertheless, these studies are scarce in the Iberian Chain, and norelevant information on the major crustal structures, in particular theMoho structure, has been obtained so far. The aim of this paper is topresent the results of a refraction/wide-angle reflection seismic profileand amagnetotelluric (MT) survey across the central part of the IberianChain, in order to gain new insights on its crustal structure. Further-more, the velocity model is converted to density and the predictedBouguer anomaly is compared with the new Bouguer anomaly datacompiled by Ayala et al. (Submitted for publication ) (the map anddata can be downloaded from http://geodb.ictja.csic.es/BBDDTopoIberia/)
2. Geological setting
The major geological features of the Iberian Peninsula are the resultof the geological events produced during the Phanerozoic. The Variscanbasement crops out extensively in the Variscan Iberian Massif (in thewestern half of the peninsula), but also within the Iberian Chain,which occupies the central and eastern peninsular areas (Fig. 1). TheVariscan basement was strongly eroded during the latest Paleozoic,a generalized planation surface being developed over it. This surfacewas covered by the Lower Triassic red beds (Buntsandstein facies) and
d after Guimerà, 2004, 2013). A seismic line (Fig. 5); (C and D) Geological cross sections ofts are shown. UTM coordinates (30 T, ED50) are also shown in km.
0 10 20 km
HercinianBasement
Palaeogene
Upper Cretaceous
Triassic
Lower and Middle Jurassic
Upper Jurassic andLower Cretacous
SSWNS NNE
0
0 m
EW
Sierra deAltomira Loranca Basin Serranía de Cuenca Thrust
TajoBasin
EbroBasin
SW NE
Castilian Branch Arch
B
C
Fig. 2. Geological sections across the central Iberian Chain. (B) After Guimerà, 2004; (C) after Guimerà and Álvaro (1990) and Muñoz Martín and de Vicente (1998). For location, see Fig. 1.
341H.Seillé
etal./Tectonophysics663
(2015)339–353
Fig. 3. Bouguer anomalymap of the study area. Details on data gathering and database can be found in Ayala et al. (Submitted for publication ) and Torne et al. (2015). Themap has beencalculated using the geodetic reference system GRS80 and a reference density of 2670 kg/m3. Color key shows Bouguer Anomaly. Contours every 10 mGal. Shading indicates elevation.White thick lines shows location of seismic profile. Yellow thick lines geological cross-section shown in Fig. 2. Red dots show MT sites. Access to the gravity data can be found athttp://geodb.ictja.csic.es/BBDDTopoIberia/.
342 H. Seillé et al. / Tectonophysics 663 (2015) 339–353
younger Mesozoic rocks in the Iberian Rift System (central and easternparts of the peninsula).
The geological structure and topography of the Iberian Peninsulaare the result of the Cenozoic contraction, which was a consequenceof the Euroasiatic and African plate convergence. The mountain rangesthat constitute the Iberian Chain, a double-vergent intraplate thrustbelt, with a NW-SE main trend (Guimerà and Álvaro, 1990), resultedfrom the contractional inversion of the Iberian Mesozoic basins—alsoinvolving the Variscan basement between the North Iberian and theSerranía de Cuenca Thrusts (Figs. 1 and 2) (Guimerà et al., 2004; Salaset al., 2001). In the areas where the basement has been involved inthe inversion, two big arches (the Castilian Branch Arch and theAragonese Branch Arch) can be distinguished, separated by theAlmazán Syncline. The Cenozoic thrust system has been proposed toinvolve only the upper crust with 10–12 km deep detachment level(Guimerà and Álvaro, 1990) or the whole crust (Anadón and Roca,1996 and Salas and Casas, 1993).
The Iberian Chain exceeds 2000 m above sea level locally, butits moving average elevations (using a 15 km radius search) rangebetween 800 and 1500 m. The elevations in the Cenozoic sedimentarybasins around the chain (Ebro, Duero, Tajo, etc.) range from 700 to1000 m in the Duero Basin, 500 to 900 m in the Tajo basin, and 200to 700 m in the Ebro basin. The only deep seismic reflection profileavailable (ECORS) in the northern Ebro Basin, north of the IberianChain, displays a crustal thickness of about 34 km (Roure et al., 1989).
In the eastern coastal areas of the Iberian Peninsula, the topographyis strongly influenced by the late Oligocene-to-Present development ofthe Western Mediterranean. NE-SW extensional faults related to thecrustal thinning between the Balearic Islands and the Iberian Peninsulaare preserved onshore, producing a descending topographic gradientfrom the elevated inner parts of Iberia to the Mediterranean coast.
The uppermost Cretaceous to lowermost Cenozoic rocks are theyoungest precontractional rocks preserved within the Iberian Chain
Fig. 4.Record section and theoretical travel time arrivals for shots T3, T2, and T1 (a, b, and c,are band-pass filtered between 3 and 20 Hz and displayed with normalized amplitude.constrained zones of the final velocity–depth model (a) Upper panel: record section and thpanel: record section and theoretical travel time arrivals for shot T1; lower panel: ray-tracinglower panel: ray-tracing coverage.
(Figs. 1 and 2). They are good markers of the overall structure of thechain and of the topography developed during the Cenozoic contrac-tion. It should be taken into account that previously to this episode,the uppermost Cretaceous rocks indicate a shift from marine environ-ments to continental with sporadic marine intercalations (Canérotet al., 1982; Gautier, 1980). This indicates that these rocks were formednear the sea level, which was at that time about 200 m above the pres-ent one. These Upper Cretaceous rocks are preserved in wide parts ofthe Iberian Chain, even in many of the more elevated areas, exceedinglocally 1800 m of altitude.
Cenozoic shortening in the Iberian Chain is estimated to varybetween 41 km (19%) in the NW across the Cameros area and 37 km(14%) in the central parts, along the sections shown in Fig. 2 (Guimerà,2013). The Cenozoic shortening is responsible for the uplift of theMesozoic and Lower Cenozoic rocks and the generation of the presentday topography of the Iberian Chain (Guimerà, 2013; Guimerà andGonzález, 1998) and for the crustal thickening, exceeding 40 kmbeneaththe chain (Salas and Casas, 1993).
3. Mapping the crust beneath the Iberian Chain from a refraction/wide-angle reflection seismic profile
In order to improve the knowledge of the IberianChain crustal archi-tecture, a 300 km long refraction/wide-angle reflection seismic profilewas carried out through the chain in December 2003 (see preliminaryreport in Gallart et al., 2004). The profile samples the area that showsa Bouguer gravity low in the range of −100 to −140 mGal by meansof three shots designed to obtain a reversed profile (Figs. 1 and 3). Theshot located at themid part of the linewas loadedwith 1 t of explosivesin a single hole, while the two others were detonated at the edges of theprofile loadedwith up to 1.5 t of explosives. All the shots were recordedby a linear array of 76 portable seismic stations from the ICTJA-CSIC,
respectively). Record sections are plotted using a 6 km/s reduction velocity. The tracesThe lower panels show the ray-tracing diagrams for each shot, depicting the well-eoretical travel time arrivals for shot T3; lower panel: ray-tracing coverage. (b) Uppercoverage. (c) Upper panel: record section and theoretical travel time arrivals for shot T2;
0 20 40 60 80 100 120 140 160 180 200 220 240 260 280 300 320distance (km)
-6
-5
-4
-3
-2
-1
0
1
2
3
4
5
6
7
8
time
(s)
T3
PiP
PgPs
PmP
Pn
a
c
-320 -300 -280 -260 -240 -220 -200 -180 -160 -140 -120 -100 -80 -60 -40 -20 0distance (km)
-6
-5
-4
-3
-2
-1
0
1
2
3
4
5
6
7
8
time
(s)
T1
PiP
PgPs
PmP
Pn
-160 -140 -120 -100 -80 -60 -40 -20 0 20 40 60 80 100 120 140 160distance (km)
-6
-5
-4
-3
-2
-1
0
1
2
3
4
5
6
7
8
time
(s)
PiP
Pg
Ps
PmP
PmP PiP
PgPgPg
Ps
T2
b
343H. Seillé et al. / Tectonophysics 663 (2015) 339–353
344 H. Seillé et al. / Tectonophysics 663 (2015) 339–353
Lithoscope and Iris Passcal pools, most of them equipped with 3-component sensors. The data were processed following a standardapproach, including time and amplitude verification for each individualtrace and gathering in a single Seg-Y file for each shot. Those files werethen displayed as record sections with a reduction velocity of 6 km/sand a band-pass Butterworth filter with corner frequencies at 3 and20 Hz was applied. The signal-to-noise ratio shows a reasonably goodoverall data quality.
Hereafter a new, detailed interpretation of this seismic profile cross-ing the Iberian Chain will be presented, confirming the existence of amoderate crustal thickening beneath the central part of the profile,where Moho depths exceeding 40 km is documented. This result isconsistent with available Bouguer gravity anomaly data (see below)and with crustal thickness recently inferred from teleseismic data.
The different seismic phases (shown and labeled in Fig. 4) are iden-tified using the conventional nomenclature: Ps, Pg, and Pn denoterefractions through the sedimentary cover, basement, and Moho dis-continuity, respectively; PiP and PmP stand for P to P reflections pro-duced at the top of the middle crust and Moho.
P-wave velocity–depth models were derived by forward modelingtravel times of diving and reflected waves using RAYINVR software(Zelt and Smith, 1992). The initial model was composed by a sedimen-tary cover directly retrieved from geological maps, overlying layersrelated to basement, upper crust,mid/lower crust andmantle. The geom-etry of the sedimentary coverwas retrieved fromgeological results, whilefor the rest of layers an initial flat interface was assumed.
Figs. 4a and c show the record sections from the shots T3 and T1,located respectively at the SW and NE edges of the profile. Up to offsetsof 40 km, clear first arrivals corresponding to Ps and Pg phases areobserved between 0 and 1 s of reduced time. In the 40–100 km offsetrange, the arrivals are interpreted as PiP phases reflected in a mid-crustal discontinuity. A clear time delay can be identified at offsets of60–80 km for the T1 shot evidencing variations along the profile inthe upper part of the crust. Between 100 and 200 km distances, mostconspicuous arrivals are observed between 4 and 2 s of reduced timesand correspond to PmP phases.
At record sections from shots T1 and T3 at the edges of the profileand for offset ranges between 100 and 150 km, the correlations ofarrivals from the bottom of the crust appear to be complex. Ratherthan a single PmP phase correlation, one can observe there (see Fig. 4aand c) different arrivals, delayed one another by ~0.2–0.4 s, which sug-gest a Moho complexity that we have turned into the model as tworeflective horizons separated ~3–4 km, with velocity contrasts around0.06–0.1 km/s. The rather low density of wide-angle measurements
Fig. 5. Final velocity–depth model along the profile. Color scale shows the corresponding P-wupper crust beneath the Iberian Chain basement involved thrust system.
along the profile does not allow us to refine such a model structure.These phases from the lowermost part of the crust arrive significantlylater for shot T1 than for shot T3; for example, at offsets of 160 km themain energy arrives at of 2.6 s for shot T1 and 1.9 s for shot T3. This ob-servation already suggests a certain amount of crustal thickening at thecentral part of the profile. It is also worthy to note that the energy ofthese phases vanishes quickly at distances of 200 km. At greater offsets,much less energy is present on the sections, although first arrivals delin-eating an apparent velocity around 8 km/s, and thus interpreted as Pnphases, could be nicely observed in selected sites, especially from shot1 (Fig. 4c). Shot T2, located at the central part of the profile, shows asignificant lateral difference in crustal structure between its two sides(Fig. 4b). The PiP phase arrives clearly later in the NE section, thus sug-gesting a thickening of the upper crust in that part of theprofile. Regard-ing PmP phases, the apparent velocity is higher to the NE than in theSW section of the profile hence suggesting also variations in the Mohogeometry along this transect.
The final model is presented at Fig. 5. The phases reflected at theuppermost sedimentary layer are reproduced properly assuming the apriori geometry and velocities ranging from 4.2 to 5.1 km/s. The base-ment depth ranges between 3 and 5 km along the profile, and it ismodeled with velocities between 5.1 and 5.4 km/s. In order to matchthe delay observed70kmwest of shot T1 for the Pg phases, a local thick-ening has to be introduced in themodel, reaching depths of 7–8 km. Theupper crust has velocities close to 6.0 km/s and a variable thicknessranging from about 5 km at the edges of the profile to 12 km around50 km NE of shot T2. The preferred model has a middle crust layerwith velocities ranging between 6.3 and 6.45 km/s and showing smoothvertical gradients.
In order to reproduce properly the double arrivals identified aroundthe PmP phases, a 3–4 km thin layer has been introduced in the modelto account for the crust–mantle transition, with velocities increasingfrom 7.0 to 8.0 km/s. Although such a layer could be regarded aslower crust, its unusual low thickness clearly favors an interpretationin which the Moho is not a sharp discontinuity but a relatively thicktransitional zone resulting in complex energy trends associated toPmP arrivals. In this preferred hypothesis, the thin layer overlying theMoho is just a way to model this complex crust–mantle transition. Arecent contribution from multichannel seismic in the Central Iberiandomain (Ehsan et al., 2014) has also reported a thin Moho reflectiveband (~1 s TWT) in the NE part of the sampled area, toward the CentralSystem. The crust–mantle transition on wide-angle reflection profilesin the same area is modeled as a 2–3 km thick zone with interbandingof lower crust and upper mantle materials (Ehsan et al., 2015).
ave velocities. Main superficial geological features are located. Note the thickening of the
345H. Seillé et al. / Tectonophysics 663 (2015) 339–353
The main result inferred from the velocity–depth modeling of thiswide-angle seismic profile is the presence of a 100 km wide crustalthickened area located NE of shot T2 (Fig. 5). This thickening is mainlycontrolled by the time differences of the PmP phases recorded forshots T1 and T3. Similar differences can also be identified among thePmP phases recorded toward both sides of shot T2, which providefurther constrains. In the final model, the top of the transitional zoneassociated with theMoho evolves from 30 km at the SE part of the pro-file to a maximum depth of about 40 km at offsets close to 200 km, tothen thin again to 30 km at the NE edge of the profile. The thickenedarea, located up to 70 kmNE of shot T2, is consistent with results arisingfrom passive seismic experiments (Mancilla and Diaz, 2015), as well aswith the gravity minimum observed in the Bouguer anomaly map(Fig. 3), which will be tested by gravity modeling in the next section.
Moreover, to fit the travel times of the PmPand Pn phases, and to ex-plain the abrupt decrease of the PmP energy at offsets beyond 200 km inshots T1 and T3, it has required to introduce in the model a local rise ofthe Moho boundary at distances of 110–150 km from the SW edge.
4. Gravity model
The lateral crustal thickness variations and the topography of theMoho inferred from the 300 km long refraction/wide-angle reflectionseismic profile (Fig. 5) are outstanding results that have been checkedagainst the observed gravity signature. The gravity anomaly in thestudy area (Figs. 3 and 6a) is characterized by a remarkably low from−100 to−140 mGal that extends from the NE segment of the Serraniade Cuenca to the Montalbán thrust. Low values in the range of −80 to−140 mGal are observed all along the NE-SW trending Iberian Chaincoinciding with the highest relief. The average elevation in the IberianChain is 845 m (Casas-Sainz and de Vicente, 2009), although alongthe study profile elevation can reach maximum values above 1600 m(e.g., central part of the model profile). Toward the NE and SW gravity,anomalies increase with a gentle gradient to values above−40mGal in
a
b
Fig. 6. Crustal densitymodel obtained by gravity forwardmodeling along the seismic transect shanomalymap (Fig. 3) along a 30 kmwide strip. The vertical bars denote standard variation at eaccrustalmodel shown in (b). Blue line indicates differences between the observed and calculated(Fig. 5). T1, T2 and T3 indicate shot-point location. Color key shows density values in kg/m3. S
the Ebro foreland basin and above−60mGal in the Loranca basin. Of in-terest is also the pronounced positive gradient toward the Neogene Va-lencia Trough (in the SE) indicating the crustal thinning from inland tothe axis of the Trough (Fig. 3).
To account for lateral variations to the strike of the profiles, gravitydata have been averaged over a strip ±30 km of the profile with theresulting standard deviation computed and displayed in Fig. 6a. Thegravity model has been built using the geometry and velocities derivedfrom seismic modeling. An averaged P-wave velocity value has beencalculated for each layer in the seismic model, and the correspondingdensitieswere calculated for the crystalline crust using the empirical re-lations by Christensen and Mooney (1995), while for the sedimentaryandMesozoic cover, we have used field samples andwell data collectedand presented in theworks of Gómez-Ortiz et al. (2005) and Pueyo et al.(in press).
Thus, the crust is modeled as sedimentary layers with densities thatrange from 2300 to 2400 kg/m3 for the Tertiary infill of the Loranca andEbro sedimentary basins and from 2450 to 2650 kg/m3 for theMesozoiccover, which is mainly composed of dolomites, gypsum and shales(Triassic), marine limestones and dolostones (Jurassic and Upper Creta-ceous), and continental and marine deposits (Lower Cretaceous). Forthe underlying Paleozoic basement, mainly formed by metamorphicshales, sandstones, and limestones, we have taken an average densityof 2750 kg/m3. For the upper crust, we have taken an average value of2800 kg/m3, 2830 kg/m3 for the middle crust, 2970 kg/m3 for the tran-sition zone, and 3330 kg/m3 for the lithospheric mantle.
Calculations of the densitymodel response are based on themethodsof Talwani et al. (1959), and Talwani and Heirtzler (1964), and thealgorithms described in Won and Bevis (1987). As observed in Fig. 6b,the modeled crustal density distribution is well compatible with theP-wave seismic velocity models, particularly from the upper crust touppermost mantle levels. To achieve a satisfactory fit to the gravity, thegeometry of the top of the upper crust had to be slightly modified fromca. 190 to 230 km to fit the positive gradient observed from km 200onward. Also, we had to slightly modify the geometry of the Mesozoic
own in Fig. 5. (a) 2Dprojection of the gravitymeasures extracted from the Bouguer gravityhmeasure point, and the red line denotes the calculated anomaly generated by thedensityanomaly. (b) distribution of the densitieswithin the crust, inferred from the seismicmodelee text for detailed explanation.
346 H. Seillé et al. / Tectonophysics 663 (2015) 339–353
and sedimentary layers, which are not so well constrained from seismicdata particularly away from shots location.
Unlike the mid-low crustal levels where velocities and densitiesare laterally fairly homogeneous, the upper levels of the crust show agreater heterogeneity both in the seismic section and on the gravityprofile. Thus, fitting the short to medium wavelength of the gravityanomaly requires slight lateral variations in the densities of the Tertiarysediments, from 2300 to 2400 kg/m3, and in the Mesozoic cover, from2450 to 2650 kg/m3. To the contrary, the upper and middle crustshows laterally a rather homogeneous P-wave velocity and densityprofile. At deeper levels, the most remarkable results are the presenceof a 5 km thin high-velocity (7.0 km/s) high-density (2970 kg/m3) tran-sitional layer: a crustal keel in theNEhalf of the profilewheremaximumcrustal thickness above 45 km are locally recorded, the local rise of theMoho boundary from 100 to 160 km, right underneath the Serrania deCuenca Thrust, and the gentle thinning of the crust toward the Loranca(Tajo) and Ebro basins, where the crust is about 35 km thick.
5. Magnetotelluric data
The MT data consists of 37 broad-band magnetotelluric sites alongtwoNE-SWprofiles crossing the tectonic structures andmain geologicalunits of the Iberian chain (Fig. 1). These datawere acquired during threecampaigns between 2011 and 2013 in the framework of the Topo-Iberiaproject. Space betweenMT sites along the two profiles is approximately5 to 7 km. The five componentsweremeasured at all sites, two horizon-tal electric and magnetic field components recorded in N-S and E-Wdirections and the vertical magnetic field component. Recording timewas between 24 h and 48 h at each site, and the transfer functionsobtained have periods that range from 0.001 s to 1000 s. The estimatesof the MT impedance and the geomagnetic transfer functions were ob-tained using standard robust processing method (Egbert and Booker,1986). The data were in general of good quality and only some sites,located close topopulated areas,were affected by cultural noise. Remotereference processing was not able to remove such noise from theresponses, leaving some gaps especially around 10 s.
5.1. Dimensionality analysis and induction vectors
The dimensionality and the preferred strike direction of two pro-files (Fig. 7) were studied using two approaches: (1) the multisite/multiperiod analysis (McNeice and Jones, 2001) based on the Groom
A06
A07
A08
A09
A10
A11
A12
A13
A14
A15
A17
A18
A19
A20
A21
A22
A23
A24
A25
A26
A27
A28
A29
1
2
3
4
log
10 [
per
iod
(s)]
-3-6-9-12-15
Skew (degrees
0 63
-2
-1
0
-3Profile A
aSW
Fig. 7. Phase tensor ellipses of the observed data. The ellipses are normalized by φmax and filleoriented. (a) Phase tensor ellipses of the observed data for profile A. (b) Phase tensor ellipses o
and Bailey (1989) tensor decomposition and (2) the phase tensor anal-ysis (Caldwell et al., 2004).
Following the multisite/multiperiod analysis, the strike directionwas calculated in a period range for all sites by minimizing the globalχ2misfit between experimental and theoretical Groom–Bailey imped-ance tensor. For the analysis of both profiles, an error floor of 13%had to be applied to the data in order to obtain a χ2 error within the95% confidence interval for a period range of 1–1000 s. A strike directionof N130° for the profile A and N125° for the profile B were obtainedin the analyses. The necessary high error floor was a consequence of(1) some of the sites were affected by noise, mainly in the periodrange between 4 and 40 s; (2) clear 3D effects in a number of sites. In-duction arrows that represent the complex ratio between vertical andhorizontal magnetic fields can be used to infer lateral conductivity var-iations. In a 2D case, the induction arrows should be perpendicular tothe geoelectric strike. We show in Fig. 8 the real induction arrows inthe Wiese convention (pointing away from the conductor). At longperiods (between 100 s and 1000 s), the vectors in the southernpart of the profile point to NW, being parallel to the supposed strikedirection. This is an indication of off-profile conductivity variationsthat could not be considered using a 2D inversion scheme. The presenceof the Mediterranean Sea to the SE of the profile, at a distance between90 km and 130 km from the sites, was tested through 3D forwardmodeling and resulted not to be responsible for the behavior affect-ing the inductions arrows, i.e., arrows pointing to the NW with highmagnitudes.
The phase tensor is defined by the relation φ = X−1 Y, betweenthe real (X) and imaginary (Y) part of the impedance tensor, respec-tively. As any second rank tensor, the phase tensor can be represent-ed graphically as an ellipse, which is defined by three invariants ofthe phase tensor: the principal axes φmax, φmin, showing the maxi-mum and minimum phase difference between the magnetic andelectric fields and a third coordinate invariant parameter │β│ (theskew angle) that represents the asymmetry of the phase response.The skew angle │β│ provides a measure of the significance of 3Deffects in the MT phase response. Fig. 7 shows the phase tensor ellip-ses of the observed data as pseudosections for both profiles. It wasobtained using the MTpy python toolbox (Krieger and Peacock,2014). The color used to fill the ellipses shows that low │β│ values(b6°) are generally present at high frequencies (b1 s) but increases togreater values at long periods, and at sites/areas where data quality islower. Some sites present great │β│ values (N10°) at high frequencies.
B01
B02
B03
B04
B05
B06
B07
B08
B09
B10
A02
A03
A04
A05
)
15129
Profile B
Profile B
Profile A
North
b c Profiles Map NE SW NE
d with the skew angle β. The ellipses are plotted so that the vertical axes are north–southf the observed data for profile B. (c) Location of two profiles, black dots are the MT sites.
-1° -2° -1°
40°
41°
-2° -1°
40°
41°
T=465 s
41°
TT=100 s
-2° -1°
40°
-1°
Re=1 - Calculated
Re=1 - Observed
Re=1 - Calculated
Re=1 - Observed
Re=1 - Calculated
Re=1 - Observed
Re=1 - Calculated
Re=1 - Observed
-2°
40°
41°
T=1000 s
-2°
40°
41°
T=217 s
Fig. 8. Real induction vectors for 4 selected periods (Wiese convention). Data (black) and responses of the 3D inverse model (red).
347H. Seillé et al. / Tectonophysics 663 (2015) 339–353
The criterion for which the data is considered to be affected by 3Dstructures is reached when │β│ N 3° (Caldwell et al., 2004). This com-plex behavior suggests a multidimensional character, localized andregional.
The dimensionality analysis carried out shows that both profilespresent 3D effects. In this context, the 3D interpretation of the datawas thought to bemore suitable, in order to accuratelymodel 3D effectsthat could bias a 2D interpretation of the data sets. Accordingly, we pro-ceed with 3D inversion of the whole data set.
5.2. 3D Magnetotelluric inversion
The 3D inversion was carried out using on the 3D non-linear con-jugate gradient (NLCG) algorithm (Rodi and Mackie, 2001, 2012) sub-sequently implemented by Schlumberger. The algorithm minimizesthe misfit between observed and computed data, using a regularizationoperator to produce a smoothly varying resistivity volume.
In order to optimize the starting model and the mesh used in theinversion and to test the influence of several factors, we performedfirst several forward modeling. The presence of the topography andthe Mediterranean Sea to the East slightly affected the responses andwas included in the starting model. This effect consists on a moderatesplit of both polarizations for the phases at long periods (around1000 s) and small real induction arrows pointing to the west. Themeshwas rotated to N40°E in order to align with the profiles, to reducethe cells number of themesh and to align it more closely to the orienta-tion of the Iberian Chain. The data were accordingly reproduced for acoordinate system with the x-axis oriented in direction N40°E.
The cell size in the central part of the mesh has horizontal dimen-sions of 2000 m × 2000 m, and a thickness of 100 m in the superficialarea to define the topography. Themeshwas defined in order to have1–3 cells between neighboring sites. Outside of the central area, themesh extends long enough laterally and in depth to respect theboundary and continuity conditions. The mesh has a dimension of1324 km × 717 km × 742 km in the x, y, and z directions, respectively(as mentioned before, the mesh has been rotated and the x direction
is oriented N40°E). The total number of cells of the mesh is 483,084cells (126 × 54 × 71).
The startingmodel is a homogeneousmodel of 100Ωmthat includestopography and theMediterranean Sea to the East, defined by a resistiv-ity value of 0.33 Ωm. The full impedance tensor (all four complex com-ponents) and the magnetic transfer function were used in the 3Dinversion. Sixteen frequencies between 0.01 s and 1000 swere inverted.
Several inversions were carried out with different error floors andregularizers. The final model was obtained using error floors of 2% forthe Zxy component of the impedance tensor and 3% for the Zyx compo-nent. Rotating the data to N40E we noticed that the YX component wasmore affected by noise than the XY, accordingly we decided to give it aslight higher error floor. For the magnitude of the diagonal elementsof the impedance tensor an error floor of 10% was applied and for themagnetic transfer function tensor the absolute error applied was 0.05.The last inversion was started using a Lagrange multiplier tau equal to0.1. In a second step, we followed with the inversion reducing tau to0.05. Reducing the Lagrange multiplier permitted to improve the datafit in the final iterations. Several inversions with different tau valueswere tested, and this workflow resulted to be the one that was givingthe best RMS obtaining a smoothmodel that was geologically plausible.The inversion using a tau equal to 0.1 converged after 52 iterations. Inthe second step using a Lagrange multiplier equal to 0.05, the conver-gence occurred after 13 more iterations. In total after 63 iterations, theinversion reached a RMS of 1.88. All sites present a good RMS, withouthaving any region poorly fitted. The resistivity distribution obtained inthe shallow part of themodel was able to fit well the data and no signif-icant shift appeared in the apparent resistivity responses comparedwith the data.
Fig. 9 shows the data andmodel responses pseudosections of the off-diagonal components along two profiles A and B grouping the sites(Fig. 7c). All the components of the impedance tensor are well fitted,with the exception of the diagonal components at high frequencies. InFig. 10, we show the fitting of four sites, showing both the diagonal andthe off-diagonal components of the impedance tensor. The calculated in-duction arrows (Fig. 8) fit the behavior of themagnetic transfer function,
12
3
log1
0 [p
erio
d(s)
]
-2-1
0
12
3
-2-1
0
12
3
-2-1
0
12
3
-2-1
0
12
3
-2-1
0
12
3
-2-1
0
12
3
-2-1
0
12
3
-2-1
0
12
3
-2-1
0
12
3
-2-1
0
12
3
-2-1
0
12
3
-2-1
0
12
3
-2-1
0
12
3
-2-1
0
12
3
-2-1
0
12
3
-2-1
0log1
0 [p
erio
d(s)
]
0 40 80 120 160Distance [km]
0 40 80 120 160Distance [km]
0 40 80 120 160Distance [km]
0 40 80 120 160Distance [km]
0 40 60Distance [km]20 0 40 60
Distance [km]20
0 40 60Distance [km]20 0 40 60
Distance [km]20
Observed Profile A Calculated Profile A Observed Profile B Calculated Profile B
Pha
se [°
]A
pp. R
es [o
hm.m
]
0
30
60
90
5
30
170
1000
Pha
se [°
]A
pp. R
es [o
hm.m
]
0
30
60
90
5
30
170
1000
YX
XY
Fig. 9. Pseudosections (data and response of the 3D inverse model) of the off-diagonal components along the two profiles A and B shown in Fig. 7c.
348H.Seillé
etal./Tectonophysics663
(2015)339–353
-1
0
1
2
3
4
Log 10
App
.R
es.
[ohm
.m]
A06XY YX
-3
-2
-1
0
1
2
Log 10
App
.R
es.
[ohm
.m]
A06XX YY
-2 -1 0 1 2 30
45
90
Log10 T [s]
Pha
se[d
egre
e s]
-2 -1 0 1 2 3-180
-90
0
90
180
Log10 T [s]
Pha
se[d
egre
es]
-1
0
1
2
3
4
Log 10
App
.R
es.
[ohm
.m]
A21XY YX
-3
-2
-1
0
1
2
Log 10
App
.R
es.
[ohm
.m]
A21XX YY
-2 -1 0 1 2 30
45
90
Log10 T [s]
Pha
se[d
egre
es]
-2 -1 0 1 2 3-180
-90
0
90
180
Log10 T [s]
Pha
se[d
egr e
e s]
-1
0
1
2
3
4
Log 10
App
.R
es.
[ohm
.m]
B01XY YX
-3
-2
-1
0
1
2
Log 10
App
.R
es.
[oh m
.m]
B01XX YY
-2 -1 0 1 2 30
45
90
Log10 T [s]
Pha
se[d
egr e
e s]
-2 -1 0 1 2 3-180
-90
0
90
180
Log10 T [s]
Pha
se[d
egre
e s]
-1
0
1
2
3
4
Log 10
App
.R
es.
[ohm
.m]
B09XY YX
-2
-1
0
1
2
3
Log 10
App
.R
es.
[ ohm
.m]
B09XX YY
-2 -1 0 1 2 30
45
90
Log10 T [s]
Pha
se[d
egre
es]
-2 -1 0 1 2 3-180
-90
0
90
180
Log10 T [s]
Pha
se[d
egre
es]
RMS=1.68 RMS=2.68
RMS=2.2 RMS=1.29
Fig. 10. The four components of the impedance tensor at four selectedMT sites. Data (points) and 3Dmodel responses (continuous line). The components are rotated toN40°E (see text forexplanation).
349H. Seillé et al. / Tectonophysics 663 (2015) 339–353
especially the deviation of the induction arrows from the profile lines inthe southern part of the profiles.
5.3. 3D resistivity model
The model is presented in Fig. 11, where several cross sections fromthe 3Dmesh, oriented NW-SE (X sections) and NE-SW (Y sections), areshown alongwith their location on the geologicalmap of the area. In thefirst kilometers of the model several conductors and resistors coincidewith geological features such as Tertiary basins (low resistivities),Palaeozoic basement (high resistivities) outcrops, and known faultzones (elongated conductors). At greater depths, from southwest tonortheast, the model is divided into three zones. From 0 to 50 km, be-neath de Serrania de Cuenca Thrust, the mid–lower crust is moderatelyconductive (less than 100Ωm), the central zone, from 50 to 150 km, themiddle-lower crust is homogeneous and highly resistive (higher than1000 Ωm) and in the northern zone, from 150 km to the NE edge,beneath the Aragonese Branch Arc and the North Iberian Thrust, thecrust is moderately conductive (less than 100Ωm). The causes invokedfor the high conductivity at mid–lower crust are the presence of aque-ous fluids, partial melting and conductive mineralization. In the IberianChain, no evidences for other causes than fluids occur. Therefore, thehomogeneously high resistive mid–lower crust in the central zonereveals low porosities and no fluids circulation. By contrast, only in thenortheast and in the southwest zones, themain faults, the North Iberianthrust in the NE, and the Serrania de Cuenca Thrust in the SW, whichdominate the structure of the Iberian Chain, reach greater depths.
Main conductive anomalies identified in these cross sections areindicated by letters and correlated to the geological structures presentin surface (Fig. 11). A is a mid-crustal conductor located at 25 kmdepth, which fades toward the SE. B is an upper crustal conductor dip-ping southwestward and reaching a depth of 10 km, as shown in theY16 to Y35 cross sections. It correlates with the southwest dippingthrust fault located above sections Y35, Y32, and Y29, the latter coincid-ing with a change in the thrust orientation as it appears on the surfaceover section Y22. C is a shallow conductive anomaly present in all theY sections reaching 6 km depth. This conductor correlates with thecore of the Almazan Syncline (see also Fig. 1). D has been subdividedinto two different conductors, D1 and D2. Conductor D1, located in sec-tion Y29, dips southwestward reaching 10 kmdepth and correlates wellwith the southwestward dipping thrust observed on surface. More tothe east, on sections Y32 and Y35, this conductor changes and dipsnortheastward (conductorD2) coincidingwith the dipping of the thrustobserved on surface, which dips northeastward (Montalbán thrust) inthis eastern part of the geological map (Fig. 1). This D2 conductorreaches a depth of 10 km. Section X89, located in the north part of themodel, shows that the set of both conductors (D1–D2) is also extending(dipping) eastward, revealing a complex 3D behavior of the volumewith high fluid content related to the complex system faults in thisarea. Note in Fig. 1 over sites A9 to A12, there is a convergence of anumber of thrusts dipping in different directions. Conductor E, in thenortheastern part of the model, is visible in sections Y29 to Y35. It dipssouthwestward reaching a depth of 10 kmprobably being the signatureat depth of the North Iberian thrust, located north of the studied area
A
Y32
Y35 X17
X36
X48
N40°ENorth
X36X48
X70
X89
X17
Upper Jurassic andLower Cretaceous
Neogene
Palaeogene
Upper Cretaceous
X17Y29Y22
Y32Y35
Y16
B
D1
E
Y29Y16
Y16
Y16
Y16
Y22
010203040D
epth
(km
)
A
A
B
B
B
C
C
C
C
C
B
C
E
E
D2
F
E
D2
F
E
D2
F
E
B
D2
D1
C
50
100
150
0
50
100
150
100
150
X17
02040
2040
40
(km) Distance (km)
X70
X89
010203040 D
epth
(km
)
D2
F
E
F
E
D
150
150
C
D2
F
Fig. 11. 3D resistivity model: Y are NE-SW sections and X are NW-SE sections. Dashed lines link the corner of each section to its position on the geological map. Red dots are the MT sites, letters indicate the main conductors.
350H.Seillé
etal./Tectonophysics663
(2015)339–353
351H. Seillé et al. / Tectonophysics 663 (2015) 339–353
(Fig. 1). The mid-crustal conductor F located at 15–20 km depth ispresent in section Y35. This conductor merges with the northeastwarddipping conductor D2 and the southwestward dipping conductor Esuggesting the branching of the Montalbán thrust to the North IberianThrust.
Several others resistive and conductive features correlate with thesurface geology (Fig. 11). The Palaeozoic outcrops (Variscan basement)observed in the geological map are characterized by shallow resistivebodies. Their outcropping is generally a consequence of thrusts, dippingnorth or south. The Neogene sediments are characterized by shallowconductive bodies. The correlation between the surface geology andtheMT conductive-resistive features gives us confidence on the reliabil-ity of the anomalies and their structures at depth, allowing us to inter-pret the shape and extension of the faults in the crust.
Inversion of the full impedance tensor jointly with the magnetictransfer function constrains the model at shallow and intermediatedepths. Using small error floors and having reached a very good fittingfor all the sites give us confidence on the sensitivity of the data to thefeatures found in the lower parts of the model. However, the mainfeatures presented at depth need to be tested, in order to concludethat our data are sensitive to those structures, and that they are notdue to artifact of the inversion process or to data noise fitting. Testingthe sensitivity of the structures was done removing separately eachone of the conductors appearing in the model, substituting them byresistivity values surrounding the conductors. To account for theirpresence, we ran the forward modeling and compared the responsesof the different models. As a result, we conclude that (1) the modelloses resolution beneath a depth of 30 km, (2) the conductor A locatedat a depth of 25 km is not well resolved given that it is located beneatha shallow conductor (the one at 8–10 km depth, section Y22 in Fig. 11)and it presents an average resistivity of 100 Ωm in front of 1000 Ωm ofthe surrounding areas, and (3) all the other features at middle shallowdepths are well resolved.
6. Discussion and conclusions
The multidisciplinary study integrating seismic refraction/wide-angle reflection, magnetotellurics, and gravity modeling allowed us tobetter constraint the crustal and uppermost mantle structure alongthe modeled profiles. Since they have different depth resolutions and
−60
−50
−40
−30
−20
−10
0
Dep
th (
km)
0 20 40 60 80 100 120 140
Distan
SWT3
Serrania de CuencThrustTajo Basin
Ib
Mesozoic-Cenozoiccover
0
B
Fig. 12. Contours of the seismic velocity model superimposed on the Y32 section of the 3D resiPalaeozoic.
are sensitive to different physical parameters they, consequently, com-plement each other.
The magnetotelluric method provides the electrical conductivity dis-tribution at depth. The 3DMT inversion images the subsurface from shal-low tomiddle-lower crust, taking into account the sensitivity test and theperiod range used. The orientation and depth of the main faults imagedare consistent with the geological structures observed at surface.
At greater depths, the Moho discontinuity is not resolved by MTsince theMoho does not have a significant change in the electrical resis-tivity (e.g., Jones, 2013). Seismic refraction method allows us to detectthe discontinuity between crust and uppermantle. TheMoho transitionis generally very clear, marked by a notable increase in the seismicvelocity. However, in our case, the sparsity of the stations along theprofile and the low sensitivity of the seismic refraction technique tolateral variations do not constraint the shallow structures. Therefore,the integration of seismic refraction and magnetotellurics was optimalin order to obtain a complete image of the Iberian Chain at crust andupper mantle depths.
After the resistivity model, the Mesozoic and Cenozoic cover(e.g., the Calatayud–Montalban Tertiary basin, with resistivity valuesfrom 1 to 50 Ωm) can be distinguished from the Variscan Basement(with resistivity greater than 200 Ωm). Variscan rock outcrops(Montalbán Variscan Basement) and Cenozoic basins are clearlydepicted. The North Iberian thrust, which bounds to the north thebasement involved areas of the Iberian Chain, can be recognized by alow resistivity zone emerging at the northern end of the profiles(Figs. 11 and 12, conductors E and F). This major thrust reaches adepth of 15 km andmerges with the conductor D2 (Fig. 11), suggestingthe branching of the Montalbán thrust to the North Iberian Thrust asshown in Fig. 12. The Serranía de Cuenca Thrust, which bounds to thesouth the basement involved areas, crops out south of the MT sitesand it is not well recognized in the resistivity model. However, the con-ductor beneath sites A29 (Fig. 12) can be attributed to this thrust, thenreaching 7 km depth. The Almazán Syncline megastructure appearsclearly depicted (Figs. 11 and 12), correlating with themaximum thick-ness of the shallow conductive area C (Fig. 11).
The reversed refraction/wide-angle reflection seismic profile acrossthe Iberian Chain (Fig. 5) has revealed a moderate crustal thickening,exceeding 40 km, beneath its central part, having 33 and 32 km in itsSW and NE ends. The thicker crust is needed to support the Iberian
160 180 200 220 240 260 280 300
ce (km)
NET2 T1
Almazánsyncline Pz North Iberian
Thrust
aPz basement
erian Chain basement involved thrust system
50 100 150
Montalbán Thrust
F
D2EC
Log Resistivity(ohm·m)
0 1 2 3
Y32
stivity model. Letters are the conductors of Fig. 11. Red lines are the interpreted faults. Pz:
352 H. Seillé et al. / Tectonophysics 663 (2015) 339–353
Chain topography, and is consistent with the minimum observed in theBouguer gravity anomaly map (Figs. 3 and 6) a result which also is inagreement with the work presented by Rivero et al. (2008) andGuimerà et al. (Submitted for publication ). The 30 km thick crust be-neath the Iberian Chain-Ebro Basin boundary (Fig. 1) is in agreementwith the 34 km deep seismic Moho beneath the northern Ebro basincomputed in the ECORS deep reflection seismic profile (Roure et al.,1989). Note that the crustal thickening is mostly concentrated in theupper crust (Fig. 5), being consistent with the MT and gravity results,which show that the Cenozoic thrusts involve only the upper crust aswas previously proposed by Guimerà and Álvaro (1990), who consid-ered a thrust system detachment at 7–11 km deep.
The pattern of the seismic arrivals reflected in the Moho suggeststhat the limit between the crust and the upper mantle is not a sharpboundary but a relatively thick transitional zone. A local rise of theMoho boundary (30 km deep) at distances of 110–150 km from theSW edge of the profile needs to be introduced in the seismic velocity/depth model to adjust the seismic data. This feature does not have anytopographic expression, nor in the superficial geological structure. Ten-tative interpretations include an origin related with a relict feature ofthe Variscan crust, with the late Jurassic–Early Cretaceous extensionor, alternatively, with the compressional tectonics during the Paleogeneinversion of the rifted areas. Further research is required to provide asatisfying interpretation for the enigmatic local rise of the Moho andits possible connection with conductor A (Figs. 11 and 12).
Summing up our conclusions, the major thrusts depicted after MTdata reach a maximum depth of 15 km and after the seismic data thecrustal thickening is concentrated in the upper crust. This is consistentwith a Cenozoic thrust system involving only the upper crust. Themod-erate crustal thickening, reaching around 40 kmbeneath its central part,is consistent with results arising from passive seismic experiments(Mancilla and Díaz, 2015) and with the minimum observed in theBouguer gravity anomaly map. The results from the interpretation ofthese complementary geophysical data sets, deep seismic profiles,magnetotellurics, and gravity provided the first images of the crustalstructure of the Iberian Chain. They are consistent with a Cenozoicshortening responsible of the upper crust thickening as well as ofthe uplift of the Iberian Chain and the generation of its present daytopography.
Acknowledgments
Funding for the research was provided by the I + D + I researchprojects: BTE-2001-4993-E, CGL2008-04916, and by the Consolider-Ingenio 2010 programme, under CSD 2006-00041 “Topo-Iberia.” Wethank Schlumberger IEM CoE for allowing us to carry out the 3D MTinversion using Schlumberger’s proprietary 3D MT NLCG algorithm.Constructive reviews by two anonymous reviewers and the GuestEditor Manel Fernandez are greatly appreciated.
References
Almeida, E., Monteiro Santos, F., Mateus, A., Heise, W., Pous, J., 2005. Magnetotelluricmeasurements in SW Iberia: new data for the Variscan cristal structures. Geophys.Res. Lett. 32, L08312. http://dx.doi.org/10.1029/2005GL022596.
Anadón, P., Roca, E., 1996. Geological setting of the Tertiary basins of Northeast Spain. In:Friend, P.F., Dabrio, C.J. (Eds.), Tertiary basins of Spain. Cambridge University Press,pp. 43–48.
Ayala, C., Bohoyo, F., Maestro, A., Reguera, M.I., Rubio, F.M., Torné, M., Fernàndez, M.,García-Lobón, J.L., 2015n. Updated Bouguer anomaly map of the Iberian peninsula:a new perspective to interpret the regional geology. J. Maps (Submitted forpublication)
Banda, E., Suriñach, E., Aparicio, A., Sierra, J., Ruiz De La Parte, E., 1981. Crust and uppermantle structure of the central Iberian Meseta (Spain). Geophys. J. R. Astron. Soc.67, 779–789.
Caldwell, T.G., Bibby, H.M., Brown, C., 2004. The magnetotelluric phase tensor. Geophys.J. Int. 158, 457–469.
Campanya, J., Ledo, J., Queralt, P., Marcuello, A., Liesa, M., Muñoz, J.A., 2012. Newgeoelectrical characterisation of a continental collision zone in the west-central
Pyrenees: constraints from long period and broadband magnetotellurics. EarthPlanet. Sci. Lett. 333–334, 112–121.
Canérot, J., Cugny, P., Pardo, G., Salas, R., Villena, J., 1982. Ibérica Central-Maestrazgo. ElCretácico en España. Universidad Complutense, Madrid, pp. 273–344.
Carbonell, R., Simancas, F., Juhlin, C., Pous, J., Pérez-Estaún, A., Gonzalez-Lodeiro, F.,Muñoz, G., Heise, W., Ayarza, P., 2004. Geophysical evidence of a mantle derivedintrusion in SW Iberia. Geophys. Res. Lett. 31, L11601. http://dx.doi.org/10.1029/2004GL019684.
Casas-Sainz, A.M., de Vicente, G., 2009. On the tectonic origin of the Iberian topography.Tectonophysics 474 (1-2), 214–235. http://dx.doi.org/10.106/j.tecto.2009.01.030.
Choukroune, P., Ecors-Pyrenees Team, 1989. The ECORS Pyrenean deep seismic profile.Reflection data and the overall structure of an orogenic belt. Tectonics 8 (1), 23–39.
Christensen, N.I., Mooney, W.D., 1995. Seismic velocity structure and composition of thecontinental crust: a global view. J. Geophys. Res. 100 (B7), 9761–9788.
Daignières, M., Gallart, J., Banda, E., 1981. Lateral variation of the crust in the NorthPyrenean Zone. Ann. Geophys. 37 (3), 435–456.
Dañobeitia, J.J., Arguedas, M., Gallart, J., Banda, E., Makris, J., 1992. Deep seismic configura-tion of the Valencia Trough and its Iberian and Balearic borders from extensiverefraction-wide angle reflection seismic profiling. Tectonophysics 302, 37–55.
Díaz, J., Gallart, J., 2009. Crustal structure beneath the Iberian Peninsula and surroundingwaters: a new compilation of deep seismic sounding results. Phys. Earth Planet. Inter.173, 181–190.
E.C.O.R.S. Pyrenean Team, 1988. The ECORS deep reflection seismic survey across thePyrenees. Nature 331, 508–511.
Egbert, G., Booker, J.R., 1986. Robust estimation of geomagnetic transfer functions.Geophys. J. R. Astron. Soc. 87, 173–194.
Ehsan, S.A., Carbonell, R., Ayarza, P., Martí, D., Pérez-Estaún, A., Martínez-Poyatos,D.J., Simancas, J.F., Azor, A., Mansilla, L., 2014. Crustal deformation styles alongthe reprocessed deep seismic reflection transect of the Central Iberian Zone(Iberian Peninsula). Tectonophysics 621, 159–174. http://dx.doi.org/10.1016/j.tecto.2014.02.014.
Ehsan, S.A., Carbonell, R., Ayarza, P., Martí, D., Martínez-Poyatos, D.J., Simancas, J.F., Azor,A., Ayala, C., Torné, M., Pérez-Estaún, A., 2015. Lithospheric velocity model acrossthe Southern Central Iberian Zone (Variscan Iberian Massif): the ALCUDIA wide-angle seismic reflection transect. Tectonics 34, 535–554. http://dx.doi.org/10.1002/2014TC003661.
Ehsan, S.A., Carbonell, R., Ayarza, P., Martí, D., Martínez Poyatos, J., Marzán, I., Pérez-Estaún, A., 2015. The structure and nature of the Moho beneath Central Iberia.Tectonophysics 663, 277–291.
Gallart, J., Banda, E., Daignieres, M., 1981. Crustal structure of the Paleozoic Axial Zoneof the Pyrenees and transition to the North Pyrenean Zone. Ann. Geophys. 37 (3),457–480.
Gallart, J., Vidal, N., Dañobeitia, J.J., 1995. Multichannel seismic image of the crustalthinning at the NE Iberian margin combining normal and wide angle reflectiondata. Geophys. Res. Lett. 22 (4), 489–492.
Gallart, J., Salas, R., Guimerà, J., Mas, R., Díaz, J., Ruiz, M., 2004. A refraction/wide-anglereflection seismic profile through the Iberian Chain: preliminary report. Geo-Temas6 (2), 183–186.
Gautier, F. (1980): Mapa geológico de España 1:50 000. Núm 543 (Villarluengo), 2ª serie,1ª ed. Memoria 45 p. IGME, Servicio de Publicaciones del Ministerio de Industria yEnergía. Madrid.
Gómez-Ortiz, D., Tejero-López, R., Babín-Vich, R., Rivas-Ponce, A., 2005. Crustal densitystructure in the Spanish Central System derived from gravity data analysis (CentralSpain). Tectonophysics 403 (1), 131–149.
Groom, R.W., Bailey, R., 1989. Decomposition of magnetotelluric impedance tensor inthe presence of local three-dimensional galvanic distortion. J. Geophys. Res. 94,1913–1925.
Guimerà, J., 2004. Cadenas con cobertera: Las Cadenas Ibérica y Costera Catalana. In: Vera,J.A. (Ed.)Geologia de España, pp. 602–615.
Guimerà, J., 2013. Estructura y evolución tectónica de la Cadena Ibérica. In: Simón, J.L.,Liesa, C.L. (Eds.), XLVII Curso de Geología Práctica. Teruel 15-20 de Jul. 2013.Universidad de Verano de Teruel, pp. 1–9.
Guimerà, J., Álvaro, M., 1990. Structure et évolution de la compression alpine dans laChaîne Ibérique et la Chaine côtière catalane (Espagne). Bull. Soc. Geol. Fr. 6 (2),339–348.
Guimerà, J., González, A., 1998. El relieve de la Cadena Ibérica como un producto delacompresión alpina. Geogaceta 24, 163–166.
Guimerà, J., Mas, R., Alonso, A., 2004. Intraplate deformation in the NW Iberian Chain:Mesozoic extension and Tertiary contractional inversion. J. Geol. Soc. Lond. 161,291–303.
Guimerà, J., Rivero, Ll, Salas, R., Casas, A., 2015. Moho depth inferred from gravity andtopography in the Iberian Chain (Submitted for publication).
Jones, A.G., 2013. Imaging and observing the electrical Moho. Tectonophysics 609,423–436. http://dx.doi.org/10.1016/j.tecto.2013.02.025.
Krieger, L., Peacock, J.R., 2014. MTpy: a Python toolbox for magnetotellurics. Comput.Geosci. 72, 167–175.
Mancilla, F., Díaz, J., 2015. High resolution Moho topography map beneath Iberia andnorthern Morocco from receiver function analysis. Tectonophysics 663, 205–213.
McNeice, G., Jones, A.G., 2001. Multisite, multifrequency tensor decomposition ofmagnetotelluric data. Geophysics 66, 158–173.
Monteiro Santos, F.A., Pous, J., Almeida, E.P., Queralt, Marcuello, Matias, H., Mendes-Victor, L.A., 1999. Magnetotelluric survey of the electrical conductivity of the crustacross the Ossa Morena Zone and South Portuguese Zone suture. Tectonophysics313, 449–462.
Muñoz Martin, A., de Vicente, G., 1998. Cuantificación del acortamiento alpino yestructura profunda del extremo sur-occidental de la Cordillera Ibérica (Sierras de
353H. Seillé et al. / Tectonophysics 663 (2015) 339–353
Altomira y Bascuñana). Revista de la Sociedad Geológica de España 11 (3-4),233–252.
Muñoz, G., Mateus, A., Pous, J., Heise, W., Monteiro Santos, F., Almeida, E., 2008.Unravelling middle-crust conductive layers in Paleozoic Orogens through 3D model-ing of Magnetotelluric data: The Ossa-Morena Zone case study (SW IberianVariscides). J. Geophys. Res. 113, B06106. http://dx.doi.org/10.1029/2007JB004987.
Pedreira, D., Pulgar, J.A., Gallart, J., Díaz, J., 2003. Seismic evidence of Alpine crustal thick-ening and wedging from the western Pyrenees to the Cantabrian Mountains (northIberia). J. Geophys. Res. 108, 2204. http://dx.doi.org/10.1029/2001JB001667 (B4).
Pous, J., Ledo, J., Marcuello, A., Daignères, M., 1995. Electrical resistivity model of the crustand upper mantle from an MT Survey through the Central Pyrenees. Geophys. J. Int.121, 750–762.
Pous, J., Queralt, P., Ledo, J., Roca, E., 1999. A high electrical conductive zone at lower crustaldepth beneath the Betic Chain (Spain). Earth Planet. Sci. Lett. 167, 35–45.
Pous, J., Queralt, P., Marcuello, A., 2001. Magnetotelluric signature of the WesternCantabrian Mountains. Geophys. Res. Lett. 28, 1795–1798.
Pous, J., Muñoz, G., Heise, W., Melgarejo, J.C., Quesada, C., 2004. Electromagnetic imagingof Variscan crustal structures in SW Iberia: the role of interconnected graphite. EarthPlanet. Sci. Lett. 217, 435–450.
Pous, J., García Poyatos, D., Heise, W., Monteiro Santos, F., Galindo, J., Ibarra, P., Pedrera, A.,Ruiz-Constán, A., Anahnah, F., Goncalves, R., Mateus, A.A., 2011. Constraints on thecrustal structure of the internal Variscan Belt in SW Europe: a magnetotellurictransect along the eastern part of Central Iberian Zone (Iberian massif). J. Geophys.Res. http://dx.doi.org/10.1029/2010JB007538.
Pueyo, E., Izquierdo-Lavall, E., Rodríguez-Pintó, A., Rey, C., Oliva-Urcia, B., Casa, A., Calvín,P., Ayala, C., Rmajo, J., del Rio, P., Rubio, F., García-Lobón, J., 2015. Map of petrophysicalproperties in the Iberian Range and sorroundign áreas (NE-Spain): 1-Density. J. Maps(in press).
Rivero, L., Guimerà, J., Salas, R., Casas, A., 2008. Constraints on the Moho depth in theIberian Chain. Geo-Temas 10, 673–675.
Rodi,W.,Mackie, R.L., 2001. Nonlinear conjugate gradients algorithm for 2-Dmagnetotelluricinversions. Geophysics 66, 174–187.
Rodi, W.L., Mackie, R.L., 2012. The inverse problem. In: Chave, A.D., Jones, A.G. (Eds.), TheMagnetotelluric Method, Theory and Practice. Cambridge Univ. Press, Cambridge,pp. 347–420.
Roure, F., Choukroune, P., Berástegui, X., Muñoz, J.A., Villien, P., Matheron, P., Bareyt, M.,Séguret, M., Camara, P., Déramond, J., 1989. ECORS deep seismic data and balancedcross sections: geometric constraints on the evolution of the Pyrenees. Tectonics 8,41–50.
Ruiz-Constán, A., Pedrera, A., Galindo-Zaldívar, J., Pous, J., Arzate, J., Roldán-García, F.J.,Marin-Lechado, C., Anahnah, F., 2012. Constraints on the frontal crustal structure
of a continental collision from an integrated geophysical research: the central–western Betic Cordillera (SW Spain). Geochem. Geophys. Geosyst. 13, Q08012.http://dx.doi.org/10.1029/2012GC004153.
Salas, R., Casas, A., 1993. Mesozoic extensional tectonics, stratigraphy, and crustalevolutionduring the Alpine cycle of the eastern Iberian basin. Tectonophysics 228,33–55.
Salas, R., Guimera, J., Mas, R., Martin-Closas, C., Melendez, A., Alonso, A., 2001. Evolution ofthe Mesozoic Central Iberian Rift system and its Cenozoic inversion. In: Ziegler, P.A.,et al. (Eds.), Peri-Tethyan Rift/Wrench Basins and Passive MarginsMémoire. Mus.Nat. Hist. Nat, Paris, pp. 145–185.
Simancas, F., González-Lodeiro, A., Jhulin, C., Ayarza, A., Carbonell, R., Kashubin, A., Azor,D., Martínez-Poyatos, G.R., Almodovar, E., Pascual, R., Saenz, Expósito, I., 2003. Crustalstructure of the transpressional Variscan orogen of SW Iberia: SW Iberia deep seismicreflection profile (IBERSEIS). Tectonics 22 (6), 1962.
Talwani, M., Heirtzler, J.R., 1964. Computation of magnetic anomalies caused by twodimensional bodies of arbitrary shape. In: Parks, G.A. (Ed.), Computers in the mineralindustries, Part 1Geol. Sci. 9. Stanford Univ. Publ., Stanford, Calif., pp. 464–480.
Talwani, M., Worzel, J.L., Landisman, M., 1959. Rapid gravity computations for two-dimensional bodies with application to the Mendocino submarine fracture zone.J. Geophys. Res. 64 (1), 49–59. http://dx.doi.org/10.1029/JZ064i001p00049.
Torne, M., Pascal, G., Buhl, P., Watts, A.B., Mauffret, A., 1992. Crustal and velocity structureof the Valencia trough (western Mediterranean): part I. A combined refraction/wide-angle reflection and near-vertical reflection study. Tectonophysics 203, 1–20.
Torne, M., Fernandez, M., Vergés, J., Ayala, C., Salas, M.C., Jimenez-Munt, I., Buffett, G.G.,Díaz, 2015. Crustal and mantle lithospheric structure of the Iberian Peninsuladeduced from potential field modelling and thermal analysis. Tectonophysics 663,419–433.
Watts, A.B., Torné, M., Buhl, P., Mauffret, A., Pascal, G., Pinet, B., 1990. Evidence for reflec-tors in the lower continental crust before rifting in the Valencia trough. Nature 348,631–635.
Won, I.J., Bevis, M., 1987. Computing the gravitational and magnetic anomalies due to apolygon; algorithms and Fortran subroutines. Geophysics 52 (2), 232–238.
Zelt, C.A., Smith, R.B., 1992. Seismic traveltime inversion for 2-D crustal velocity structure.Geophys. J. Int. 108, 16–34. http://dx.doi.org/10.1111/j.1365-246x.1992.tb00836.x.
Zeyen, H.J., Banda, E., Gallart, J., Ansorge, J., 1985. A wide-angle seismic reconnaissance ofthe crust and upper mantle in the Celtiberian chain (Spain). Earth Planet. Sci. Lett. 75,393–402.
