- Email: [email protected]
Mantle transition zone beneath central-eastern Greenland: Possible evidence for a deep tectosphere from receiver functions
Tectonophysics 728–729 (2018) 34–40
Contents lists available at ScienceDirect
Tectonophysics journal homepage: www.elsevier.com/locate/tecto
Mantle transition zone beneath central-eastern Greenland: Possible evidence for a deep tectosphere from receiver functions
T
⁎
Helene Anja Krafta, Lev Vinnikb, , Hans Thyboc,d a
University of Copenhagen, Copenhagen, Denmark Institute of Physics of the Earth, Moscow, Russia c Eurasia Institute of Earth sciences, Istanbul Technical University, Istanbul, Turkey d Centre for Earth Evolution and Dynamics, University of Oslo, Norway b
A R T I C L E I N F O
A B S T R A C T
Keywords: Hotspot Upper mantle Mantle discontinuities Mantle transition zone P receiver functions Tectosphere
We investigate the mantle of central-eastern Greenland by using recordings with data from 24 local broad-band seismograph stations. We apply P wave receiver function technique and evaluate the difference in the arrival times of seismic phases that are formed by P to SV mode conversion at the 410-km and 660-km seismic discontinuities. These boundaries mark the top and bottom of the mantle transition zone (MTZ). The difference in the arrival time of the phases from the 410-km and 660-km discontinuities is sensitive to the thickness of the MTZ and relatively insensitive to volumetric velocity anomalies above the 410-km discontinuity. Near the east coast of Greenland in the region of the Skaergaard basalt intrusions we find two regions where the differential time is reduced by more than 2 s. The 410-km discontinuity in these regions is depressed by more than 20 km. The depression may be explained by a temperature elevation of ~150 °C. We hypothesize that the basaltic intrusions and the temperature anomalies at a depth of ~400 km are, at least partly, effects of the passage of Greenland over the Iceland hotspot at about 55 Ma. This explanation is consistent with the concept of tectosphere and implies that the upper mantle to a depth of ~400 km translates coherently with the Greenland plate.
1. Introduction Greenland has a geologic history of almost 4 Gyr. It is a Precambrian shield, with an Archean block in the south and Proterozoic mobile belts in the north. Since about 1.6 Ga Greenland was part of Laurentia and major geologic development took place mainly along its margins (Henriksen et al., 2009). In the late Ordovician (around 450 Ma) the closure of the Iapetus Ocean between Laurentia and Baltica led to the collision between the two continents and the Caledonian Orogeny. The opening of the Central Atlantic in the Cretaceous (around 130 Ma) reached southern Greenland at around 80 Ma. Initially, sea-floor spreading began at the west side of Greenland but at around 40 Ma it shifted to the eastern side (Torsvik et al., 2002). This shift is close in time to the main magmatic phase of the North Atlantic Igneous Province at ca. 60.5 and 54.5 Ma (Jolley and Bell, 2002). At about the same time, according to the plate reconstruction of Lawver and Müller (1994) east Greenland passed over the Iceland hotspot. A recent reconstruction (Torsvik et al., 2015) suggests that the Iceland hotspot was close to Greenland's east coast between ca. 70 and 40 Ma. This means that the Tertiary basaltic outcrops at the east coast (Fig. 1) with an age of around 55 Myr (Henriksen et al., 2009) may be at least partly related to ⁎
the Iceland hotspot. The crust and upper mantle beneath Greenland has been the focus of recent geophysical studies. The crustal structure of Greenland has been studied in a series of seismic experiments in the ice-free coastal regions. For the region under the ice sheet there are sparse estimates of crustal thickness ranging typically from 40 km to 50 km (Dahl-Jensen et al., 2003; Artemieva and Thybo, 2008, 2013). These estimates are obtained mostly by receiver function techniques at less than 10 locations and only one refraction seismic profile (Shulgin and Thybo, 2015). In the coastal regions the crust is relatively thin (20–30 km). The shear velocity structure of the upper mantle was investigated by Rayleigh wave tomography (Darbyshire et al., 2004). The results show high uppermantle velocities in central regions and somewhat lower velocities in the east. The seismological lithosphere has an average thickness of ~180 km. A thick mantle lithosphere beneath central Greenland is also shown by Lebedev et al., 2017. Estimates of geothermal flux (Rogozhina et al., 2016) are indicative of a long east-west oriented geothermal anomaly which may be related to the passage of Greenland over the Iceland hotspot. The hotspots may have affected not only the crust and the upper mantle but also the mantle transition zone (MTZ). However, the mantle at depth of 400–700 km is still rarely sampled by
Corresponding author. E-mail address: [email protected] (L. Vinnik).
https://doi.org/10.1016/j.tecto.2018.02.008 Received 29 September 2017; Received in revised form 10 February 2018; Accepted 12 February 2018 Available online 14 February 2018 0040-1951/ © 2018 Published by Elsevier B.V.
Tectonophysics 728–729 (2018) 34–40
H.A. Kraft et al.
Fig. 1. Map of the North Atlantic region (topography from Amante and Eakins (2012), left) and of the study area (right). The white area on the right shows the extent of the inland ice, surface geology is from Henriksen (2008). Seismograph stations are shown by triangles. Black triangles mark the stations of the temporary deployment from 2009 to 2012. Orange triangles are the stations from GLISN/GLATIS networks. The stations with blue labels were installed on ice. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)
Fig. 2. Epicenters of seismic events that were used in receiver function calculations. The concentric circles mark 30°, 60°, and 90° epicentral distance to the centre of the array (marked with a blue triangle). (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)
which were installed in central-eastern Greenland in the region between Scoresby Sund and Summit from June 2009 to May 2012. Eight of these stations were installed on bedrock, and the remaining 10 were installed on the ice. These stations were complemented by stations from other networks (Dahl-Jensen et al., 2003): DBG, ICESG, SCO, SOE, SUMG, HJO (Fig. 1).
available models (e.g., Rickers et al., 2013) and in the present study we address this issue. 2. Data and methods We primarily used data from 18 STS-2 broadband seismometers 35
Tectonophysics 728–729 (2018) 34–40
H.A. Kraft et al.
Fig. 3. Stacked receiver functions for sub-regions: A, B, C, D, E, F, G. Each trace corresponds to the conversion depth shown on the y-axis. P410s and P660s seismic phases are marked by arrows. IASP91 reference times are shown by dash lines.
variations in the crust and upper mantle above the 410-km boundary. We solve this problem by calculating the differential time between arrivals of the P660s and P410s phases. The ray paths of P660s and P410s phases in the crust and upper mantle are close to each other for the same seismic recording, and, as a result, the differential time is relatively insensitive to the properties of the Earth above the MTZ. To detect the P660s and P410s phases and to map the differential time, a large number of PRFs are stacked. One possibility is to apply a version of CCP (Common Conversion Point) stacking: to divide the Earth's surface into cells and to stack the PRFs, for which the surface projections of the conversion points fall into the same cell, with appropriate move-out time corrections. The surface projections of the conversion points for P410s and P660s phases for the same recording are at different distances (around 1° and 2°, respectively) from the seismograph station, and the set of PRFs thus selected for the detection of the P410s phase differs from that for the P660s phase. Therefore the differential time of P660s and P410s can be affected by lateral heterogeneity in the crust and upper mantle above the MTZ. In order to avoid this, we find the conversion points in the middle of the MTZ (at a depth of 535 km) and stack the PRFs which have the surface projections of the conversion points within the same cell. By this procedure the P410s and P660s phases for each cell are detected in the same set of
Our data base consists of recordings of 547 seismic events (Fig. 2) with magnitudes greater than 5.4 and epicentral distances between 35° and 90°. Most events are located to the north and south-west of the network. We used the P receiver function technique (PRF) of Vinnik (1977). To construct PRF we rotate the seismograms from the ZNE to the LQT coordinate system, where L corresponds to the principal direction of the P wave motion and Q is normal to L in the wave-propagation plane. In order to remove source effects we deconvolve the Q components from the respective L components in the time domain. In the PRF calculations we apply bandpass filters with corner periods near 30 s and 6 s. These periods provide the clearest signals from the 410 km(P410s) and 660 km- (P660s) discontinuities. Some of the stations were operating on up to 3290 m thick ice (at station 18), which causes a delay of up to ~1 s of the P-to-SV converted phases. To remove this effect we introduced time shifts in the receiver functions for the stations on the ice sheet. The corrections are calculated by assuming 3980 m/s and 1840 m/s for the P and S wave velocities in the ice (Wittlinger and Farra, 2012). The largest correction is −0.96 s for station 18. The main difficulty when interpreting the travel times of P410s and P660s phases is the separation of the effects of topography on the 410 km and 660 km discontinuities and of volumetric velocity 36
Tectonophysics 728–729 (2018) 34–40
H.A. Kraft et al.
Fig. 4. Bootstrap resampling results. Each trace represents one of 50 resamplings for each of the 7 regions (from A to G). Left: stacks with moveout for P410s. Right: stack with moveout correction for P660s. Dashed lines mark standard delay times according to IASP91.
within a fraction of a second: −0.2 s, 0.2 s and 0.5 s for B, C and F. These residuals are close to the estimated uncertainty in our measurements. Assuming the MTZ velocities of IASP91 (Kennett and Engdahl, 1991), a residual of −1 s indicates that the thickness of the MTZ is reduced by ~10 km. If the temperature in the MTZ is elevated by 100 °C, the related effect on the differential time due to lower velocities is around +0.2 s (Shen et al., 2002), which is small enough to be neglected. In addition to the thickness of the MTZ we evaluate less accurately the topography on the 410-km and 660-km discontinuities. The arrival times of the P410s phase vary within a range of 3.2 s from 43.5 s to 46.7 s. The largest values (46.2 and 46.7 s) are observed in the subregions with the lowest differential times (E and G). By comparison, the arrival times of the P660s phase of all sub-regions vary within a range of only 2.1 s (from 67.2 s to 69.3 s). If sub-region F is excluded, the range of variations of time of P660s is only 1.4 s. We note that in F the differential time is close to the standard time, but the times of P410s and P660s phases are larger than the standard times by 0.9 s for the P410s phase and 1.4 s for the P660s phase. This means that the late arrival of P660s in F is mainly an effect of low upper mantle velocities rather than of a depression on the 660-km discontinuity. These regularities suggest that the reduced differential time in E, G, A, and D is an effect of topography on the 410-km discontinuity. If the depth to the 660-km discontinuity in the study region is relatively stable, the observed lateral variations of time of the P660s phase reflect mainly volumetric velocity variations in the Earth's medium above the 410-km discontinuity. The residuals of this time relative to the IASP91 are +0.7 s, −0.7 s, +0.3 s, +0.1 s, −0.3 s, +1.4 s and +0.5 s for A, B, C, D, E, F, and G, respectively, with an average residual of +0.3 s. Assuming that these residuals present mainly the effect of anomalous velocities in the upper mantle, we infer that the lowest upper mantle velocities correspond to sub-regions A and F. The highest velocities correspond to B. All other residuals are in a range of a small fraction of
PRFs and the effect of lateral heterogeneity of the Earth above the 410 km discontinuity on the differential time is minimized. We experimented with several systems of cells (sub-regions). The critical parameters of a sub-region are its size and the number of stacked PRFs. These parameters are inter-related: the larger the subregion, the larger the number of the PRFs to be stacked. We have found that the optimum dimension of a sub-region for our observations is about 300 km and the minimum number of the stacked PRFs is 25. Based on these estimates, we divided the area into 7 sub-regions (A, B, C, D, E, F, G, Fig. 3). The number of individual conversion points and the stacked PRFs in each sub-region varies from 138 (in A) to 25 (in G). This number is sufficient for a robust detection of the P410s and P660s phases (Fig. 3). To confirm this, we conducted a bootstrap resampling analysis (Efron and Tibshirani, 1991) with 50 resamplings for P410s and P660s for each sub-region (Fig. 4). These tests demonstrate that the standard error (confidence interval of 66%) of the delay time of the P410s phase in each sub-region is around 0.1 s. The accuracy of the arrival time for the P660s phase is 0.2 s in D and G, and 0.1 s for the other sub-regions. Assuming that the errors of time measurement of P410s and P660s are independent, the differential time between P660s and P410s phases is measured typically with a standard error (confidence interval of 66%) of less than 0.2 s. We summarize the results of our measurements in Fig. 5. For each sub-region we show the time of the P410s phase, of the P660s phase, the differential time between the two and the residuals of these times relative to the IASP91 reference Earth model (Kennett and Engdahl, 1991) for a P wave slowness of 6.4 s/deg. The IASP91 predicts these times to be 44.0 s, 67.9 s and 23.9 s, respectively. The differential time of 23.9 s provides a good approximation for a global set of seismic data (e.g., Chevrot et al., 1999). We find that in 4 sub-regions of the 7 the differential time is significantly lower than the standard differential time: −2.5 s, −2.2 s, −1.0 s and −1.0 s for E, G, A and D, respectively. For the other three sub-regions the residuals of the differential time are 37
Tectonophysics 728–729 (2018) 34–40
H.A. Kraft et al.
Fig. 5. A: sub-regions A, B, C, D, E, F, G and projections of conversion points at a depth of 535 km. Seismograph stations are shown by triangles. The numbers are the differential times in seconds and the residuals of the differential time relative to the IASP91 (in brackets). B: the times of the P410s phase in seconds and the related residuals relative to the IASP91 (in brackets). C: the same as B but for the P660s phase.
central and eastern Greenland (sub-regions A, C, D and the neighborhood of them) suggest that the high-velocity root is practically missing there (might have been destroyed by recent thermal and/or metasomatic processes). This may be reconciled with the trend in the Rayleigh wave model by Darbyshire et al. (2004). Sub-region B with the large negative residual (−0.7 s) should have a pronounced high-velocity root, in agreement with this model. An intriguing detail of the stacks in Fig. 3 is a phase at a time around 50 s in sub-regions A, B, C, D and G, which follows the P410s phase with comparable amplitudes and opposite polarity. Previously a similar feature was observed for the Azores and Cape Verde hotspots (Vinnik et al., 2012) and described as an “M-shaped” phase. It was interpreted as evidence of a low-velocity layer at depths of 450 to 510 km. A lowvelocity layer in the upper MTZ has also been found in S wave receiver functions for a few other hotspots, including Iceland (Vinnik et al.,
a second and comparable with the uncertainty of the estimates. Note that our method provides accurate locations only for the anomalous structure of the MTZ. The residuals corresponding to the crust and uppermost mantle are accumulated in the vicinity of the seismograph stations and some of the related piercing points may be located outside the corresponding sub-regions in their immediate neighborhood.
3. Discussion and conclusions We observe mostly positive time residuals of the P660s phase. On a global scale the residuals of the P410 and the P660s phases relative to the IASP91 model range from −1.8 s to +5 s (Chevrot et al., 1999). Most of the data in the interval from −1.8 s to −1.0 s are obtained for Precambrian shields. The negative residuals for the shields are attributed to high-velocity mantle roots. Positive residuals for the shield of 38
Tectonophysics 728–729 (2018) 34–40
H.A. Kraft et al.
et al., 2017). In fact, coupling between the lithosphere and the underlying mantle is necessary if, as is often accepted, the plates are driven by mantle flow.
2005) and Afar (Vinnik et al., 2004). The so-called 520-km seismic discontinuity, which is often interpreted as the phase transition from wadsleyite to ringwoodite, may in hotspots represent the base of this low-velocity layer. Keshav et al. (2011) reported an abrupt drop in the solidus temperature of carbonated mantle in the pressure range corresponding to these depths. The reduced S wave velocity in this depth range may be an effect of an elevated temperature and the reduced solidus temperature. The effect of this low-velocity layer in the differential time between the P660s and P410s phases is on the order of 0.1 s and can be neglected. Observations of a highly heterogeneous zone to depths of 520 km have also been made from controlled-source P wave data from Siberia and North America and explained by either high Mg# or the transformation of pyroxene into a garnet phase (Thybo et al., 2003). The most important result of our study is the evidence of a thin MTZ in sub-regions A, D and especially in E and G. Sub-region G corresponds to an area with basaltic outcrops with ages around 55 Ma (Fig. 1). At the time of the eruptions this region may have passed over the Iceland hotspot. Sub-region E corresponds to less extensive basaltic outcrops (Fig. 1) and of about the same age. These outcrops may be of the same origin as in G or they may be related to the Jan Mayen hotspot. The deep structure of the Iceland hotspot was investigated with P receiver functions (Du et al., 2006). Iceland was divided into central, western and eastern regions. The Iceland hotspot is presently located in the central region, where the residual of the differential time between the P660s and P410s phases is −1.4 s and the thickness of the MTZ is reduced by about 15 km relative to the IASP91. This is the result of a depression of the 410-km discontinuity. Similar observations exist for many other hotspots (Courtier et al., 2007). The reduced thickness of the MTZ is usually, but not always (e.g., Vinnik et al., 2012), the effect of a depression on the 410-km discontinuity and no topography on the 660-km discontinuity. In agreement with the Clapeyron slope of 3.0 MPa/K (Bina and Helffrich, 1994) the depression of 15–20 km on the 410 km discontinuity is likely the effect of a temperature elevation of ~150 K. The stable depth of the 660-km discontinuity is the result of a zero temperature anomaly at that depth or an effect of two phase transitions with opposite Clapeyron slopes (Hirose, 2002). In the western region of Iceland the residual of the differential time is +0.2 s, which is close to the standard time. This means that the reduced width of the MTZ in eastern Greenland cannot be a simple westward extension of the anomaly related to the present-day Iceland hotspot. The presented interpretation of the receiver functions for Iceland is based on the assumption that the S-wave velocity in the MTZ of the central region is reduced relative to the neighboring regions by 3% which is consistent with results of seismic tomography. Without this reduction the 660-km discontinuity in the central region should be depressed (Du et al., 2006). The model of a depressed 660-km discontinuity in Iceland is advocated by Jenkins et al. (2016) with implication that the 660-km discontinuity in Iceland is formed by a phase transition in garnet with a positive Clapeyron slope. However, our preferred interpretation of these data is that of Du et al. (2006). The basaltic outcrops of eastern Greenland may have formed due to the passage of Greenland over the Iceland hotspot in about the same way as the present day volcanic rocks in Iceland. The depressions on the 410-km discontinuity beneath eastern Greenland are similar to the depression beneath present-day Iceland, and their location beneath the 55 Ma old basalts suggests that the mantle at a depth of ~400 km might translate coherently with the lithosphere of Greenland. Topography on the 410-km boundary may be preserved for some time, in spite of a break in supply of hot material from the lower mantle. This contradicts the idea of a decoupling between the continental lithosphere and asthenosphere at depth of ~100–200 km, but is consistent with the concept of a tectosphere – the continental mantle layer which may extend to depths of at least 400 km and translate coherently during plate motions (Jordan, 1975). This concept finds some support in other seismic data (e.g., Rocha et al., 2011; Jordan and Paulson, 2013; Vinnik
Acknowledgements This work was made possible through the support and funding of several field campaigns by the Forskningsråd for Natur og Univers (09065945, 09-072802, 0602-02060B). The data from the permanent stations were collected and distributed by the Greenland Ice Sheet Monitoring Network (GLISN) federation and its members. The work of Lev Vinnik was partially supported by the RFBR grant 15-05-04938. Helene Kraft carried out her PhD study financed from grants to Hans Thybo from Independent Research Fund Denmark Nos 11-104254 and 16/059776-15. The authors appreciate help from Sergey Oreshin and comments from M. Agius and an anonymous reviewer. References Amante, C., Eakins, B., 2012. ETOPO1, Global 1 Arc-minute Ocean Depth and Land Elevation From the US National Geophysical Data Center (NGDC). Artemieva, I.M., Thybo, H., 2008. Deep Norden: highlights of the lithospheric structure of Northern Europe, Iceland and Greenland. Episodes 31, 98–106. Artemieva, I.M., Thybo, H., 2013. EUNAseis: a seismic model for Moho and crustal structure in Europe, Greenland, and the North Atlantic region. Tectonophysics 609, 97–153. http://dx.doi.org/10.1016/j.tecto.2013.1008.1004. Bina, C.R., Helffrich, G., 1994. Phase transition Clapeyron slopes and transition zone seismic discontinuity topography. J. Geophys. Res. 99 (B8), 15853–15860. Chevrot, S., Vinnik, L., Montagner, J.-P., 1999. Global-scale analysis of the mantle Pds phases. J. Geophys. Res. Solid Earth 104 (B9), 20203–20219. Courtier, A.M., Jackson, M.G., Lawrence, J.F., Wang, Z., Lee, C.T.A., Halama, R., Warren, J.M., Workman, R., Xu, W., Hirschmann, M.M., Larson, A.M., 2007. Correlation of seismic and petrologic thermometers suggests deep thermal anomalies beneath hotspots. Earth Planet. Sci. Lett. 264 (1), 308–316. Dahl-Jensen, T., Larsen, T.B., Woelbern, I., Bach, T., Hanka, W., Kind, R., Gregersen, S., Mosegaard, K., Voss, P., Gudmundsson, O., 2003. Depth to Moho in Greenland: receiver-function analysis suggests two Proterozoic blocks in Greenland. Earth Planet. Sci. Lett. 205 (3–4), 379–393. Darbyshire, F.A., Larsen, T.B., Mosegaard, K., Dahl-Jensen, T., Gudmundsson, O., Bach, T., Gregersen, S., Pedersen, H.A., 2004. A first detailed look at the Greenland lithosphere and upper mantle, using Rayleigh wave tomography. Geophys. J. Int. 158 (1), 267–286. Du, Z., Vinnik, L.P., Foulger, G.R., 2006. Evidence from P-to-S mantle converted waves for a flat “660-km” discontinuity beneath Iceland. Earth Planet. Sci. Lett. 241 (1–2), 271–280. Efron, B., Tibshirani, R., 1991. Statistical data analysis in the computer age. Science 253, 390–395. Henriksen, N. (Ed.), 2008. Geological History of Greenland – Four Billion Years of Earth Evolution. Geological Survey of Denmark and Greenland. Henriksen, N., Higgins, A., Kalsbeek, F., Pulvertaft, T.C.R., 2009. Greenland From Archaean to Quaternary: Descriptive Text to the 1995 Geological Map of Greenland, 1: 2 500 000. Hirose, K., 2002. Phase transitions in pyrolitic mantle around 670-km depth: implications for upwelling of plumes from the lower mantle. J. Geophys. Res. Solid Earth 107 (B4). http://dx.doi.org/10.1029/2001JB000597. Jenkins, J., Cottaar, S., White, R.S., Deuss, A., 2016. Depressed mantle discontinuities beneath Iceland: evidence of a garnet controlled 660 km discontinuity? Earth Planet. Sci. Lett. 433, 159–168. Jolley, D.W., Bell, B.R., 2002. The evolution of the North Atlantic Igneous Province and the opening of the NE Atlantic rift. Geol. Soc. Lond. Spec. Publ. 197 (1), 1–13. Jordan, T.H., 1975. The continental tectosphere. Rev. Geophys. 13 (3), 1–12. Jordan, T.H., Paulson, E.M., 2013. Convergence depths of tectonic regions from an ensemble of global tomographic models. J. Geophys. Res. Solid Earth 118 (8), 4196–4225. Kennett, B.L.N., Engdahl, E.R., 1991. Traveltimes for global earthquake location and phase identification. Geophys. J. Int. 105 (2), 429–465. Keshav, S., Gudfinnsson, G.H., Presnall, D.C., 2011. Melting phase relations of simplified carbonated peridotite at 12–26 GPa in the systems CaO–MgO–SiO2–CO2 and CaO–MgO–Al2O3–SiO2–CO2: highly calcic magmas in the transition zone of the Earth. J. Petrol. 52 (11), 2265–2291. Lawver, L.A., Müller, R.D., 1994. Iceland hotspot track. Geology 22, 311–314. Lebedev, S., Schaeffer, A.J., Fullea, J., Pease, V., 2017. Seismic tomography of the Arctic region: inferences for the thermal structure and evolution of the lithosphere. Geol. Soc. Lond., Spec. Publ. 460, SP460–10. Rickers, F., Fichtner, A., Trampert, J., 2013. The Iceland–Jan Mayen plume system and its impact on mantle dynamics in the North Atlantic region: evidence from full-waveform inversion. Earth Planet. Sci. Lett. 367, 39–51. Rocha, M.P., Schimmel, M., Assumpção, M., 2011. Upper-mantle seismic structure beneath SE and Central Brazil from P-and S-wave regional traveltime tomography. Geophys. J. Int. 184 (1), 268–286.
39
Tectonophysics 728–729 (2018) 34–40
H.A. Kraft et al.
Steinberger, B., Corfu, F., Ashwal, L.D., Griffin, W.L., et al., 2015. Continental crust beneath southeast Iceland. Proc. Natl. Acad. Sci. 112 (15), E1818–E1827. Vinnik, L.P., 1977. Detection of waves converted from P to SV in the mantle. Phys. Earth Planet. Inter. 15 (1), 39–45. Vinnik, L.P., Farra, V., Kind, R., 2004. Deep structure of the Afro-Arabian hotspot by S receiver functions. Geophys. Res. Lett. 31, L11608. http://dx.doi.org/10.1029/ 2004GL019574. Vinnik, L.P., Foulger, G.R., Du, Z., 2005. Seismic boundaries in the mantle beneath Iceland: a new constraint on temperature. Geophys. J. Int. 160 (2), 533–538. Vinnik, L., Silveira, G., Kiselev, S., Farra, V., Weber, M., Stutzmann, E., 2012. Cape Verde hotspot from the upper crust to the top of the lower mantle. Earth Planet. Sci. Lett. 319, 259–268. Vinnik, L.P., Oreshin, S.I., Makeyeva, L.I., 2017. Siberian traps: hypotheses and seismology data. Izv. Phys. Solid Earth 53 (3), 332–340. Wittlinger, G., Farra, V., 2012. Observation of low shear wave velocity at the base of the polar ice sheets: evidence for enhanced anisotropy. Geophys. J. Int. 190, 391–405.
Rogozhina, I., Petrunin, A.G., Vaughan, A.P., Steinberger, B., Johnson, J.V., Kaban, M.K., Calov, R., Rickers, F., Thomas, M., Koulakov, I., 2016. Melting at the base of the Greenland ice sheet explained by Iceland hotspot history. Nat. Geosci. 9 (5), 366–369. Shen, Y., Solomon, S.C., Bjarnason, I.T., Nolet, G., Morgan, W.J., Allen, R.M., Vogfjörd, K., Jakobsdóttir, S., Stefánsson, R., Julian, B.R., Foulger, G.R., 2002. Seismic evidence for a tilted mantle plume and north–south mantle flow beneath Iceland. Earth Planet. Sci. Lett. 197 (3), 261–272. Shulgin, A., Thybo, H., 2015. Seismic explosion sources on an ice cap–technical considerations. Polar Sci. 9 (1), 107–118. Thybo, H., Nielsen, L., Perchuc, E., 2003. Seismic scattering at the top of the mantle Transition Zone. Earth Planet. Sci. Lett. 216, 259–269. Torsvik, T.H., Carlos, D., Mosar, J., Cocks, L.R.M., Malme, T.N., 2002. Global reconstructions and North Atlantic paleogeography 440 Ma to recent. In: BATLAS—Mid Norway Plate Reconstruction Atlas With Global and Atlantic Perspectives, pp. 18–39. Torsvik, T.H., Amundsen, H.E., Tronnes, R.G., Doubrovine, P.V., Gaina, C., Kusznir, N.J.,
40
Contents lists available at ScienceDirect
Tectonophysics journal homepage: www.elsevier.com/locate/tecto
Mantle transition zone beneath central-eastern Greenland: Possible evidence for a deep tectosphere from receiver functions
T
⁎
Helene Anja Krafta, Lev Vinnikb, , Hans Thyboc,d a
University of Copenhagen, Copenhagen, Denmark Institute of Physics of the Earth, Moscow, Russia c Eurasia Institute of Earth sciences, Istanbul Technical University, Istanbul, Turkey d Centre for Earth Evolution and Dynamics, University of Oslo, Norway b
A R T I C L E I N F O
A B S T R A C T
Keywords: Hotspot Upper mantle Mantle discontinuities Mantle transition zone P receiver functions Tectosphere
We investigate the mantle of central-eastern Greenland by using recordings with data from 24 local broad-band seismograph stations. We apply P wave receiver function technique and evaluate the difference in the arrival times of seismic phases that are formed by P to SV mode conversion at the 410-km and 660-km seismic discontinuities. These boundaries mark the top and bottom of the mantle transition zone (MTZ). The difference in the arrival time of the phases from the 410-km and 660-km discontinuities is sensitive to the thickness of the MTZ and relatively insensitive to volumetric velocity anomalies above the 410-km discontinuity. Near the east coast of Greenland in the region of the Skaergaard basalt intrusions we find two regions where the differential time is reduced by more than 2 s. The 410-km discontinuity in these regions is depressed by more than 20 km. The depression may be explained by a temperature elevation of ~150 °C. We hypothesize that the basaltic intrusions and the temperature anomalies at a depth of ~400 km are, at least partly, effects of the passage of Greenland over the Iceland hotspot at about 55 Ma. This explanation is consistent with the concept of tectosphere and implies that the upper mantle to a depth of ~400 km translates coherently with the Greenland plate.
1. Introduction Greenland has a geologic history of almost 4 Gyr. It is a Precambrian shield, with an Archean block in the south and Proterozoic mobile belts in the north. Since about 1.6 Ga Greenland was part of Laurentia and major geologic development took place mainly along its margins (Henriksen et al., 2009). In the late Ordovician (around 450 Ma) the closure of the Iapetus Ocean between Laurentia and Baltica led to the collision between the two continents and the Caledonian Orogeny. The opening of the Central Atlantic in the Cretaceous (around 130 Ma) reached southern Greenland at around 80 Ma. Initially, sea-floor spreading began at the west side of Greenland but at around 40 Ma it shifted to the eastern side (Torsvik et al., 2002). This shift is close in time to the main magmatic phase of the North Atlantic Igneous Province at ca. 60.5 and 54.5 Ma (Jolley and Bell, 2002). At about the same time, according to the plate reconstruction of Lawver and Müller (1994) east Greenland passed over the Iceland hotspot. A recent reconstruction (Torsvik et al., 2015) suggests that the Iceland hotspot was close to Greenland's east coast between ca. 70 and 40 Ma. This means that the Tertiary basaltic outcrops at the east coast (Fig. 1) with an age of around 55 Myr (Henriksen et al., 2009) may be at least partly related to ⁎
the Iceland hotspot. The crust and upper mantle beneath Greenland has been the focus of recent geophysical studies. The crustal structure of Greenland has been studied in a series of seismic experiments in the ice-free coastal regions. For the region under the ice sheet there are sparse estimates of crustal thickness ranging typically from 40 km to 50 km (Dahl-Jensen et al., 2003; Artemieva and Thybo, 2008, 2013). These estimates are obtained mostly by receiver function techniques at less than 10 locations and only one refraction seismic profile (Shulgin and Thybo, 2015). In the coastal regions the crust is relatively thin (20–30 km). The shear velocity structure of the upper mantle was investigated by Rayleigh wave tomography (Darbyshire et al., 2004). The results show high uppermantle velocities in central regions and somewhat lower velocities in the east. The seismological lithosphere has an average thickness of ~180 km. A thick mantle lithosphere beneath central Greenland is also shown by Lebedev et al., 2017. Estimates of geothermal flux (Rogozhina et al., 2016) are indicative of a long east-west oriented geothermal anomaly which may be related to the passage of Greenland over the Iceland hotspot. The hotspots may have affected not only the crust and the upper mantle but also the mantle transition zone (MTZ). However, the mantle at depth of 400–700 km is still rarely sampled by
Corresponding author. E-mail address: [email protected] (L. Vinnik).
https://doi.org/10.1016/j.tecto.2018.02.008 Received 29 September 2017; Received in revised form 10 February 2018; Accepted 12 February 2018 Available online 14 February 2018 0040-1951/ © 2018 Published by Elsevier B.V.
Tectonophysics 728–729 (2018) 34–40
H.A. Kraft et al.
Fig. 1. Map of the North Atlantic region (topography from Amante and Eakins (2012), left) and of the study area (right). The white area on the right shows the extent of the inland ice, surface geology is from Henriksen (2008). Seismograph stations are shown by triangles. Black triangles mark the stations of the temporary deployment from 2009 to 2012. Orange triangles are the stations from GLISN/GLATIS networks. The stations with blue labels were installed on ice. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)
Fig. 2. Epicenters of seismic events that were used in receiver function calculations. The concentric circles mark 30°, 60°, and 90° epicentral distance to the centre of the array (marked with a blue triangle). (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)
which were installed in central-eastern Greenland in the region between Scoresby Sund and Summit from June 2009 to May 2012. Eight of these stations were installed on bedrock, and the remaining 10 were installed on the ice. These stations were complemented by stations from other networks (Dahl-Jensen et al., 2003): DBG, ICESG, SCO, SOE, SUMG, HJO (Fig. 1).
available models (e.g., Rickers et al., 2013) and in the present study we address this issue. 2. Data and methods We primarily used data from 18 STS-2 broadband seismometers 35
Tectonophysics 728–729 (2018) 34–40
H.A. Kraft et al.
Fig. 3. Stacked receiver functions for sub-regions: A, B, C, D, E, F, G. Each trace corresponds to the conversion depth shown on the y-axis. P410s and P660s seismic phases are marked by arrows. IASP91 reference times are shown by dash lines.
variations in the crust and upper mantle above the 410-km boundary. We solve this problem by calculating the differential time between arrivals of the P660s and P410s phases. The ray paths of P660s and P410s phases in the crust and upper mantle are close to each other for the same seismic recording, and, as a result, the differential time is relatively insensitive to the properties of the Earth above the MTZ. To detect the P660s and P410s phases and to map the differential time, a large number of PRFs are stacked. One possibility is to apply a version of CCP (Common Conversion Point) stacking: to divide the Earth's surface into cells and to stack the PRFs, for which the surface projections of the conversion points fall into the same cell, with appropriate move-out time corrections. The surface projections of the conversion points for P410s and P660s phases for the same recording are at different distances (around 1° and 2°, respectively) from the seismograph station, and the set of PRFs thus selected for the detection of the P410s phase differs from that for the P660s phase. Therefore the differential time of P660s and P410s can be affected by lateral heterogeneity in the crust and upper mantle above the MTZ. In order to avoid this, we find the conversion points in the middle of the MTZ (at a depth of 535 km) and stack the PRFs which have the surface projections of the conversion points within the same cell. By this procedure the P410s and P660s phases for each cell are detected in the same set of
Our data base consists of recordings of 547 seismic events (Fig. 2) with magnitudes greater than 5.4 and epicentral distances between 35° and 90°. Most events are located to the north and south-west of the network. We used the P receiver function technique (PRF) of Vinnik (1977). To construct PRF we rotate the seismograms from the ZNE to the LQT coordinate system, where L corresponds to the principal direction of the P wave motion and Q is normal to L in the wave-propagation plane. In order to remove source effects we deconvolve the Q components from the respective L components in the time domain. In the PRF calculations we apply bandpass filters with corner periods near 30 s and 6 s. These periods provide the clearest signals from the 410 km(P410s) and 660 km- (P660s) discontinuities. Some of the stations were operating on up to 3290 m thick ice (at station 18), which causes a delay of up to ~1 s of the P-to-SV converted phases. To remove this effect we introduced time shifts in the receiver functions for the stations on the ice sheet. The corrections are calculated by assuming 3980 m/s and 1840 m/s for the P and S wave velocities in the ice (Wittlinger and Farra, 2012). The largest correction is −0.96 s for station 18. The main difficulty when interpreting the travel times of P410s and P660s phases is the separation of the effects of topography on the 410 km and 660 km discontinuities and of volumetric velocity 36
Tectonophysics 728–729 (2018) 34–40
H.A. Kraft et al.
Fig. 4. Bootstrap resampling results. Each trace represents one of 50 resamplings for each of the 7 regions (from A to G). Left: stacks with moveout for P410s. Right: stack with moveout correction for P660s. Dashed lines mark standard delay times according to IASP91.
within a fraction of a second: −0.2 s, 0.2 s and 0.5 s for B, C and F. These residuals are close to the estimated uncertainty in our measurements. Assuming the MTZ velocities of IASP91 (Kennett and Engdahl, 1991), a residual of −1 s indicates that the thickness of the MTZ is reduced by ~10 km. If the temperature in the MTZ is elevated by 100 °C, the related effect on the differential time due to lower velocities is around +0.2 s (Shen et al., 2002), which is small enough to be neglected. In addition to the thickness of the MTZ we evaluate less accurately the topography on the 410-km and 660-km discontinuities. The arrival times of the P410s phase vary within a range of 3.2 s from 43.5 s to 46.7 s. The largest values (46.2 and 46.7 s) are observed in the subregions with the lowest differential times (E and G). By comparison, the arrival times of the P660s phase of all sub-regions vary within a range of only 2.1 s (from 67.2 s to 69.3 s). If sub-region F is excluded, the range of variations of time of P660s is only 1.4 s. We note that in F the differential time is close to the standard time, but the times of P410s and P660s phases are larger than the standard times by 0.9 s for the P410s phase and 1.4 s for the P660s phase. This means that the late arrival of P660s in F is mainly an effect of low upper mantle velocities rather than of a depression on the 660-km discontinuity. These regularities suggest that the reduced differential time in E, G, A, and D is an effect of topography on the 410-km discontinuity. If the depth to the 660-km discontinuity in the study region is relatively stable, the observed lateral variations of time of the P660s phase reflect mainly volumetric velocity variations in the Earth's medium above the 410-km discontinuity. The residuals of this time relative to the IASP91 are +0.7 s, −0.7 s, +0.3 s, +0.1 s, −0.3 s, +1.4 s and +0.5 s for A, B, C, D, E, F, and G, respectively, with an average residual of +0.3 s. Assuming that these residuals present mainly the effect of anomalous velocities in the upper mantle, we infer that the lowest upper mantle velocities correspond to sub-regions A and F. The highest velocities correspond to B. All other residuals are in a range of a small fraction of
PRFs and the effect of lateral heterogeneity of the Earth above the 410 km discontinuity on the differential time is minimized. We experimented with several systems of cells (sub-regions). The critical parameters of a sub-region are its size and the number of stacked PRFs. These parameters are inter-related: the larger the subregion, the larger the number of the PRFs to be stacked. We have found that the optimum dimension of a sub-region for our observations is about 300 km and the minimum number of the stacked PRFs is 25. Based on these estimates, we divided the area into 7 sub-regions (A, B, C, D, E, F, G, Fig. 3). The number of individual conversion points and the stacked PRFs in each sub-region varies from 138 (in A) to 25 (in G). This number is sufficient for a robust detection of the P410s and P660s phases (Fig. 3). To confirm this, we conducted a bootstrap resampling analysis (Efron and Tibshirani, 1991) with 50 resamplings for P410s and P660s for each sub-region (Fig. 4). These tests demonstrate that the standard error (confidence interval of 66%) of the delay time of the P410s phase in each sub-region is around 0.1 s. The accuracy of the arrival time for the P660s phase is 0.2 s in D and G, and 0.1 s for the other sub-regions. Assuming that the errors of time measurement of P410s and P660s are independent, the differential time between P660s and P410s phases is measured typically with a standard error (confidence interval of 66%) of less than 0.2 s. We summarize the results of our measurements in Fig. 5. For each sub-region we show the time of the P410s phase, of the P660s phase, the differential time between the two and the residuals of these times relative to the IASP91 reference Earth model (Kennett and Engdahl, 1991) for a P wave slowness of 6.4 s/deg. The IASP91 predicts these times to be 44.0 s, 67.9 s and 23.9 s, respectively. The differential time of 23.9 s provides a good approximation for a global set of seismic data (e.g., Chevrot et al., 1999). We find that in 4 sub-regions of the 7 the differential time is significantly lower than the standard differential time: −2.5 s, −2.2 s, −1.0 s and −1.0 s for E, G, A and D, respectively. For the other three sub-regions the residuals of the differential time are 37
Tectonophysics 728–729 (2018) 34–40
H.A. Kraft et al.
Fig. 5. A: sub-regions A, B, C, D, E, F, G and projections of conversion points at a depth of 535 km. Seismograph stations are shown by triangles. The numbers are the differential times in seconds and the residuals of the differential time relative to the IASP91 (in brackets). B: the times of the P410s phase in seconds and the related residuals relative to the IASP91 (in brackets). C: the same as B but for the P660s phase.
central and eastern Greenland (sub-regions A, C, D and the neighborhood of them) suggest that the high-velocity root is practically missing there (might have been destroyed by recent thermal and/or metasomatic processes). This may be reconciled with the trend in the Rayleigh wave model by Darbyshire et al. (2004). Sub-region B with the large negative residual (−0.7 s) should have a pronounced high-velocity root, in agreement with this model. An intriguing detail of the stacks in Fig. 3 is a phase at a time around 50 s in sub-regions A, B, C, D and G, which follows the P410s phase with comparable amplitudes and opposite polarity. Previously a similar feature was observed for the Azores and Cape Verde hotspots (Vinnik et al., 2012) and described as an “M-shaped” phase. It was interpreted as evidence of a low-velocity layer at depths of 450 to 510 km. A lowvelocity layer in the upper MTZ has also been found in S wave receiver functions for a few other hotspots, including Iceland (Vinnik et al.,
a second and comparable with the uncertainty of the estimates. Note that our method provides accurate locations only for the anomalous structure of the MTZ. The residuals corresponding to the crust and uppermost mantle are accumulated in the vicinity of the seismograph stations and some of the related piercing points may be located outside the corresponding sub-regions in their immediate neighborhood.
3. Discussion and conclusions We observe mostly positive time residuals of the P660s phase. On a global scale the residuals of the P410 and the P660s phases relative to the IASP91 model range from −1.8 s to +5 s (Chevrot et al., 1999). Most of the data in the interval from −1.8 s to −1.0 s are obtained for Precambrian shields. The negative residuals for the shields are attributed to high-velocity mantle roots. Positive residuals for the shield of 38
Tectonophysics 728–729 (2018) 34–40
H.A. Kraft et al.
et al., 2017). In fact, coupling between the lithosphere and the underlying mantle is necessary if, as is often accepted, the plates are driven by mantle flow.
2005) and Afar (Vinnik et al., 2004). The so-called 520-km seismic discontinuity, which is often interpreted as the phase transition from wadsleyite to ringwoodite, may in hotspots represent the base of this low-velocity layer. Keshav et al. (2011) reported an abrupt drop in the solidus temperature of carbonated mantle in the pressure range corresponding to these depths. The reduced S wave velocity in this depth range may be an effect of an elevated temperature and the reduced solidus temperature. The effect of this low-velocity layer in the differential time between the P660s and P410s phases is on the order of 0.1 s and can be neglected. Observations of a highly heterogeneous zone to depths of 520 km have also been made from controlled-source P wave data from Siberia and North America and explained by either high Mg# or the transformation of pyroxene into a garnet phase (Thybo et al., 2003). The most important result of our study is the evidence of a thin MTZ in sub-regions A, D and especially in E and G. Sub-region G corresponds to an area with basaltic outcrops with ages around 55 Ma (Fig. 1). At the time of the eruptions this region may have passed over the Iceland hotspot. Sub-region E corresponds to less extensive basaltic outcrops (Fig. 1) and of about the same age. These outcrops may be of the same origin as in G or they may be related to the Jan Mayen hotspot. The deep structure of the Iceland hotspot was investigated with P receiver functions (Du et al., 2006). Iceland was divided into central, western and eastern regions. The Iceland hotspot is presently located in the central region, where the residual of the differential time between the P660s and P410s phases is −1.4 s and the thickness of the MTZ is reduced by about 15 km relative to the IASP91. This is the result of a depression of the 410-km discontinuity. Similar observations exist for many other hotspots (Courtier et al., 2007). The reduced thickness of the MTZ is usually, but not always (e.g., Vinnik et al., 2012), the effect of a depression on the 410-km discontinuity and no topography on the 660-km discontinuity. In agreement with the Clapeyron slope of 3.0 MPa/K (Bina and Helffrich, 1994) the depression of 15–20 km on the 410 km discontinuity is likely the effect of a temperature elevation of ~150 K. The stable depth of the 660-km discontinuity is the result of a zero temperature anomaly at that depth or an effect of two phase transitions with opposite Clapeyron slopes (Hirose, 2002). In the western region of Iceland the residual of the differential time is +0.2 s, which is close to the standard time. This means that the reduced width of the MTZ in eastern Greenland cannot be a simple westward extension of the anomaly related to the present-day Iceland hotspot. The presented interpretation of the receiver functions for Iceland is based on the assumption that the S-wave velocity in the MTZ of the central region is reduced relative to the neighboring regions by 3% which is consistent with results of seismic tomography. Without this reduction the 660-km discontinuity in the central region should be depressed (Du et al., 2006). The model of a depressed 660-km discontinuity in Iceland is advocated by Jenkins et al. (2016) with implication that the 660-km discontinuity in Iceland is formed by a phase transition in garnet with a positive Clapeyron slope. However, our preferred interpretation of these data is that of Du et al. (2006). The basaltic outcrops of eastern Greenland may have formed due to the passage of Greenland over the Iceland hotspot in about the same way as the present day volcanic rocks in Iceland. The depressions on the 410-km discontinuity beneath eastern Greenland are similar to the depression beneath present-day Iceland, and their location beneath the 55 Ma old basalts suggests that the mantle at a depth of ~400 km might translate coherently with the lithosphere of Greenland. Topography on the 410-km boundary may be preserved for some time, in spite of a break in supply of hot material from the lower mantle. This contradicts the idea of a decoupling between the continental lithosphere and asthenosphere at depth of ~100–200 km, but is consistent with the concept of a tectosphere – the continental mantle layer which may extend to depths of at least 400 km and translate coherently during plate motions (Jordan, 1975). This concept finds some support in other seismic data (e.g., Rocha et al., 2011; Jordan and Paulson, 2013; Vinnik
Acknowledgements This work was made possible through the support and funding of several field campaigns by the Forskningsråd for Natur og Univers (09065945, 09-072802, 0602-02060B). The data from the permanent stations were collected and distributed by the Greenland Ice Sheet Monitoring Network (GLISN) federation and its members. The work of Lev Vinnik was partially supported by the RFBR grant 15-05-04938. Helene Kraft carried out her PhD study financed from grants to Hans Thybo from Independent Research Fund Denmark Nos 11-104254 and 16/059776-15. The authors appreciate help from Sergey Oreshin and comments from M. Agius and an anonymous reviewer. References Amante, C., Eakins, B., 2012. ETOPO1, Global 1 Arc-minute Ocean Depth and Land Elevation From the US National Geophysical Data Center (NGDC). Artemieva, I.M., Thybo, H., 2008. Deep Norden: highlights of the lithospheric structure of Northern Europe, Iceland and Greenland. Episodes 31, 98–106. Artemieva, I.M., Thybo, H., 2013. EUNAseis: a seismic model for Moho and crustal structure in Europe, Greenland, and the North Atlantic region. Tectonophysics 609, 97–153. http://dx.doi.org/10.1016/j.tecto.2013.1008.1004. Bina, C.R., Helffrich, G., 1994. Phase transition Clapeyron slopes and transition zone seismic discontinuity topography. J. Geophys. Res. 99 (B8), 15853–15860. Chevrot, S., Vinnik, L., Montagner, J.-P., 1999. Global-scale analysis of the mantle Pds phases. J. Geophys. Res. Solid Earth 104 (B9), 20203–20219. Courtier, A.M., Jackson, M.G., Lawrence, J.F., Wang, Z., Lee, C.T.A., Halama, R., Warren, J.M., Workman, R., Xu, W., Hirschmann, M.M., Larson, A.M., 2007. Correlation of seismic and petrologic thermometers suggests deep thermal anomalies beneath hotspots. Earth Planet. Sci. Lett. 264 (1), 308–316. Dahl-Jensen, T., Larsen, T.B., Woelbern, I., Bach, T., Hanka, W., Kind, R., Gregersen, S., Mosegaard, K., Voss, P., Gudmundsson, O., 2003. Depth to Moho in Greenland: receiver-function analysis suggests two Proterozoic blocks in Greenland. Earth Planet. Sci. Lett. 205 (3–4), 379–393. Darbyshire, F.A., Larsen, T.B., Mosegaard, K., Dahl-Jensen, T., Gudmundsson, O., Bach, T., Gregersen, S., Pedersen, H.A., 2004. A first detailed look at the Greenland lithosphere and upper mantle, using Rayleigh wave tomography. Geophys. J. Int. 158 (1), 267–286. Du, Z., Vinnik, L.P., Foulger, G.R., 2006. Evidence from P-to-S mantle converted waves for a flat “660-km” discontinuity beneath Iceland. Earth Planet. Sci. Lett. 241 (1–2), 271–280. Efron, B., Tibshirani, R., 1991. Statistical data analysis in the computer age. Science 253, 390–395. Henriksen, N. (Ed.), 2008. Geological History of Greenland – Four Billion Years of Earth Evolution. Geological Survey of Denmark and Greenland. Henriksen, N., Higgins, A., Kalsbeek, F., Pulvertaft, T.C.R., 2009. Greenland From Archaean to Quaternary: Descriptive Text to the 1995 Geological Map of Greenland, 1: 2 500 000. Hirose, K., 2002. Phase transitions in pyrolitic mantle around 670-km depth: implications for upwelling of plumes from the lower mantle. J. Geophys. Res. Solid Earth 107 (B4). http://dx.doi.org/10.1029/2001JB000597. Jenkins, J., Cottaar, S., White, R.S., Deuss, A., 2016. Depressed mantle discontinuities beneath Iceland: evidence of a garnet controlled 660 km discontinuity? Earth Planet. Sci. Lett. 433, 159–168. Jolley, D.W., Bell, B.R., 2002. The evolution of the North Atlantic Igneous Province and the opening of the NE Atlantic rift. Geol. Soc. Lond. Spec. Publ. 197 (1), 1–13. Jordan, T.H., 1975. The continental tectosphere. Rev. Geophys. 13 (3), 1–12. Jordan, T.H., Paulson, E.M., 2013. Convergence depths of tectonic regions from an ensemble of global tomographic models. J. Geophys. Res. Solid Earth 118 (8), 4196–4225. Kennett, B.L.N., Engdahl, E.R., 1991. Traveltimes for global earthquake location and phase identification. Geophys. J. Int. 105 (2), 429–465. Keshav, S., Gudfinnsson, G.H., Presnall, D.C., 2011. Melting phase relations of simplified carbonated peridotite at 12–26 GPa in the systems CaO–MgO–SiO2–CO2 and CaO–MgO–Al2O3–SiO2–CO2: highly calcic magmas in the transition zone of the Earth. J. Petrol. 52 (11), 2265–2291. Lawver, L.A., Müller, R.D., 1994. Iceland hotspot track. Geology 22, 311–314. Lebedev, S., Schaeffer, A.J., Fullea, J., Pease, V., 2017. Seismic tomography of the Arctic region: inferences for the thermal structure and evolution of the lithosphere. Geol. Soc. Lond., Spec. Publ. 460, SP460–10. Rickers, F., Fichtner, A., Trampert, J., 2013. The Iceland–Jan Mayen plume system and its impact on mantle dynamics in the North Atlantic region: evidence from full-waveform inversion. Earth Planet. Sci. Lett. 367, 39–51. Rocha, M.P., Schimmel, M., Assumpção, M., 2011. Upper-mantle seismic structure beneath SE and Central Brazil from P-and S-wave regional traveltime tomography. Geophys. J. Int. 184 (1), 268–286.
39
Tectonophysics 728–729 (2018) 34–40
H.A. Kraft et al.
Steinberger, B., Corfu, F., Ashwal, L.D., Griffin, W.L., et al., 2015. Continental crust beneath southeast Iceland. Proc. Natl. Acad. Sci. 112 (15), E1818–E1827. Vinnik, L.P., 1977. Detection of waves converted from P to SV in the mantle. Phys. Earth Planet. Inter. 15 (1), 39–45. Vinnik, L.P., Farra, V., Kind, R., 2004. Deep structure of the Afro-Arabian hotspot by S receiver functions. Geophys. Res. Lett. 31, L11608. http://dx.doi.org/10.1029/ 2004GL019574. Vinnik, L.P., Foulger, G.R., Du, Z., 2005. Seismic boundaries in the mantle beneath Iceland: a new constraint on temperature. Geophys. J. Int. 160 (2), 533–538. Vinnik, L., Silveira, G., Kiselev, S., Farra, V., Weber, M., Stutzmann, E., 2012. Cape Verde hotspot from the upper crust to the top of the lower mantle. Earth Planet. Sci. Lett. 319, 259–268. Vinnik, L.P., Oreshin, S.I., Makeyeva, L.I., 2017. Siberian traps: hypotheses and seismology data. Izv. Phys. Solid Earth 53 (3), 332–340. Wittlinger, G., Farra, V., 2012. Observation of low shear wave velocity at the base of the polar ice sheets: evidence for enhanced anisotropy. Geophys. J. Int. 190, 391–405.
Rogozhina, I., Petrunin, A.G., Vaughan, A.P., Steinberger, B., Johnson, J.V., Kaban, M.K., Calov, R., Rickers, F., Thomas, M., Koulakov, I., 2016. Melting at the base of the Greenland ice sheet explained by Iceland hotspot history. Nat. Geosci. 9 (5), 366–369. Shen, Y., Solomon, S.C., Bjarnason, I.T., Nolet, G., Morgan, W.J., Allen, R.M., Vogfjörd, K., Jakobsdóttir, S., Stefánsson, R., Julian, B.R., Foulger, G.R., 2002. Seismic evidence for a tilted mantle plume and north–south mantle flow beneath Iceland. Earth Planet. Sci. Lett. 197 (3), 261–272. Shulgin, A., Thybo, H., 2015. Seismic explosion sources on an ice cap–technical considerations. Polar Sci. 9 (1), 107–118. Thybo, H., Nielsen, L., Perchuc, E., 2003. Seismic scattering at the top of the mantle Transition Zone. Earth Planet. Sci. Lett. 216, 259–269. Torsvik, T.H., Carlos, D., Mosar, J., Cocks, L.R.M., Malme, T.N., 2002. Global reconstructions and North Atlantic paleogeography 440 Ma to recent. In: BATLAS—Mid Norway Plate Reconstruction Atlas With Global and Atlantic Perspectives, pp. 18–39. Torsvik, T.H., Amundsen, H.E., Tronnes, R.G., Doubrovine, P.V., Gaina, C., Kusznir, N.J.,
40