Roadmap | The review process | Home
   Subduction magmatism
Subduction-triggered magmatic pulses: a new class of plumes?

C. Faccenna1, T.W. Becker2, S. Lallemand3, Y. Lagabrielle3, F. Funiciello1 & C. Piromallo4

1Dip. Scienze Geologiche, Università “Roma Tre” and  CNR, IGAG-Rome, Italy, ;

2Dep. Earth Sciences, University Southern California, Los Angeles, USA,

3Laboratoire Geosciences Montpellier, Université Montpellier II, CNRS, France, ;

4Istituto Nazionale di Geofisica e Vulcanologia, Roma, Italy,


This webpage is a summary of: Faccennaa, C., T.W. Beckerb, S. Lallemandc, Y. Lagabriellec, F. Funiciellod and C. Piromallo, Subduction-triggered magmatic pulses: A new class of plumes?, Earth Planet. Sci. Lett., 54-68, 2010.



A variety of thermal plumes - splash (Davies & Bunge, 2006), baby (e.g., Wilson & Dowes, 2006) and edge (King & Ritsema, 2000) plumes - have been recently described as focused vertical upwellings that are not directly rooted in the lower mantle, as predicted by the original hotspot model (Morgan, 1971). Moreover, high-resolution seismological images have shown many apparent small-scale convective heterogeneities in the uppermost mantle at margins such as the western US (e.g., Sigloch et al., 2008; West et al., 2009) and the Mediterranean (Faccenna & Becker, 2010). Small scale mantle flow could explain also the velocity distribution beneath the Rio Grande Rift  (van Wijk et al., 2008), the uplift and magmatism of the Colorado plateau (Roy et al., 2009) and extension beneath the Great basin (West et al., 2009). Subcontinental small-scale convection may be excited by sharp temperature gradients at a craton’s edge, where decompression melting may cause volcanism (King & Anderson, 1995). In this work, we suggest an indirect connection to slab return flow, which may interact with a hydrated layer in the transition zone (Lehay & Bercovici, 2007) to facilitate localized upwellings.

Subduction-related upwelling

In several regions, there is evidence for volcanism that is spatially and temporally connected to subduction zones but not associated with mantle wedge melting. Related magmas show ocean island basalt (OIB)-type signatures developing from volcanoes located either far off the arc, ahead of the trench, or at slab edges. Relationships between subduction and anomalous volcanism, though already postulated to explain regional cases of intraplate magmatic activity (e.g., Changbai volcano, East Asia: Zhao et al., 2009; New Hebrides-North Fiji: Lagabrielle et al., 1997; Mediterranean-European: Goes et al., 1999, Piromallo et al., 2008, Lustrino & Wilson, 2007), have never been framed and modeled in a subduction-related convecting system.

We demostrate that subduction within the upper mantle triggering return flow can generate focused, sub-lithospheric, non-thermal mantle upwellings.  The time-dependent evolution of subduction-induced mantle circulation has been explored running 3D dynamically self-consistent numerical models using the finite element code CitcomCU (e.g., Zhong et al., 1998; Moresi & Solomatov, 1995). In the adopted simplified setup, an old single slab fixed in the far field sinks into the mantle. This is approximated by a visco-plastic material with temperature-dependent viscosity. The evolution of slab sinking into the mantle (Figure 1) shows a progressive increase of the subduction velocity during the fall of the slab into the upper mantle (Becker et al., 1999). The sinking slows once the slab reaches the base of the upper mantle and it temporarily ponds where the viscosity increases (Funiciello et al., 2003). Subsequently, the subduction process is taken up by trench rollback. As discussed by Funiciello et al. (2004), the pattern of mantle circulation is strongly variable during the three aforementioned evolutionary stages. To describe better subduction-induced mantle flow, it is instructive to perform decomposition into the toroidal and poloidal components (e.g., Tackley, 2000).

Figure 1: Evolution of the reference numerical model. Each panel is composed of two parts. The upper part shows the lateral cross-section of the model taken through the middle of the plate. The color plot gives the magnitude of the non-dimensional lithospheric temperature, assuming the initial mantle temperature is fixed at 1. Arrows illustrate the x-z flow pattern in the mantle. The lower part shows the horizontal cross-sections taken at x km depths from the top. In this case, the color plot gives the magnitude of the vertical velocity component. Arrows illustrate the x-y flow pattern in the mantle. Non-dimensional units and unity time correspond to ~15 Ma. Click here or on Figure for enlargement.

Our modelling shows that the rapid sinking of the slab during the initial stages of subduction, with a dominance of poloidal flow, induces a convective cell ahead of the slab with a wavelength on the order of the upper mantle thickness. The maximum vigor of the poloidal flow is attained just before the slab encounters the 660-km discontinuity. The toroidal/poloidal ratio (TPR) is less than 0.5 (for selected model parameters) during slab subduction into the upper mantle and reaches maximum values on the order of 0.6 after the slab interaction with the 660-km discontinuity. The poloidal cell involves upwelling components:

  1. ahead of the slab, mainly active during the transient stage of slab descent into the upper mantle; and
  2. along the slab’s edges, active throughout the model.

These upwelling mantle currents, combined with decompression melting, could lead to a thinning of the lithospheric thermal boundary layer and explain the existence/timing of:

  1. some of the anomalous outside-arc alkaline volcanism;
  2. positive non-isostatic topography; and
  3. melting zones seismically recorded in the mantle as low velocity anomalies.

Combining modeling results with natural data from  key regions - western North America (Figure 2), the Central Mediterranean, the North Fiji/Lau Basin, the West Philippine basin, Mt. Etna (Figure 3) and others, we confirmed the physical relationships between subduction and anomalous volcanism.

Figure 2: a) Map of Western North America (redrawn from Xue & Allen, 2007). Three main magmatic provinces can be recognized: the Columbia River Basalts, the Yellowstone Hotspot Track (with red showing the location of magmatic centers and related ages) and the High Lava Plains of Oregon, or Newberry Hotspot Track (with red showing the location of magmatic centers and related ages, and black thin solid lines indicating the location of magmatic centers and related ages). Dashed lines show the depth of the subducting lithosphere. MC indicates McDermitt Caldera. The A-B line indicates the cross-section of panels b-g. b) The tomographic cross-section along the Western North America margin from Li et al. (2008). c-g) Reconstruction of the evolution of the Juan de Fuca subduction zone and of the position of the volcanic centers back over the last 17 Ma, starting from the distribution of the velocity anomaly in panel b). Arrows indicate the net motions of the subducting plate, the overriding plate and the trench. The age of the volcanic centers is expressed in Ma. This shows that the emergence of volcanism occurred just as the slab arrived at the 660-km discontinuity and that the position of the volcanism lines up with the slab tip (Figure 2b). Based on seismological observations and plate-tectonic reconstructions, we speculate that a mechanism of subduction-driven upwelling, similar to the one shown in Figure 1, could be adapted to the Yellowstone case. We also speculate that the massive onset of volcanism in the surrounding area could be triggered by a shallow, upper-mantle source, perhaps triggered by the separation of the oldest portion of the slab and the onset of the new subduction cycle (Sigloch et al., 2008).


Figure 3: Mantle structure beneath and around Mount Etna from the PM01 model (Piromallo & Morelli, 2003). Note the low velocity anomalies located around the Calabrian slab. a) Sections at 150 and 250 km depth. b) Topography and elevation of the MIS 5.5 marine terraces along the section (from Ferranti et al., 2006) illustrating uplift in correspondence of Mount Etna. c) AA' cross section shown in a) of the seismic tomography model, showing a marked low velocity anomaly around the slab. Our modeling results show that the contradiction between these classes of models, slab-induced asthenospheric flow and the hot spot model, is only just apparent. We propose that Mt. Etna can be considered an upwelling structure confined in the upper mantle as a consequence of the complex 3-D mantle circulation triggered by the subducting lithosphere.



The combination of numerical modeling and tectonic reconstructions reveals (Figure 4):

  1. The initiation of strong volcanic off-arc activity is expected during peaks of poloidal mantle circulation. This condition is reached when the slab approaches the transition zone. In this phase, the slab attains its peak velocity (Funiciello et al., 2006). The case of Yellowstone clearly shows this correlation (Figure 2). Volcanism is likely positioned between 600 and 700 km from the trench. During this phase, the source of magmatism is entirely related to decompression melting. At the surface this process is accompanied by  broad uplift, elevated heat flow, and normal faulting.
  2. Volcanism remains active during the entire subduction process. The locus of volcanism can either follow the moving subduction zone or be rather stable once the upwelling is rooted in the high-velocity anomaly stagnating in the transition zone (Figure 4). It is possible that, during this phase, slab dehydration could contribute to melt in the asthenosphere and feed volcanism (Richard & Iwamori, 2010). Petrological and numerical modelling are needed to constrain this  mechanism better.
  3. Slabs can show peculiar volcanism positioned at slab edges. Examples of such kinds of volcanoes are frequent (e.g., South Sandwich, Kamchatka, Taiwan, Tonga and Mt. Etna; Figure 3). The mechanism propelling melting is still decompression, due to the vertical component of the upper mantle circulation around the slab edges being active since the beginning of the subduction process. However, it cannot be excluded that slab dewatering could be another efficient mechanism active in this particular position. Some of the cited natural examples, in fact, exhibit a complex geochemical signature that could reveal mixing between deep components and shallow ones. Shear heating between the slab and the surrounding mantle could be invoked as an alternative interpretation to localized low-velocity zones at slab edges, but its contribution has been calculated to be extremely weak (Rupke et al., 2004).


Figure 4: Cartoons showing scenarios able to trigger decompression melting produced by slab return flow, and the resulting development of off-axis volcanism. a) Slab-mantle interaction; b) Steady-state condition of subduction with the slab lying on the 660 km discontinuity.


  • Becker, T.W., Faccenna, C., O'Connell, R.J., Giardini, D., 1999. The development of slabs in the upper mantle: insight from numerical and laboratory experiments. J. Geophys. Res., 104, 15207-15225.
  • Davies, H.J., Bunge, H.P., 2006. Are splash plumes the origin of minor hotspots? Geology, 34, 349–352, doi: 10.1130/G22193.1.
  • Faccenna, C., Becker, T.W., 2010. Shaping mobile belt from small scale convection. Nature, 465,  doi:10.1038/nature09064.
  • Ferranti, L., et al., 2006. Markers of the last interglacial sea level highstand along the coast of Italy: Tectonic implications. Quat. International, 145–146, 30–54, doi:10.1016/j.quaint.2005.07.009.
  • Funiciello, F., Faccenna, C., Giardini, D., 2004. Flow in the evolution of subduction system: Insights from 3-D laboratory experiments. Geophys. J. Int., 15, 1393–1407.
  • Funiciello, F., Faccenna, C., Giardini, D., Regenauer-Lieb K, 2003. Dynamics of retreating slabs (part 2): insights from 3D laboratory experiments. J. Geophys. Res., 108, doi:10.1029/ 2001JB000896.
  • Funiciello, F., Moroni, M., Piromallo, C., Faccenna, C., Cenedese, A., Bui, H.A., 2006. Mapping flow during retreating subduction: laboratory models analyzed by Feature Tracking. J. Geophys. Res., 111, doi:10.1029/2005JB003792
  • Goes, S., Spakman, W., Bijwaard, H., 1999. A lower mantle source for central European volcanism. Science, 286, 1928–1931.
  • King, S.D., Anderson, D.L., 1995. An Alternative Mechanism of Flood Basalt Formation. Earth Planet. Sci. Lett., 136,  269-279.
  • King, S.D., Ritsema, J., 2000. African hot spot volcanism: Small-scale convection in the upper mantle beneath cratons. Science, 290, 1137-1140.
  • Lagabrielle, Y., Goslin, J., Martin, H., Thirot, J.L.Auzende, J.M., 1997. Multiple active spreading centers in the hot North Fiji Basin (Southweast Pacific): a possible model for Archean seafloor dynamics ? Earth Planet. Sci. Lett., 149, 1–13.
  • Leahy, G.M., Bercovici, D, 2007. On the dynamics of a hydrous melt layer above the transition zone. J. Geophys. Res., 112, B07401.
  • Li, C., van der Hilst, R.D., Engdahl, E.R., Burdick, S., 2008. A new global model for P-wavespeed variations in Earth's mantle. Geochemistry Geophysics Geosystems, 9, Q05018, doi:10.1029/2007GC001806.
  • Lustrino, M., Wilson, M., 2007. The circum-Mediterranean anorogenic Cenozoic igneous province. Earth Sci. Rev., doi:10.1016/j.earscirev.2006.09.002.
  • Moresi, L., Solomatov, V.S., 1995. Numerical Investigations of 2D Convection with Extremely Large Viscosity Variations. Phys. Fluids, 7, 2154-2162.
  • Morgan, W. J., 1971.  Convection plumes in the lower mantle. Nature, 230, 42–43.
  • Piromallo, C, Morelli, A., 2003. P wave tomography of the mantle under the Alpine-Mediterranean area. J. Geophys. Res., 108, 2065, doi: 10.129/2002JB001757
  • Piromallo, C., Gasperini, D., Macera, P., Faccenna, C. 2008. A late Cretaceous contamination episode of the European–Mediterranean mantle. Earth Planet. Sci. Lett., 268, 15–27.
  • Richard, G.C., Iwamori, H., 2010. Stagnant slab, wet plumes and Cenozoic volcanism in East Asia, PEPI,doi:10.1016/j.pepi.2010.02.009
  • Roy, M., Jordan, T.H., Pederson, J. 2009. Colorado Plateau magmatism and uplift by warming of heterogeneous lithosphere. Nature, 459, doi:10.1038/nature08052
  • Rupke, L.H., Morgan, J.P., Hort, M., Connolly, J.A.D., 2004. Serpentine and the subduction zone water cycle. Earth Planet. Sci. Lett., 223, 17-34.
  • Sigloch, K., McQuarrie N., Nolet, G., 2008. Two-stage subduction history under North America inferred from multiple-frequency tomography. Nature Geoscience, 1, 458-462.
  • Tackley, P.J. 2000. Self-consistent generation of tectonic plates in time-dependent, three-dimensional mantle convection simulations 1. Pseudoplastic yielding. Geochemistry Geophysics Geosystems, 1, 2000GC000036.
  • van Wijk, J., van Hunen, J., Goes, S., 2008. Small-scale convection during continental rifting: Evidence from the Rio Grande rift. Geology, 36, 575-578, doi: 10.1130/G24691A.1.
  • West, J.D., Fouch, M.J., Roth, J.B., 2009.  Vertical mantle flow associated with a lithospheric drip beneath the Great Basin. Nature Geoscience, 2, 438-443.
  • Wilson, M., Downes, H., 2006. Teriary–Quaternary intra-plate magmatism in  Europe and its relationships to mantle dynamics. In: Gee, D., Stephenson, R.  (Eds.), European Lithosphere Dynamics. Geol. Soc. Lond. Mem., 32, 147–166.
  • Xue, M., Allen, R. M. 2010. Mantle structure beneath the western United States and its implications for convection processes.  J. Geophys. Research, 115, B07303, doi:10.1029/2008JB006079
  • Zhao, D., Tian,, Y., Lei, J., Liu, L., Zheng, S., 2009. Seismic image and origin of the Changbai intraplate volcano in East Asia: Role of big mantle wedge above the stagnant Pacific slab. Phys. Earth Planet. Inter., 173, 197-206.
  • Zhong, S., Gurnis, M., Moresi, L., 1998. Role of faults, nonlinear rheology, and viscosity structure in generating plates from instantaneous mantle flow models. J. Geophys. Res., 103, 15255–15268.
last updated 23rd February, 2011