Radial viscous fingering of hot asthenosphere within the Icelandic plume beneath the North Atlantic Ocean

https://doi.org/10.1016/j.epsl.2017.03.036Get rights and content

Highlights

  • Tomographic images show 5 horizontal low-Vs fingers radiating away from Iceland.

  • Residual depth and Vs estimates constrain temperature and mobility variation.

  • Thickness, mobility and no. of fingers are consistent with experimental observations.

  • Radial miscible viscous fingering may explain irregular planform of Icelandic plume.

Abstract

The Icelandic mantle plume has had a significant influence on the geologic and oceanographic evolution of the North Atlantic Ocean during Cenozoic times. Full-waveform tomographic imaging of this region shows that the planform of this plume has a complex irregular shape with significant shear wave velocity anomalies lying beneath the lithospheric plates at a depth of 100–200 km. The distribution of these anomalies suggests that about five horizontal fingers extend radially beneath the fringing continental margins. The best-imaged fingers lie beneath the British Isles and beneath western Norway where significant departures from crustal isostatic equilibrium have been measured. Here, we propose that these radial fingers are generated by a phenomenon known as the Saffman–Taylor instability. Experimental and theoretical analyses show that fingering occurs when a less viscous fluid is injected into a more viscous fluid. In radial, miscible fingering, the wavelength and number of fingers are controlled by the mobility ratio (i.e. the ratio of viscosities), by the Péclet number (i.e. the ratio of advective and diffusive transport rates), and by the thickness of the horizontal layer into which fluid is injected. We combine shear wave velocity estimates with residual depth measurements around the Atlantic margins to estimate the planform distribution of temperature and viscosity within a horizontal asthenospheric layer beneath the lithospheric plate. Our estimates suggest that the mobility ratio is at least 20–50, that the Péclet number is O(104), and that the asthenospheric channel is 100±20km thick. The existence and planform of fingering is consistent with experimental observations and with theoretical arguments. A useful rule of thumb is that the wavelength of fingering is 5±1 times the thickness of the horizontal layer. Our proposal has been further tested by examining plumes of different vigor and planform (e.g. Hawaii, Cape Verde, Yellowstone). Our results support the notion that dynamic topography of the Earth's surface can be influenced by fast, irregular horizontal flow within thin, but rapidly evolving, asthenospheric fingers.

Introduction

It is generally agreed that a substantial convective upwelling or plume centered beneath Iceland has had a significant effect on the stratigraphic evolution of the North Atlantic Ocean (White and McKenzie, 1989; Jones et al., 2012). This plume developed during Early Cenozoic times and its inception is usually linked with the appearance of basaltic magmatism at 64 Ma. It is bisected by a mid-oceanic ridge which provides a helpful window into the detailed temporal evolution of this globally significant feature (Parnell-Turner et al., 2014). Fluctuations in plume activity over the last 50 Ma are recorded in the pattern of diachronous V-shaped ridges that are imaged in the oceanic basins on either side of the Reykjanes Ridge. During the Neogene period, regional bathymetric changes associated with these fluctuations appear to have moderated overflow of Northern Component Water, the ancient precursor of North Atlantic Deep Water (Poore et al., 2011).

The present-day planform of the Icelandic plume is determined from a combination of three different sets of observations (Fig. 1). The simplest and most striking manifestation is the pattern of long wavelength (700–2500 km) free-air gravity anomalies. A positive anomaly of 30–50 mGal is centered on Iceland. Together, other anomalies form an irregular planform that reaches from Baffin Island to western Scandinavia, and from the Charlie-Gibbs fracture zone to Svalbard. The inference that this pattern of long wavelength anomalies is a manifestation of mantle convective upwelling is strengthened by the existence of significant residual depth anomalies throughout adjacent oceanic basins. Hoggard et al. (2017) built a database of seismic reflection and wide-angle profiles that they used to accurately calculate water-loaded depths to oceanic basement as a function of plate age. In this way, residual depth anomalies are determined that build upon previous analyses (White, 1997; Marquart and Schmeling, 2004). These combined results show that oceanic crust surrounding Iceland is considerably shallower than expected (Fig. 1a). For example, residual depth anomalies of up to 2 km are recorded adjacent to Iceland. This regional shallowing dies out gradually with increasing distance from Iceland. The match between residual depth measurements and long wavelength gravity anomalies is reasonable, although a notable exception is observed north of Greenland. The relationship between the gravity field and residual depth measurements suggests that the water-loaded admittance is Z+25 mGal km−1, in agreement with global studies (Crosby and McKenzie, 2009).

Finally, the presence of a mantle convective anomaly is corroborated by earthquake tomographic models which suggest that an extensive and irregular patch of low shear wave velocity lies beneath the lithospheric plates (Bijwaard and Spakman, 1999; Ritsema et al., 2011). The most striking of these studies is that of Rickers et al. (2013) who use full-waveform tomography to build a high resolution shear wave velocity model of the North Atlantic region from the surface to a depth of 1300 km (Fig. 1b). A significant negative velocity anomaly of >10% with respect to their reference model is centered beneath Iceland, in agreement with earlier studies. One notable feature of their model is the existence of narrow, slow velocity fingers that protrude beneath the fringing continental margins. Two prominent fingers reach beneath the British Isles and western Norway. In both cases, the associated negative shear wave velocity anomalies are >2% and sit within a 100±20 km thick horizontal layer immediately beneath the lithospheric plate (Fig. 2). Rickers et al. (2013) show that there is a reasonable match between the loci of these fingers and long wavelength gravity anomalies. Significantly, both fingers also coincide with crustal isostatic anomalies and with the general pattern of Neogene vertical movements observed across the northwest shelf of Europe (Anell et al., 2009; Davis et al., 2012). In the southern North Sea, the fast (i.e. cooler) region between these fingers has a water-loaded subsidence anomaly of ∼500 m that grew in Neogene times and represents a significant departure from the expected thermal subsidence trajectory (Fig. 2b–d; Kooi et al., 1991). This region probably subsided as a result of small-scale convective downwelling between the two warm fingers.

Here, we combine these different geologic and geophysical observations to investigate the causes and consequences of radial fingering within the asthenospheric mantle. In a series of contributions pioneered by Weeraratne et al. (2003), it has been suggested that some combination of rectilinear viscous fingering instabilities, small-scale convection, and shear-driven upwelling may play a role in explaining the observed pattern of seismic velocity anomalies beneath the southern portion of the East Pacific Rise (Weeraratne et al., 2007; Harmon et al., 2011; Ballmer et al., 2013). Although there are significant geometric and mechanical differences, our analysis evidently builds upon these previous contributions and upon the analysis of Morgan et al. (2013).

Our approach is divided into three steps. First, we present the physical characteristics of the Icelandic plume, such as its size, shape and vigor. By combining the correlation between shear wave velocity anomalies and the pattern of regional Neogene epeirogeny with a global empirical relationship between shear wave velocity and temperature, we estimate how viscosity within the plume head spatially varies. Secondly, we compare these observations of plume behavior beneath Iceland and elsewhere with published laboratory experiments that investigate the development of radial miscible viscous fingering. Thirdly, the development of radial fingering is discussed using a suite of theoretical and heuristic arguments. We conclude by exploring the implication of our hypothesis for a small selection of well-known plumes.

Section snippets

Physical characteristics of Icelandic plume

The temperature structure of the Icelandic plume can be estimated in a variety of related ways. In the North Atlantic Ocean, a mid-oceanic spreading center transects this plume and provides the most straightforward method for determining this structure. Within the region of influence, the average thickness of oceanic crust increases from 7 to 14 km and the seabed is anomalously shallow by up to 2 km (Fig. 1; White, 1997). Both of these observations are consistent with an average temperature

Saffman–Taylor instability

When a less viscous fluid displaces a more viscous fluid, the boundary between the two fluids can become unstable and promote viscous fingering (Saffman and Taylor, 1958). A considerable amount of experimental and theoretical work has been carried out on this fingering process for a variety of geometries under different dynamic conditions. The general aim is to predict conditions under which fingering occurs and to estimate the number of fingers that develop (Homsy, 1987). Here, the relevant

Discussion

Fig. 5 captures the idealized geometry of the Icelandic plume. On Iceland itself, the putative conduit is located at Vatnajökull in southeast Iceland and has a diameter of ∼100 km (Fig. 1; Shorttle et al., 2010). We have shown that M=24±10, Pe=(2.1±0.4)×104 and b=100±20 km for the Icelandic plume. By combining this scaling analysis with the results of laboratory experiments on radial miscible viscous fingering, we suggest that a Saffman–Taylor instability will manifest itself at the distal

Conclusions

We use a combination of geophysical and geologic observations from the North Atlantic Ocean to confirm that the Icelandic plume has an irregular planform. Sub-plate and physiographic evidence shows that about five radial fingers of hot asthenosphere protrude beneath adjacent continental margins. A quantitative comparison with appropriately scaled laboratory experiments suggests that these fingers are generated by the classic Saffman–Taylor instability. This manifestation of viscous fingering

Correspondence

Correspondence and requests for materials should be addressed to N.J. White.

Author contributions

This project was conceived and managed by NJW. CMS processed data with guidance from NJW. DP provided mathematical analysis and insight in consultation with NJW. The paper was written by NJW with contributions from CMS and DP.

Competing financial interests

The authors declare no competing financial interests.

Acknowledgments

CMS is supported by Shell Exploration and by the British Geological Survey. We thank M. Hoggard, C. Richardson, and A. Woods for their help. M. Ballmer provided a constructive review. This paper is dedicated to Maeve White. Figures were prepared using Generic Mapping Tools. Cambridge Earth Sciences contribution number esc. 3904.

References (50)

  • R.B. Smith et al.

    Geodynamics of the Yellowstone hotspot and mantle plume: seismic and GPS imaging, kinematics, and mantle flow

    J. Volcanol. Geotherm. Res.

    (2009)
  • U.S.N. Arachchige

    Formation of Intraplate Seamount Chains by Viscous Fingering Instabilities in the Asthenosphere Using Low Reynolds Number Miscible Fluids with a Moving Surface Boundary

    (2016)
  • M.D. Ballmer et al.

    Non-hotspot volcano chains produced by migration of shear-driven upwelling toward the East Pacific Rise

    Geology

    (2013)
  • M.H.P. Bott et al.

    The Cenozoic uplift and earthquake belt of mainland Britain as a response to an underlying hot, low-density upper mantle

    J. Geol. Soc.

    (2004)
  • J.D. Chen

    Growth of radial viscous fingers in a Hele–Shaw cell

    J. Fluid Mech.

    (1989)
  • J. Chui

    Understanding the Evolution of Miscible Viscous Fingering Patterns

    (2012)
  • J.Y.Y. Chui et al.

    Interface evolution during radial miscible viscous fingering

    Phys. Rev. E

    (2015)
  • R.C. Courtney et al.

    Anomalous heat flow and geoid across the Cape Verde Rise: evidence for dynamic support from a thermal plume in the mantle

    Geophys. J. R. Astron. Soc.

    (1986)
  • A.G. Crosby et al.

    An analysis of young ocean depth, gravity and global residual topography

    Geophys. J. Int.

    (2009)
  • M.W. Davis et al.

    Crustal structure of the British Isles and its epeirogenic consequences

    Geophys. J. Int.

    (2012)
  • R.A. Greenkorn et al.

    Two-phase flow in Hele–Shaw models

    AIChE J.

    (1967)
  • G. Hirth et al.

    Rheology of the Upper Mantle and the Mantle Wedge: A View from the Experimentalists

    (2003)
  • M.J. Hoggard et al.

    Oceanic residual depth measurements, the plate cooling model, and global dynamic topography

    J. Geophys. Res., Solid Earth

    (2017)
  • K.E. Holloway et al.

    Viscous fingering with a single fluid

    Can. J. Phys.

    (2005)
  • P.M. Holm et al.

    Sampling the Cape Verde mantle plume: evolution of melt compositions on Santo Antão, Cape Verde Islands

    J. Petrol.

    (2006)
  • Cited by (38)

    • Is the tilt of the Thames valley towards the east caused by a distal lobe of the Icelandic mantle plume?

      2023, Proceedings of the Geologists' Association
      Citation Excerpt :

      In contrast with uplift onshore, the North Sea is identified by Schoonman et al. (their Fig. 1) as an area of anomalous subsidence. “This [North Sea] region probably subsided as a result of small-scale convective downwelling between the two [British and Norwegian] warm fingers” (Schoonman et al., 2017, page 52). The eastward tilt of the Red Crag in the Thames Valley onshore continues offshore beneath the southern North Sea, but at a steeper angle.

    • Chapter 2 Scottish mineral Geological Conservation Review Sites – Magmatic Minerals

      2022, Proceedings of the Geologists' Association
      Citation Excerpt :

      Helium isotopic studies of garnet megacrysts by Kirstein et al. (2004) support the view that the Permo-Carboniferous suite of upper mantle xenoliths were contaminated by a lower mantle plume under Scotland. However, advanced seismographic techniques clearly indicate that the Icelandic mantle plume has an irregular shape with radial fingers extending outwards of which one occurs underneath Scotland’s thin crust (Schoonman et al., 2017). Gernon et al. (2013, 2016) proposed that the REE patterns signify that the garnet was retained in a mantle source when peridotite melted in the mantle garnet-lherzolite facies.

    • Postglacial relative sea level change in Norway

      2022, Quaternary Science Reviews
      Citation Excerpt :

      Though the duration and magnitude of these effects would depend on the amount of ice change and the solid earth structure underlying the region, estimates from GIA modeling suggest that between 13 and 11.6 ka EIS readvance diminished isostatic uplift in parts of southwest Norway by as much as 25 m, while the increase in gravitational attraction raised RSL by less than 1 m (Fjeldskaar and Amantov, 2018). Recent tomographic studies in Norway suggest a strong northeast-southwest gradient in seismic wave speed along the Norwegian coast, which likely translates into temperature and mantle viscosity (e.g. Barnhoorn et al., 2011; Schoonman et al., 2017; Lei et al., 2020). Though a cold continental root underlies much of Scandinavia, asthenospheric fingers radiating from the Iceland plume may have weakened the earth structure beneath southern Norway (Rickers et al., 2013).

    • Control of the Paleogene sedimentary record of the Anglo-Paris Basin by both the Iceland mantle plume and the Massif Central hotspot

      2020, Proceedings of the Geologists' Association
      Citation Excerpt :

      The suggested correlation between lithospheric thickness and igneous activity is also compatible with the suggestion by Al Kindi et al. (2002) that a finger of the Paleogene Iceland plume extended from north to south through the area of the present-day Irish Sea. Schoonman et al. (2017) show a similar hot finger from the present-day Iceland plume extending north to south through the same area. As with the Massif Central, these observations to the northwest indicate present-day dynamic support of the flanks of the southwestern arm of the Palaeogene North Sea (Fig. 6).

    View all citing articles on Scopus
    View full text