Bermuda and the Bermuda Rise – A Poor Fit to the Classical Mantle Plume Model

Peter R. Vogt
Marine Science Institute, University of California at Santa Barbara, Ptr_vogt@yahoo.com

Woo-Yeol Jung
Code 7420, Naval Research Laboratory, Washington, DC 20375-5320

(See Vogt, P.R. & W.-Y. Jung, Origin of the Bermuda volcanoes and Bermuda Rise: History, observations, models, and puzzles, in Plates, Plumes, and Planetary Processes, G.R. Foulger and D.M. Jurdy (Eds.), Geological Society of America Special Paper 430, in press, for more complete coverage of the topics discussed in this webpage).

Click here to download a PDF version of this webpage

Age dating

Age-dating the Bermuda volcanism and rise development plays a crucial role in testing any model, especially the stationary deep mantle plume model. Boreholes on Bermuda, as well as DSDP Site 386 located in a fracture valley 140 km SE of Bermuda (Tucholke & Vogt, 1979) indicate that submarine eruption building up the shield began no later than the late Middle or early Late Eocene, when erosional debris from the emergent edifice first arrived at Site 386. While K-Ar whole-rock dates of  52 Ma (1958 borehole; Gees, 1969) and   47 and 91 ± 5 Ma (1972 borehole; Reynolds & Aumento, 1974) were reported for these pillow lavas, their alteration makes these dates very unreliable. In a later igneous episode, highly titaniferous dikes (sheets) were intruded into the edifice. However, the sheets have yielded consistent and credible dates  of  33-34 Ma, confirmed by recent Ar-Ar dating reported by Williamson et al. (2006). Phlogopite weathered out of the eroding sheets first arrived at Site 386 (The Shipboard Scientific Party, 1979) in the middle to upper Oligocene (ca 30-25 Ma), consistent with the radiometric dates. The initiation of Bermuda Rise uplift has been stratigraphically dated (early to middle part of the Middle Eocene, i.e., ca 48-45 Ma; The Shipboard Scientific Party, 1979; Tucholke & Vogt, 1979) by the cessation of biogenic turbidites that had covered the central and western parts of the present rise area. The initiation of igneous activity and uplift may therefore have been coeval. However, while evidence to date suggests Eocene shield formation was succeeded by 33-34 Ma intrusions, analysis of seismic reflection returns from successively tilted turbidites in the Kane Fracture Zone valley suggests that Bermuda Rise uplift continued into Miocene, with 400-500 m of uplift occurring after the sheet intrusion phase.

A Bermuda Plume?

Bermuda ranks very low when observational data are compared to predictions of the mantle plume model (Courtillot et al., 2003; Anderson, 2005). Both the 100-km-long volcanic lineament and the 1500-km-long Bermuda Rise trend northeast, nearly at right angles to the trace predicted from North America "absolute" motion models (Figures 1 and 2; Morgan, 1983; Duncan, 1984; Mueller et al., 1993). Bathymetry shows no evidence for any geologically young igneous activity along the predicted trace, which extends east  650 km from Eocene Bermuda to the predicted present site of a putative Bermuda hotspot (Vogt & Jung, 2007). Moreover, Jaroslaw & Tucholke (1994) found no evidence for migratory uplift associated with the Bermuda Rise. The shallow depth of volcanic basement under Bermuda shows that little or no subsidence has occurred. Morgan & Crough (1979), who were aware of the lack of a hotspot trace, proposed a causal relation between uplift of the Cape Fear Arch on the North American margin, and the passage of the plate over a Bermuda plume; however, more recent studies of this arch are inconsistent with this idea. Those authors also related Cretaceous kimberlites and other igneous rocks in Eastern North America to the passage of the plate over a Bermuda plume.  

Some authors (e.g., Cox & Van Arsdale, 2002) have held on to the Morgan-Crough Bermuda plume model, which would require a a pulsating  (“lava lamp”) plume and/or  severe control by lithospheric structure on melts rising into or erupting onto the crust. In support,  Cox & Van Arsdale (2002) note that predicted Bermuda hotstpot tracks cross the ca. 65 Ma igneous activity in Mississippi, and also the ca 115 Ma activity in Kansas (Figure 2). However, there is no “LIP” (large igneous plateau) in Kansas or elsewhere that might represent the effects of a  “plume head” at the beginning of the putative Bermuda plume track.  Moreover, Cretaceous and Cenozoic igneous rocks elsewhere in North America (some are noted in simplified form in Figure 2) would necessarily require other plumes (McHone, 1996) [Ed: See also CAMP page].

Figure 2.   Alternative traces of the Bermuda “hotspot” relative to the North America plate, as predicted by the plate/mantle motion models of Morgan (1983), Duncan (1984) and Mueller et al. (1993), with filled circles indicating the predicted hotspot location at 0, 50 and 100 Ma for each trace. Thin lines within the Mississippi Embayment are the Mississippi Valley graben faults, and open squares show the average location of the earliest (ca. 115 Ma) and latest (ca. 65 Ma) eruptions along a migratory path of igneous activity. CFA and GSD denote Cape Fear Arch and Great Stone Dome. Modified from Figure 1 of Cox & van Arsdale (2002).

Models

A number of attempts have been made – mostly in the late 1970s-1980s – to model the formation and present state of the Bermuda Rise, in many cases along with some other, similar mid-plate swells (e.g., Crough, 1978; Sclater & Wixon, 1986; Detrick et al., 1986; Louden et al., 1987; Liu & Chase, 1989; Sheehan & McNutt, 1989). In any case, the geoid and depth anomalies require the Bermuda Rise to be supported by a low density root at depth within the lithosphere or/and upper asthenosphere – it is not thickened oceanic crust. If anything, the crust under the Bermuda Rise (except at the Bermuda volcanoes) is anomalously thin (Lizzaralde et al., 2004). Little or no low-seismic-wave-speed anomaly has been detected by tomography in the mantle below Bermuda (Zhao, 2004). The lack of current subsidence and the low or nonexistent heat flow anomaly have been problematic for most models. However, published models suggest that certain combinations of thermal expansion, dynamic uplift, and perhaps residual lower-density melt accumulation below the rise, may account for observations. (The latter was proposed by Phipps Morgan et al., 1995, for the Hawaii and Cape Verde "hotspots"). Thermal expansion by itself cannot explain the rise except in combination with convection – purely conductive models require too high a temperature anomaly.  Some of the models have to be rejected or at least refined to account for continuous but non-migratory Eocene-Miocene uplift (Jaroslaw & Tucholke, 1994) and possible lack of a mantle-related heatflow anomaly.

Non-plume alternatives

If Bermuda was not formed by a pulsating lava-lamp-type plume, what can account for the observations? Whatever is happening under Bermuda is evidently traveling with the North America plate. Sclater & Wixon (1986), Vogt (1991), King & Anderson (1998) and King (2007) [Ed: See also EDGE page] have examined an "edge-driven" convection that models predict should be generated where thick, cold continental lithosphere abuts an oceanic plate, once the ocean has widened sufficiently. The models predict rising convection currents ca. 1000 km seawards of the boundary. The persistent Eocene to Miocene uplift of the Bermuda Rise, as well as the rise location and orientation, are qualitatively consistent with edge-driven convection.

Vogt & Jung (2007) note that Bermuda Rise formation and volcano formation were coeval, within error bounds, with the Hawaii-Emperor Bend (long dated at ca 43 Ma, but recently redated to ca. 50 Ma; Sharp & Clague, 2006). This and other coeval events might represent an abrupt rearrangement of mantle convection, plate motions, and intra-plate stresses  triggered by the closing of the Tethys at about that time. Such a global rearrangement has been suggested by Rona & Richardson (1978), among others. Alternatively or relatedly, slabs accumulating near the 660-km discontinuity may have sunk into the lower mantle at that time (Fukao et al., 2001), triggering rearrangements of mantle flow.

While such a global change might have triggered the "Bermuda event", what can explain its geographic location and rise orientation? Edge-driven convection (e.g., King, 2007) can explain this, as noted. "Lithosphere pre-conditioning" is another (Vogt & Jung, 2007). Both the rise and the volcanic lineament parallel the structure of the underlying oceanic crust, i.e. the isochrons. The rise was developed in crust and mantle lithosphere known to have been formed at very low spreading rates, and exhibiting the rough basement typical of slow spreading. Lizzaralde et al. (2004) found this crust to be anomalously thin, with the underlying mantle atypical. This led Vogt & Jung (2007) to speculate that the nature of the mantle lithosphere may have made it more vulnerable to partial melting at depth. Alternatively, the mantle below the North American plate that happened to underlie the Bermuda rise region during the Eocene might have been anomalous in composition and/or temperature.

Future research opportunities

How can future research discriminate among the wealth of models so far presented in the literature, and perhaps others still to be developed? First, earthquake-source seismic tomography with even better resolution than e.g., that presented in Fig. 19 of Zhao (2004) may delineate a "slow" region in the upper 400 km of the mantle below the Bermuda Rise. Below some such depth, the mantle must be decoupled from the motion of the North America plate. Explosion-source seismic experiments with OBS arrays on the rise, expensive and perhaps not feasible due to concerns for marine mammals, would probably be needed to test for a buoyant refractory root (e.g., Phipps Morgan et al., 1995; Holm, 2006) which might extend from ca. 50 km (the depth of swell compensation deduced from geoid data) to 200 km below the rise. (The depth of origin of the Bermuda sheets is ≥ 150 km; Olsen, 2005). Airgun-source multichannel profiling of the type conducted by Lizarralde et al. (2004) across the southern Bermuda Rise might detect cooled intrusions which reached the upper mantle lithosphere. However, any reduced wave-speed anomalies would have to allow for melt retention under the slow spreading rates forming most of the crust under the Bermuda Rise (Lizarralde et al., 2004). Spreading-rate-dependent mantle velocities should change only gradually northwards along isochrons, whereas any anomalies associated with the Bermuda Rise should reach extrema under the rise summit. Future deep seismic experiments in the ocean surrounding Bermuda are influenced by concerns – whether scientifically justified or not – that the airgun sources will adversely impact marine mammals, particularly whales.

Additional drilling into the igneous basement of Bermuda and its three smaller satellites (Plantagenet/Argos and Challenger banks, and Bowditch seamount) is essential. Given the geologic complexity we know from volcanic islands (e.g., the Cape Verde archipelago; Holm et al., 2006), it seems highly unlikely that even the Bermuda edifice was formed by a simple two-stage process of an Eocene tholeiitic shield, followed after ca. 6-12 Ma of quiescence by the 33-34 Ma "Bermudite" (Aumento & Gunn, 2006) sheet intrusions.

Further, we have assumed – with no direct evidence! – that the three satellite edifices are of the same age as Bermuda.  This remains pure speculation until they (and/or their flanking debris aprons) are cored, preferably to several km depth. A detailed magnetic and gravity survey of all four edifice summits and upper flanks should be conducted in advance of any drilling or further deep-tow, AUV, or manned submersible investigation, to help map the structural layout of intrusive sheets, lava accumulations, central conduits and flanking volcaniclastic debris aprons. We have proposed such work in cooperation with the Bedford Institute of Oceanography (Drs. Steve Blasco and Marie-Claude Williamson).

Several deep holes similar to those drilled at DSDP Site 386 (Figure 1) and 385 (New England Seamounts; Tucholke & Vogt, 1979) should also be placed around the bases of the four Bermuda edifices, as close as possible to the bases, but still practically penetrating the volcaniclastic debris and flows (i.e. the "inner" seismic reflector Av of Tucholke & Mountain, 1979), to recover and biostratigraphically date the youngest sediments overlain by the oldest Bermudan rocks. Recovery of larger, less altered rock fragments would also be more likely closer to the base of the volcanic edifices.

A transect of a few boreholes across and along the Bermuda Rise, just deep enough to sample the oldest hemipelagic sediment just above the Eocene biosiliceous turbidites, should be able to resolve the detailed spatial-temporal pattern of BR uplift initiation. Such boreholes might also recover the time when bottom currents were first steered by the BR (e.g., Ayer & Laine, 1982). Several boreholes should be placed along the Kane fracture zone to calibrate the uplift history deduced by Jaroslaw & Tucholke (1994) from seismic reflection mapping of more local turbidites. Abyssal plains, with gradients of 1:1000 or less (the present  BR is surrounded on three sides by modern or at least late Pleistocene abyssal plains; Pilkey & Cleary, 1986), should be extremely sensitive to small elevation changes. The Mid-Eocene turbidite offlap pattern – in time and space – would depend on the uplift mechanism.

A plume model predicts rise uplift migrating radially outwards from a region above the upwelling plume head, whereas a "distributed source" model with simultaneous partial melting or/and temperature increase would predict simultaneous uplift over the entire area of the present Bermuda Rise. A plume-type model (Griffiths & Campbell, 1991; Campbell, 2006) predicts a possible lag of about 2 Ma between the first uplift above the center of the plume head, and uplift on the outer fringes of the expanding asthenosphere (I. Campbell, personal communication, 2005). Such a lag should be  recorded by the offlapping turbidites, and is probably measurable from biostratigraphic dating of the first hemipelagics deposited on the last turbidites. The Campbell (2006) model also predicts a lag of  ca 2-4 Ma between uplift initiation and onset of volcanism. Current dating (Middle Eocene for onset of uplift; late Middle to early Late Eocene or earlier for the volcanism) make such a lag possible, but not proven. The plume model also predicts a possibly testable early central uplift several hundred meters higher in the beginning, before spreading below the plate flattens the head and reduces swell height during the next ca. 2 Ma. The "swell root spreading" models of Phipps Morgan et al. (1995; their Fig.6 ) make even more specific and testable predictions about rise uplift and swell radius as a function of time.

Uplift resolution might be further refined by correlation of individual turbidites from one borehole to the next. Some of the thicker and compositionally distinctive Quaternary turbidites have been correlated from one core to another in modern abyssal plains surrounding Bermuda (Pilkey & Cleary, 1986). Readers interested in joining the authors in proposing the Bermuda Rise borehole transects discussed above are welcome. Please email us about your interest!

Acknowledgments

We thank Marie-Claude Williamson for sharing results from her reanalysis of Bermuda Deep Drill-72 igneous rock samples from Bermuda. More complete acknowledgments are given in Vogt & Jung (2007). 

References

• Anderson, D.L., Scoring hotspots: The plume and plate paradigms, in Foulger, G.R., Natland, J.H., Presnall, D.C., and Anderson, D.L., eds., Plates, plumes, and paradigms, Geol. Soc. Amer. Special Paper 388, 31-54, 2005.
• Aumento, F., and B.M. Gunn, Geochemistry of Igneous Rocks, See Oceanic Islands Basalts/Atlantic/Bermuda, 2006. 
• Aumento, F., P.H. Reynolds, and B.M. Gunn, The Bermuda Seamount; a reactivated section of an older oceanic crust (Abs.), Eos Trans. Amer. Geophys. Union, 55, 455, 1974.
• Ayer, E.A., and E.P. Laine, Seismic stratigraphy of the northern Bermuda Rise, Mar. Geol., 49, 169-186, 1982.
• Campbell, I.H., Large Igneous Provinces and the mantle plume hypothesis, Elements, 1, 265-269, 2006.
• Courtillot, V., A. Davaille, J. Besse, and J. Stock, Three distinct types of hotspots in the Earth's mantle, Earth Planet. Sci. Lett., 205, 295-308, 2003.
• Cox, R.T., and R.B. Van Arsdale, The Mississippi Embayment, North America; a first order continental structure generated by the Cretaceous superplume mantle event, J. Geodynamics, 34, 163-176, 2002.
• Crough, S.T., Thermal origin of mid-plate hotspot swells, Geophys. J. R. Astron. Soc., 55, 451-470, 1978.
• DeLaughter, J.E., C.A. Stein, and S. Stein, Hotspots: A view from the swells, in Foulger, G.R., Natland, J.H., Presnall, D.C., and Anderson, D.L., eds., Plates, Plumes and Paradigms, Geol. Soc. Amer. Spec. Paper 388, 257-278, 2005, doi:10.1130/2005.2388 (16).
• Detrick, R.S., P.R. Von Herzen, B. Parsons, D. Sandwell, and M. Dougherty, Heat flow observations on the Bermuda Rise, J. Geophys. Res., 91, 3701-3723, 1986.
• Duncan, R.A., Age progressive volcanism in the New England Seamounts and the opening of the central Atlantic Ocean, J. Geophys. Res., 89, 9980-9990, 1984.
• Fukao, Y., Widiyantoro, S., and Obayashi, M., Stagnant slabs in the upper and lower mantle transition region, Rev. Geophys., 39, 291-323, 2001.
• Gees, R.A., The age of Bermuda Seamount, Maritime Sediments, 5, 56-57, 1969.
• Griffiths, R.W., and I. Campbell, Interaction of mantle plume heads with the Earth’s surface and onset of small-scale convection, J. Geophys. Res., 96, 18295-18310, 1991.
• Gurriet, P., A thermal model for the origin of post-erosional alkalic lava, Hawaii, Earth Planet. Sci. Lett., 82, 153-158, 1987.
• Harris, R.N., R.P. Von Herzen, M.K. McNutt, G. Garven, and K. Jordahl, Submarine hydrology of the Hawaiian archipelagic apron, Part 1: Heat flow patterns north of Oahu and Maro Reef, J. Geophys. Res., 105, 21353-21370, 2000.
• Holm, P.M., J.R. Wilson, B.P. Christensen, L. Hansen, S.L. Hansen, K.M. Hein, A.K. Mortensen, R. Pedersen, S. Plesner, and M.K. Runge, Sampling the Cape Verde mantle plume: Evolution of melt compositions on Santa Antao, Cape Verde Islands, J. Pet., 47, 145-189, 2006.
• Hyndman, R.D., G.K. Muecke, and F. Aumento, Deep-Drill-1972, Heat flow and heat production in Bermuda, Can. J. Earth Sci., 11, 809-818, 1974.
• Jaroslow, G.E., and B.E. Tucholke, Mesozoic-Cenozoic sedimentation in the Kane Fracture Zone, western North Atlantic, and uplift history of the Bermuda Rise, Geol. Soc. Amer. Bull., 106, 319-337, 1994.
• King, S.D., and D.L. Anderson, Edge-driven convection, Earth Planet. Sci. Lett., 160, 289-296, 1998.
• Klitgord, K.D., and  H. Schouten, Plate kinematics of the central Atlantic, in Vogt, P.R., and B.E. Tucholke, eds., The Geology of North America, v. M, The Western Atlantic Region, Geological Society of America, Boulder, 351-378, 1986.
• Liu, M., and C.G. Chase, Evolution of midplate hotspot swells; numerical solutions, J. Geophys. Res., 94, 5571-5584, 1989.
• Lizarralde, D., J.B. Gaherty, J.A. Collins, G. Hirth, and S. Kim, Spreading-rate dependence of melt extraction at mid-ocean ridges from mantle seismic refraction data, Nature, 432, 744-747, 2004.
• Louden, K.E., D.O. Wallace, and R.C. Courtney, Heat flow vs age for the Mesozoic Northwest Atlantic Ocean: Results from the Sohm abyssal plain and implications for the Bermuda Rise, Earth Planet. Sci. Lett., 83, 109-122, 1987.
• McHone, J.G., Constraints on the mantle plume model for Mesozoic alkaline intrusions in northeastern North America, Can. Mineralogist, 34, 325-334, 1996.
• McNutt, M.K., Heat-flow variations over the Hawaiian Swell controlled by near-surface processes, not plume properties, in Hawaiian Volcanoes: Deep Underwater Perspectives, Geophys. Monograph 128, American Geophysical Union, Washington, DC, 355-364, 2002.
• Morgan, W.J., and S.T. Crough, Bermuda hotspot and the Cape Fear Arch, Eos Trans. Amer. Geophy. Union, 60, 392-393, 1979.
• Morgan, W.J., Hotspot tracks and the early rifting of the Atlantic, Tectonophysics, 94, 123-139,1983.
• Mueller, R.D., J-Y. Royer, and L.A. Lawver, Revised plate motions relative to hotspots from combined Atlantic and Indian Ocean hotspot tracks, Geology, 21, 275-278, 1993.
• Olsen, S.D., Petrogenesis of Bermuda's igneous basement (M.S. Thesis), Univ of Copenhagen, 254 pp., 2005.
• Phipps Morgan, J., W.J. Morgan, and E. Price, Hotspot melting generates both hotspot volcanism and a hotspot swell?  J. Geophys. Res., 100, 8045-8062, 1995.
• Pilkey, O.H., and W.J. Cleary, Turbidite sedimentation in the northwestern Atlantic Ocean basin, in Vogt, P.R. and B.E. Tucholke, eds., Turbidite sedimentation in the northwestern Atlantic Ocean basin, v. M, The Western North Atlantic Region,  Geology of North America, Geol. Soc. Amer., Boulder, 437-450, 1986.
• Reynolds, P.R., and F.A. Aumento, Deep Drill 1972: Potassium-argon dating of the Bermuda drill core, Can. J. Earth Sci., 11, 1269-1273, 1974.
• Ritsema, J., Global seismic structure maps, in Foulger, G.R., Natland, J.H., Presnall, D.C., and Anderson, D.L., eds., Plates, plumes, and paradigms, Geol. Soc. Amer. Special Paper 388, 11-18, 2005.
• Rona, P.R., and E.S. Richardson, Early Cenozoic global reorganization, Earth Planet. Sci. Lett., 40, 1-11, 1978.
• Rowe, M.P., 1998, An explanation of the Geology of Bermuda, Bermuda Government, Ministry of the Environment, 30 pp.
• Sclater, J.G., and L. Wixon, The relationship between depth and age and heat flow and age in the western North Atlantic, in Vogt, P.R., and B.E. Tucholke, eds. The western North Atlantic region, v. M, The Geology of North America, Geol. Soc. Amer., Boulder, Colorado, 257-270, 1986.
• Sharp, W.D. and D.A. Clague, 50-Ma Initiation of Hawaiian-Emperor Bend Records Major Change in Pacific Plate Motion, Science, 313, 1281-1284, 2006.
• Sheehan, A.F., and M.K. McNutt, Constraints on thermal and mechanical structure of the oceanic lithosphere at the Bermuda Rise from geoid height and depth anomalies, Earth Planet. Sci. Lett., 93, 377-391, 1989.
• The Shipboard Scientific Party, Site 386: Fracture valley sedimentation on the central Bermuda Rise, in Tucholke, B.E., and P.R. Vogt, eds., Initial Reports of the Deep Sea Drilling Project, v. 43, U.S. Gov't. Printing Office, Washington, D.C., 195-321, 1979.
• Smith, H.F., and D.T. Sandwell, Global seafloor topography from satellite altimetry and ship depth soundings, Science, 277, 1956-1962, 1997.
• Tucholke, B.E., and G.S. Mountain, Seismic stratigraphy, lithostratigraphy, and paleosedimentation patterns in the North American Basin, in Talwani, M, W. Hay, and W.B.F. Ryan, eds., Deep Drilling Results in the Atlantic Ocean: Continental Margins and Paleoenvironment, Maurice Ewing Ser., v.  3, Amer. Geophys. Union, Washington, DC, 58-86, 1979.
• Tucholke, B.E., and P.R. Vogt, Western North Atlantic: Sedimentary evolution and aspects of tectonic history, in Tucholke, B.E. and P.R. Vogt, eds., Initial Reports of the Deep Sea Drilling Project, v. 43, U.S. Gov't. Printing Office, Washington, D.C., 791-825, 1979.
• Vacher, H.L., Rowe, M.P., and Garrett, P., 1989, The Geological Map of Bermuda, Scale 1:25,000, Oxford Cartographers, London, Bermuda Government, Ministry of Works and Engineering.
• Vogt, P.R., Bermuda and Appalachian-Labrador rises; common non-hotspot processes?, Geology, 19, 41-44, 1991.
• Vogt, P.R., and G.L. Johnson, Cretaceous sea-floor spreading in the western North Atlantic, Nature, 234, 22-25, 1971.
• Vogt, P.R., and W.-Y. Jung, Origin of the Bermuda volcanoes and Bermuda Rise: History, Observations, Models, and Puzzles, in Foulger, G., and Jurdy, D., eds., Plates, Plumes, and Planetary Processes, Geological Society of America Special Paper 430, 2007 (in press).
• Von Herzen, R.P., Geothermal evidence for continuing hydrothermal circulation in older (>60 my) ocean crust, in Davis, E.E., and E.H. Elderfield, eds., Hydrogeology of the oceanic lithosphere, Cambridge Univ. Press, Cambridge, 414-447, 2004.
• Williamson, M.-C., M. Villeneuve, and S. Blasco, Deep Drill 1972 revisited: New data on the composition and age of the Bermuda volcanic edifice (Abs.), Eos Trans. Amer. Geophys. Union, 2006.
• Zhao, D., Global tomographic images of mantle plumes and subducting slabs: insight into deep Earth dynamics, Phys. Earth and Planet. Int., 146, 3-34, 2004.
• Zhu, A., and D.A. Wiens, Thermoelastic stress in oceanic lithosphere due to hotspot reheating, J. Geophys. Res., 96, 18323-18334, 1991.