Australia: The Land Where Time Began

A biography of the Australian continent 

Oceanic Crustal Thickness Decrease After Breakup of Pangaea

Since the Archaean 2.5 Ga the mantle of the Earth has cooled by 6-11^oC every 100 My. The loss of surface heat that led to this drop in temperature may have been enhanced by the processes of plate tectonics, such as the breakup of continents, the continuous formation of oceanic lithosphere at the mid-ocean ridges and subduction at deep-sea trenches. In this study Van Avendonk et al. used a compilation of marine seismic refraction data gathered from ocean basins globally to analyse thickness changes of ocean crust over time. They found that oceanic crust that formed about 170 Ma in the Middle Jurassic is on average 1.7 km thicker than crust that is being produced at the present at mid-ocean ridge systems. If the thicker oceanic crust from the Jurassic is caused by higher mantle temperature, the upper mantle may have cooled by 15-20^oC per 100 My over this period of time. It is suggested by the difference between this and the cooling rate over the long term that modern plate tectonics coincide with grater heat loss from the mantle. Van Avendonk et al. also found that the oceanic crustal thickness increase with the age of a plate is stronger in the Atlantic Ocean and the Indian Ocean compared with the Pacific Ocean. The idea that the temperature of the upper mantle in the Jurassic was higher in the wake of the fragmentation of the supercontinent Pangaea as an effect of continental insulation is supported by this observation.

It is suggested by the geochemistry of ancient basalts that the long term cooling rate of the mantle is no more than about11^oC every 100 My (Herzberg, Condie & Korenaga, 2010); Condie, Aster & van Hunen, 2016). The role of the continents is a factor in the evolution of the loss of heat of the Earth over time. The mantle beneath large continents can be insulated for some time, which would lead to increased mantle temperatures (Gurnis, 1988; Whittaker et al., 2008). If there is evidence that formation and dispersal of Pangaea had an effect on the mantle beneath it more insight would be gained into the importance of these plate tectonic events. Van Avendonk et al. used measurements of oceanic crustal thickness as a proxy for the temperature of the mantle to test these ideas because the thicker oceanic crust is thought to form when temperature of the mantle is higher. In particular, about 60 % of the crust of the Earth has formed by decompression melting of mantle upwelling that is relatively hot beneath the mid-ocean ridges (Müller, Sdrolias, Gaina & Roest, 2008). It has been shown by numerical models of seafloor spreading (Bown & White, 1994) there will be a decrease in the amount of melting of the mantle when the full rate of spreading is less than 20 mm per year, and the a potential temperature of the mantle will lead to a higher degree of partial melting. Variations in the thickness of young oceanic crust are constrained by marine refraction data to between 3 km and 8 km, which is consistent with the numerical models (White et al., 2001). Some evidence for a small increase in the thickness of the oceanic crust with age is provided by seismic data (McClain & Atallah, 1986), though historically there have been few measurements of older oceanic crust. Seismic refraction studies of subduction zones and the deep seafloor adjacent to passive margins have provided better global coverage with estimates of the thickness of oceanic crust of various ages. Insight has been provided by these additional data into the processes of seafloor spreading in the past, assuming there has not been significant overprinting by later geologic events.

Van Avendonk et al. have shown the geographical distribution of thickness measurements of oceanic crust based on seismic refraction data gathered over the last 40 years. The relationship between age and thickness of oceanic crust in the Pacific, Atlantic and Indian oceans can be evaluated because of these seismic constraints on seafloor dating from170 Ma to recent (0.3 Ma). All the measurements were made on seafloor that formed at a full spreading rate of 30 mm per year or more, as crust that is generated at slower rates can be anomalously  thin (White et al., 2001). The list of 234 estimates of crustal thickness and their literature sources are listed in the supplementary Table 1. A large amount of scatter with a small positive correlation is shown by the distribution of seismic crustal thickness data as a function of age. Oceanic crust has been shown by a least squares regression to have decreased over time by 10 ± 1 m per My on average, which totals 23 % since 170 Ma. The linear least-squares fit of the crustal thickness measurements has an x² statistic of 1.9, where a data fit that is comparable to the uncertainties would be x²=1.0. That there are also spatial variations in the thickness of the oceanic crust at any given age, which are not modelled in this study, is shown by the additional scatter in the data. It is indicated by the linear regression that has a coefficient of determination R² of 0.28 that its variance is clearly smaller than that of the crustal thickness that is best fitting, though this improvement is modest.

Implications for global tectonics

Deep mantle overturning and cooling may have been enhanced by the partial insulation of the mantle of the Earth by large continents (Lenardic, Moresi, Jellinek & Manga, 2005), leading to a large decrease in the temperature of the mantle over the last several hundred million years. Asthenospheric mantle that had been thermally insulated beneath Pangaea (Gurnis, 1988) was emplaced in the young Atlantic and Indian Ocean, as Pangaea was breaking up in the early Jurassic. Thicker oceanic crust than that formed at the mid-ocean ridges bordering the Pacific Plate was initially generated by the hot upper mantle. Mantle overturn is slow beneath the Atlantic and Indian oceans compared with that beneath the Pacific Ocean where it is more vigorous, because the lithosphere beneath the Atlantic and Indian oceans is fixed to tectonic plates that have deep continental keels. A result of this is that the thermal state of these 2 areas may have evolved differently over time. It appears that young oceanic crust, less than 1 My old, is on average 1.3 ± 0.2  km thicker in the Pacific Ocean than in the Atlantic and Indian oceans, even though measurements of crust that formed at spreading rates of less than 20 mm per year were excluded from this study. Van Avendonk et al. suggest that this discrepancy is explained best by differences in average upper mantle temperature that are basin wide.

Van Avendonk et al. inferred a decrease rate of oceanic crustal thickening over time in the Pacific Ocean  of 8 ± 1 m per My is slightly larger than an earlier estimate of 7 m per My that had been based on fewer seismic refraction data (McClain & Atallah, 1986). The oldest west Pacific Ocean oceanic crust dating to the mid-Jurassic (170 Ma) is on average 1.4 km thicker than the crust that is newly produced at the East Pacific Rise. 170 Ma the mantle would possibly have been between 18^oC and 30^oC hotter in the Jurassic, which is a temperature anomaly that is smaller than the 50-60^oC that has been predicted from the geochemistry of the oldest Pacific basalts (Humler, Langmuir & Daux, 1999; Fisk & Kelly, 2002). Van Avendonk et al. suggest regional anomalies in the temperature of the mantle beneath the mid-ocean ridge system (Gale, Langmuir & Dalton, 2014) may account for this mismatch, as there are not many sites from the Ocean Drilling Program on old seafloor where it would be possible to make such a comparison. Enough buoyancy is provided by the estimated thickening of the Pacific oceanic crust with age to explain roughly 300 m of the flattening that has been observed of seafloor depth as a function of age (Parsons & Sclater, 1977), an amount that is comparable to the depth of the seafloor misfit of current plate cooling models for the Pacific Ocean (Hillier, 2010). Long-term changes in sea level may have been influenced by the difference in isostasy of the oceanic plate over time (Miller et al., 2005).

The mantle of the incipient Atlantic Ocean offshore from the US has recently been reported to have been 150^oC higher than beneath modern mid-ocean ridges (Brandt, Regelous, Beier & Haase, 2013) which would be consistent with the crust being much thicker than what has been observed in the seismic data. In the western central Atlantic Ocean the seafloor dating to the Jurassic is locally smooth (Whittaker et al., 2008; Minshull, 1999), which Van Avendonk et al. suggest lends proof that at the time of its formation the mantle beneath it was unusually hot 30 My after the continents of North America and Africa separated. A seismic refraction study has shown that in this area the oceanic crust is locally 8 km thick (Minshull et al., 1991), which is more easily explained by a past mantle potential temperature approximately 50^oC higher than that of the modern Mid-Atlantic Ridge system. There is a large amount of scatter arising from lateral heterogeneity in the temperature and composition of the mantle that can be seen in the long-term signal of cooling of the mantle in samples of oceanic basalt or data from marine seismic refraction data (Gale, Langmuir & Dalton, 2014; Niu & O’Hara, 2008).

It has been debated whether there had been an arrival of a deep mantle plume offshore from the southeast US about 200 Ma that accompanied the rifting and breakup of Pangaea (Wilson, 1997; McHone, 2000; Janney & Castillo, 2001). During early seafloor spreading at that location the elevated mantle temperature may be consistent with a plume, though the strong positive correlation between the thickness of the oceanic crust and the age of the tectonic plate applies more generally to the Atlantic and Indian oceans. Prior to breakup of the supercontinent the continental insulation of the mantle beneath central Pangaea warmed the asthenosphere (Humler & Besse, 2002), which contributed to the extensive syn-rift magmatism between the eastern US and Africa (Holbrook& Kelemen, 1993), leading to the formation of Jurassic oceanic crust that is on average thicker than the crust that is generated along the Mid-Atlantic Ridge of the present.

Sources & Further reading

1. Van Avendonk, H. J. A., J. K. Davis, J. L. Harding and L. A. Lawver (2017). "Decrease in oceanic crustal thickness since the breakup of Pangaea." Nature Geosci 10(1): 58-61.