On alternative models for the origin of time-progressive volcanic chains

1. On alternative models for the origin of time-progressive volcanic chains V. Puchkov, Institute of Geology, Ufimian Scientific Centre, Russia, puchkv@anrb.ru

2. The time-progressive character of some volcanic chains is well known for decades and is proven with a different degree of reliability for many of them. The recent overviews of the primary data can be found in the papers of O'Neil et al. 2005, Clouard & Bonneville 2005, enhanced by additional information from many papers and books (all references are in the poster). The above data were used by the author to compile a world scheme of time-progressive chains. The existence of such chains is a real challenge for anyone who works at the problem of origin of melting anomalies.

3. Alternative modelsfor TPVC (time-progressive volcanic chains)

5. 1. Plume model. • A standard plume and plate concept of origin of TPVC at the example of the Emperor-Hawaii ridge (Norton,2000)

6. The model was first suggested by T.Wilson (1963) as a hypothesis of “hot spots”- surficial manifestations of immovable deep mantle melting anomalies, the idea eveloped later by W. Morgan (1971)as a theory of plumes:convective upwellings of a light hot mantle substance coming from the core-mantle boundary. It was a smart explanation, and still is, though the idea of an absolute horizontal immobility of plumes is disproved now by paleomagnetic and geodetic data (Antretter et al., 2002; Norton, 2000; O’Neill et al., 2005 & oth. ). The above-presented map is in a good accordance with this concept.

7. (Mazarovich, 2000) Galapagos (O’Connor et al.) • Some TPVCs demonstrate more complicated progression of ages compared to Hawaii (e.g. Galapagos, Canaries), where volcanoes, once lit, are slow to be extinct. But the model can be the same with an admission that a plume is not necessarily head-and-tail or pillar-like.

8. The ITRF (International Terrestrial Reference Frame), 2005, http://itrf.ensg.ign.fr/, showing vectors of the modern plate movements, is also in a good conformity with the above model, though the direction of plate movements changed in time, as it is demonstrated in the left cutoff.

9. Note specially: i. the normal character of TPVCs at the East Pacific; ii. the oblique character of the chains relatively to the MOR in the Southern Atlantic, while they are subparallel to the vectors; iii. Cobb, Bowie and Yellostone chains are oriented the same way as the vectors, notwithstanding the fact that they are situated at the different sides of the plate boundary; iv. Reunion and Kerguelen chains cross the MOR, because the plates at both sides of the MOR drift in the same direction.

10. It also cannot be a coincidence that the young (active or recently extinct volcanoes at the ends of the chains sit over superswells.

11. Superswells and LIPS [Burke,Torsvik, 2004] Based on paleomag data, these authors have shown that positions of LIPs (Large Igneous Provinces) reconstructed to the time of their eruptions are situated within or at the edges of two superswells at CMB, with minor exceptions. The background map shows the SMEAN shear-wave tomography model [Becker,Boschi 2002], which is based on an average of three global intermediate wavelength shear-wave tomography models. Note that δVs color contours are highlighting the red (slower speed) and blue (higher speed) regions. We have shown in the previous slide, based on a different approach, that active or shortly extinct volcanoes at the ends of time-progressive chains follow the same pattern

12. The later scheme of Burke et al (2008) shows also some hot spots (pink crosses). It manages better with the “exceptions”, connected with Columbia River anomaly, but still has difficulties with Afar hotspot. I shall try to explain it later.

13. 2. Model of a Propagating crack of lithosphere connected with cooling stress • These forces, presumably connected with cooling stress,do not exist virtually (by themselves). They are added vectorially to much greater forces driving lithospheric plates (slab pull, mantle drag and ridge push). The resulting forces cannot produce such effect. • Therefore the model contradicts the plate tectonics (Stuart, Foulger, Barall, 2007)

14. A generalized world stress map based on the research in the frame of LITHOSPHERE Program (Zoback et al., 1992) • 1- tension, 2 – compression with thrusting, 3 – compression with formation of diagonal wrench faults, 4 – plate boundaries, 5- trajectories of plate movements

15. Lithgow-Bertelloni and Guynn,2004 • Note a scale (greater by an order) and stresses parallel to Hawaian TPVC

16. Another objections. 1. The propagation of time-progressive chains in all oceans is organized as predicted by plume-and-plate tectonics. If so, why do we need more explanations? Occam's razor: “entia non sunt multiplicanda praeter necessitatem", (”one should not increase, beyond what is necessary, the number of entities required to explain anything”). 2. Why the crack propagation is not affected by the strongest anisotropy of oceanic, transitional and continental lithosphere, crossed by many chains; 3. Why the Reunion, Kerguelen and New England-Great Meteor TPVCs overrode the active MORs and after that co-existed with them for some time; 4. Why LIPs and TPVCs are correlated with superswells, as might be predicted by the plume theory; 5. Why the “cracks” tapped fertile sources, producing OIBs, while the MORs - depleted ones?

17. To the left: a succession of a spreading onset in the Atlantic (Khain,2003) However there are TPVCs called forth by plate tectonics • The propagating crack idea by itself is useful in explanation of TPVCs connected with time-progressive graben formation. In some cases they preceded splitting apart of the supercontiment and ocean floor spreading, in some they seem to precede the future split-up of a lesser continent. • In the TPVC world map presented in the next slide such lineaments are shown tentatively in blue colour To the right: stages of development of a prograding rift (Martin, 1983)

19. Filho et al.

20. 3.The model of a drifting fertile “blob” Another alternative model is that of easily melting magma sources (pyroxenite “blobs”), drifting in an asthenosphere in the same direction as the overlying lithosphere, but quicker (Anderson, 2007).

21. The model, worked out in detail (Cuffaro, Doglioni, 2007 ), suits ad hoc the Pacific chains (except Easter and Galapagos where the orientation is opposite to predicted), but at a global scale it fails because it contradicts also to the behavior of time-progressive chains in the Eastern Atlantic and Indian oceans. • A comparison of the map of time-progressive volcanic chains,presented here , with the schemes explaining the model of Cuffaro and Doglioni (2007) shows contradictions (the next slide). The idea of a uniformly eastward-flowing mantle, with asthenosphere and lithosphere uniformly lagging behind, predicts the time succession directions for the chains of the SE Pacific, Indian and East Atlantic oceans as contrary to what is observed.

22. Puchkov (2008) Cuffaro, Doglioni Suggested vectors of plate movements Suggested mantle flow

23. The very idea of a pyroxenite “blob” (basic in composition) hanging in asthenosphere for many tens of Ma, conflicts with the nature of asthenosphere, the latter demonstrating readily the effects of Archimedes law. When situated deeper than 50 km, the “blob” exists as an eclogite which is much denser that the ambient peridotite and must sink, if not supported by a plume upwelling or heated by it. When situated higher than the phase transition zone, it turns into gabbro which is much lighter than the peridotite and therefore must finally strike the bottom of lithosphere. Nevertheless, the idea of pyroxenite as an easily melting part of the Earth's mantle seems, by itself, to be very promising. It really can be of a great benefit for the plate tectonic approach in explanation of shallow, top-asthenospheric decompression-induced melting anomalies (Anderson, 2007, Foulger, 2007). But it can be also extremely useful for the real plume model as it became evident recently from an example of the deep-sourced Hawaiian magmas (Sobolev et al., 2005; Yaxley, Sobolev 2007).

24. Conclusions. • The plume model fits in the best way the features of the chosen time-progressive volcanic chains: coincidence of their orientation and time successions with directions of plate movements, independence of the chains on “shallow” structures and processes in lithosphere and asthenosphere, connection of the young ends of chains with superswells at CMB. • On the other hand, the alternative hypotheses lead to useful ideas of a real role of crack propagation in time-progressive volcanism, importance of mantle-hosted eclogite-pyroxenite melting anomalies, and possibility of passive, purely plate tectonic rifting mechanism leading to a shallow decompressional hotspot melting. • Probably it is a way to a conciliation of competing theories.