1. Lecture 2b: Hot spots Questions –Why are there volcanoes in the middle of plates? –How do such volcanoes grow and evolve? –What is the connection between hotspots and flood basalts? Tools –Plate tectonics, geochronology, igneous petrology, isotope geochemistry, etc. 36

2. Hot spot chains 37

3. Hot spots, flood basalts, LIPs –Hawaii and Iceland are biggest, by buoyancy flux and by volume of volcanism. Chains of volcanoes in the middle of plates, if long-lived and with an age progression along the chain, are called hot spots. –There are continual arguments over how many hot spots there are – some are obvious, some are marginal (usually because it is hard to establish the age progression). The most common catalogue has ~40, others go over 100. 38

4. Hot spots: correlated with geoid? 39

5. Hot spots: correlated with geoid? 40 Burke et al, 2008 (10.1016/j.epsl.2007.09.042)

6. Hot spots: correlated with geoid? 41 Burke et al, 2008 (10.1016/j.epsl.2007.09.042)

7. Mantle Plumes An experimental starting plume (in glucose syrup) The initiation of a new plume is thought to involve a very large blob of hot material arriving at the base of the lithosphere and hence a large episode of excess volcanism. This is supported by the association between many active hotspots and older continental flood basalts or oceanic plateaux (collectively, large igneous provinces or LIPs). 42

8. Flood Basalts and hotspot tracks 43

9. Flood Basalts and hotspot tracks 120 Ma ago Today 44

10. Flood Basalts and hotspot tracks 45 Burke et al, 2008 (10.1016/j.epsl.2007.09.042)

11. Flood Basalts Flood basalts are big and erupt very quickly: – Siberian traps, 2 x 10 6 km 3 within ~1 Ma at 250 Ma (Permian-Triassic) – Deccan traps (India), 10 6 km 3 within ~1 Ma at 65 Ma (K-T) – Columbia River Basalts, 2 x 10 5 km 3 within ~1 Ma at 16 Ma. –It may or not be coincidence that big flood basalt eruptions coincide with major extinction events in the fossil record! Flood basalt petrology and chemistry: there is a general progression through… –small volume of early alkali basalt and olivine tholeiite with mantle isotope signatures and high 3 He/ 4 He – massive volume of quartz tholeiite with isotopic signature of subcontinental lithosphere –small late eruptions with wide range of compositions and evidence of crustal components 46

12. Flood Basalts We can think of this sequence as the result of emplacement of a big thermal anomaly at the base of the continental lithosphere. –The first products are small degree, high-pressure melts of the plume itself, that escape quickly. –Then heat flow, a slow process, raises the temperature of the cold non- convecting part of the mantle attached to the base of the continent until it melts over a wide area, in a process that is characterized by positive feedback between melting and heat flow, giving high magma flux for a short time. –Finally, heat from the plume head reaches the crust itself, mostly by advection of magmas, and some crustal melts occur. Oceanic plateaux are basically similar to flood basalts, except they presumably occur when a plume head comes up under oceanic lithosphere. –The biggest on earth is the Ontong-Java plateau, which is really two oceanic plateaux on top of each other, one 122 Ma and one 90 Ma. At that time the Pacific plate was hardly moving, and two plume-head like blobs came up the same conduit and hit the same area of lithosphere. 47

13. Ocean Island Volcanoes Mid-plate volcanoes in age-progressive chains are presumably the product of long-lived plume tails. They show a sequence driven by motion of new lithosphere over the plume (rather than arrival of new plume under lithosphere). At least at Hawaii, the lifecycle of one volcano is typically divided into four stages: Preshield – low flux of alkali basalt, erupted submarine, very pure plume component (high 3 He/ 4 He, etc.) Shield-building stage – very large flux and large volume of tholeiite, progressing towards an upper mantle/oceanic lithosphere affinity. Alkalic capping stage – small flux of alkali basalts, no steady shallow magma chamber. Posterosional stage – very small volume of extremely alkalic lavas that erupt ~2 Ma after end of capping stage. 48

14. Ocean Island Volcanoes Subaerial tholeiite flows have low viscosity and long cooling times and can travel far down low slopes, allowing the volcano to assume the characteristic shield shape, perhaps 50 km in diameter for each 1 km above sea level at the summit. In addition to characteristic chemistry, these stages generate characteristic morphology and structures: The pre-shield stage, erupted underwater, is mostly a big mound of pillow basalts, relatively steep sided. At present, this stage is only known from Loihi seamount; its role in other volcanoes is inferred. The main shield stage creates an edifice that emerges above sea-level. 49

15. Ocean Islands: Main Shield Stage A large summit caldera develops when the roof collapses into a shallow (<1 km below summit) magma chamber. Most lavas ascend to this summit magma chamber and degas and differentiate there, even if they erupt down on the… Rift zones that develop when gravitational stresses and push from intruding dikes break the edifice into three (or two if buttressed by an older volcano on one flank) sectors. The ongoing Pu`u-O`o eruption of Kilauea is on Kilaueas southeast rift zone. Occasionally, large sectors of a Hawaiian volcano will fail catastrophically and produce an enormous landslide, with the potential to drive km-high tsunamis. There are frequent earthquakes on shield volcano islands, sometimes on nearly horizontal faults. 50

16. Ocean Islands: Main Shield Stage 51 By way of advertisement, when you get to end of our Ph.D. program, you will have the opportunity to participate in project Pahoehoe and see a shield volcano for yourself. Here is Caltech undergrad Laura Elliott lava-dipping in 2004.

17. Ocean Islands: Hawaii Topography and bathymetry show the shield shapes, summit calderas, rift zones, Loihi seamount, and sector-collapse landslide deposits. 52

18. Ocean Islands: Post-shield stage The alkalic lavas of the post-shield capping stage are smaller in volume and more viscous. They build a steeper mound on top of the shield (see Mauna Kea at present), with many near-summit cones, but no major caldera. These lavas frequently carry xenoliths from the oceanic crust and the cumulate pile inside the volcano, implying that they pond at the base of the crust and do not pause at any shallow magma chamber. 53

19. Ocean Islands: Post-erosional stage The post-erosional lavas are easily recognized as a series of cinder cones and explosive craters on top of a major unconformity and soil horizon from 1-2 Ma of erosion and weathering. See Diamond Head on Oahu. These flows may carry mantle xenoliths, including garnet peridotites, indicating rapid ascent from mantle depths with no permanent magma chamber at any level. 54

20. Oceanic basalts, conclusion In addition to the obvious morphological differences between mid-ocean ridges and ocean island volcanoes, there are important petrological differences relating to degree of melting, volatile content, and extent of melting. Moreover, there are essential, first-order differences in trace-element ratios and radiogenic isotope ratios. –Broadly, MORB is from a depleted and degassed source, presumably the upper mantle –OIB sources tend to be less depleted, nearly primitive, or even enriched relative to bulk earth and show evidence for a primordial noble gas component, hence they are thought to sample the lower mantle in some way. –The existence of distinct isotopic reservoirs in the mantle constitutes the essential geochemical commentary on issues of whole-mantle vs. layered mantle convection, a subject on which geophysicists also have opinions. –We will return to the global geochemical dynamics of the mantle as expressed through oceanic basalts at the very end of the course. 55