Conclusions

The thesis is nearly done, here are the conclusions of three and a half years work:

1. The major and rare-earth element composition of melt generated during the extension and rifting of the continents has been successfully predicted within a geodynamic model of rifting. The melt composition calculations are modular, and can be incorporated within most models of rifting where the fraction of melt generated is calculated.

2. It has been shown that the strengthening of the mantle when small fractions of melt are generated is of fundamental importance to the evolution of rifted margins. It acts to dampen small-scale convection and reduce the volume of melt generated at slow spreading rates and mantle temperatures greater than 1300 C. I have shown that to match the global igneous crustal thickness variations, the mantle must be between 1300 and 1325 C, which is in agreement with the latest estimates from heat flux calculations (McKenzie et al, 2005).

3. Around the coast of the North Atlantic there are regions of very thickened igneous crust known as the North Atlantic Igneous Province. At the Southeast Greenland margin, the igneous thickness is up to 18 km (Hopper et al, 2003). The melt generated at the time of rifting is picritic with very high concentrations, 18 %, of MgO (Thy et al, 1998) and depleted in rare-earth elements such as TiO and La. To reproduce such volumes of melt with this chemical signature requires three processes:

Firstly, the presence of a layer of hotter mantle beneath the lithosphere. This layer is ~ 200 C hotter than the mantle, which has a potential temperature of 1325 C. Such a thermal anomaly is consistent with the lateral movement of a thermal plume based under Iceland (Holbrook et al, 2001; Chapter 4).

Secondly, rifting occurred initially at a faster rate of spreading, ~ 40mm/yr (Chapter 4). This is in agreement with observational evidence for increased spreading rates off shore Southeast Greenland (Larsen and Saunders, 1998; Smallwood and White, 2002; Hopper et al, 2003). Such increased spreading rates are essential as it increases the amount of mantle welling up and so increasing the flax of material through the solidus.

Finally, extension events prior to the rifting are crucial to the generation of excessive melt upon breakup. Passive far field extension of 125 km thick lithosphere above a thermal anomaly, even at the faster spreading rates of ~ 40 mm/yr, does not produce sufficient melt to match the 18 km thick crust observed off Southeast Greenland. The combined effect of the extension that gave the Hatton-Rockall Basin and a prior extensional basin off Southeast Greenland provides enough extension to thin the lithosphere sufficiently to enhance the upwelling of material to generate the volumes and chemistry of the igneous material observed at the Southeast Greenland margin (Chapter 5).

4. Melting at the Southeast Greenland margin began at depths close to 150 km. This deep melting was slightly damp, yet is within the garnet peridotite stability field as the the temperatures are too hot for amphibole peridotites. Given the deep damp melting, only a slight reduction in the Dy/Yb rare-earth ratio, which is a proxy for the amount of melt generated in the garnet stability field, is observed as rifting becomes sea floor spreading at the Southeast Greenland margin (Chapter 6).

5. The variation of seismic velocity within the underplate off Southeast Greenland with the igneous thickness is reproduced by the breakup scenario outlined above. The reduction in observed seismic velocities corresponds with the reduction in spreading rates and exhaustion of the thermal anomaly (Chapter 6). The geodynamic model of the extension and rifting of the continents has proven to be a successful tool for understanding the processes at the Southeast Greenland margin.