The following are some of the alternative mechanisms thought to be responsible for hotspot activity and may be useful to think about when going through the rest of the website. Some models are more controversial than others, but in the end they may all be responsible for some hoptspot regions.
Alexei Ivanov (Ivanov 2005) provides a good overview of this mechanism in the "mantleplumes" website (www.mantleplumes.org)
Reheated slabs suggests a non-convective alternative where cold, subducted slabs reach the core-mantle boundary and are transported in fragments back up through the mantle (this is refered to as penetrative convection) The model is based on observations of some seismic tomography images which do not fit well with the classical plume hypothesis (ie they are too deflected off the vertical even accounting for mantle flow) and argues that some deep negative anomalies can only be slabs as their are sheet-like in shape as opposed to being roughly cylindrical as plumes might be expected.
Geochemically the model still has a long way to go to explain some of the hotspot basalts and how those can be related to an ancient slab source passing through the mantle.
The reheated slabs model
A) Idealised plume convection model
B) Seismic tomography section under Africa (Ritsema, 1999)
C) Reheated slabs model cartoon accounting for hotspot activity under Africa
EDC is a complex process of convection arising from the instability between thick continental crust. The mechanics behind it are too involving to include in this short website and I would like to direct readers to two important articles on EDC written by Don Anderson and Scott King - "Understanding the Edge-Driven Convection Hypothesis" (King, 2004) and "Edge Driven Convection" by (King and Anderson, 1998)
EDC appears to be a fairly robust mechanism for initially driving convection and may quite often be at least partly responsible for some plume trails.
Lithospheric delamination is closely linked to EDC and suggests that the lithosphere can decouple and delaminate in some cases, allowing magma to flow further up than normal and produce melt by decompression melting. This process is summarised by Linda Elkins-Tanton and relies on very particular behaviour of the lithosphere with some gravitational instability.
Impact melting relies on a large body impact producing some decompression in certain regions underneath the imnpact. Jones et al. suggested that an iron body of around 20km travelling at 10km/s would be sufficient to produce a million cubic kilometres of melt following an impact into hot oceanic crust (Jones et al. 2003)
