where ^ is the viscosity of the mantle, p is the density of the mantle, g is the acceleration of gravity, R is the radius of the depressed area, and T is the time required to approach equilibrium. In this way, it has been shown that the viscosity of the upper mantle is between 1019 and 1020 Pas (pascal-seconds). By comparison, the viscosity of water is about 10-3 Pas, and maple syrup is 2.5 Pas. Glass is also a liquid, no matter how hard and brittle it may seem on the short timescales of days, weeks, and years: The molecules that make up glass are not ordered, so the material is technically a liquid, and does in fact flow very slowly, over great amounts of time. Glass has a viscosity of about 1019 Pas. If the mantle were as cool as your window glass, it would be several orders of magnitude more viscous yet.
Much about convective motions in the Earth's mantle is unknown. The movement of the plates proves that the mantle in convecting at shallow depths, but the patterns of convection at depth are still a matter of debate. For years many scientists thought that the Earth had layered convection: The upper mantle convected in its own pattern, and beneath the 670 discontinuity, the lower mantle convected separately. This would make separate compositional reservoirs possible, since each convecting layer would eventually become well mixed, but the layers might remain separate.
In the last few years, a technique called tomography has been developed. Using seismic waves coming from earthquakes in several directions, material in the mantle that changes their speed can be pinpointed, and even outlined. Rob van der Hilst, a geophysicist at the Massachusetts Institute of Technology, and his colleagues have used tomographic techniques to locate and outline subducting slabs of oceanic crust as they sink in the mantle. The slabs remain cold for a long time as they sink, and their coldness slows down seismic waves that move through them, making it possible to create an image of the slab in the mantle.These researchers, and others, have found evidence that subducting slabs sometimes settle to rest on the 670 discontinuity, at least temporarily, but others plunge straight through that boundary and sink to the deepest mantle.
When these studies were published, they created shockwaves in the geologic community. Here was direct evidence for two things that geologists had speculated about for years: The first is whole-mantle convection, supported by the slabs' movement through the entire mantle, and the second is mantle heterogeneity. Using geochemical studies of volcanic rocks, it has been postulated for years that there are regions in the mantle that have distinct compositions. These distinct compositions are reflected in the compositions of the volcanic rocks that are produced from them.This was sometimes taken as evidence for layered mantle convection. Here, though, is a way to make the mantle compositionally heterogeneous on a small scale:The subducted plates have a very different composition than the mantle as a whole, and they could mix into the mantle over time, creating ribbons of varying composition in the mantle. Subducted slabs also carry sediment from the sea floor, and the uppermost slab and sediment are loaded with water. Subducting slabs are thus a way that water can be added to the mantle as well.
Was this article helpful?