2.4 Why? Evidence for Decompression Melting
Plate tectonics models of the earth show the mantle convecting, flowing upward beneath divergent plate boundaries and flowing downward at convergent plate boundaries. Some people incorrectly interpret this to mean that the mantle is a liquid. Look, for example, at this image that suggests a molten mantle taken originally from an online article. Although the mantle flows, it consists of solid minerals that flow ductilely. We know this from seismic data. Sound travels through the earth as compressional waves (P-waves) and as shear waves (S-waves). S-waves are not transmitted by liquids. S-waves DO travel through the mantle, so the mantle must be solid. S-waves do not pass through the Earth's core. This means that the core, or at least the outer core, must be liquid. See the good S-wave image in Figure 2.04, modified from the IRIS website).
If the mantle is largely solid, something must be happening to cause melting at divergent boundaries. Adding water to hot rock can cause melting, but this does not not appear to be happening at divergent boundaries. Sea water does seep into the ocean crust at mid-ocean ridges. We know this because heated and particle-laden sea water rises back into the ocean at "black smoker" vents (see Figure 2.05 above and this video). But the sea water circulation hydrates mid-ocean ridge basalt, converting it to spilite, rather than causing the mantle to melt.
Because chemical equilibrium is approached by rocks at high temperatures, petrologists can use controlled experiments on rocks in the laboratory to determine
physical conditions that should lead to melting in the mantle. Melting conditions for the mineral diopside are shown here in Figure 2.06. Notice that the melting temperature of diopside rises as pressure increases. This is the case for most solids and rocks under dry conditions because the solid has a higher density than the liquid. Click on the figure to see a larger version and choices of other melting experiment results (halite, peridotite). Increasing the pressure on a solid is not likely to melt it. However, decreasing the pressure on a solid (decompression) may lead to melting, even if the temperature is not increased. Will decompression in the rising mantle beneath divergent boundaries lead to partial melting? To answer this question we need to compare the experimental results for peridotite with a plausible variation of temperature with depth in the uppermost mantle.