Dynamic hot rocks that make the earth move

When is a rock not a rock? Answer: when it is a fluid. Puzzled? Take a lump of solid rock, say the size of your fist, heat it to more than 1,000C at 200,000 times atmospheric pressure and leave in the earth's gravity field for 100 million years. Although you will not be around to see the result, the material will have deformed in accordance with the basic laws of flow of viscous fluids.

Most people know that the earth gets hotter with depth and has a layered structure comprising the crust, mantle and core. The connecting mantle, extending downwards from c.100km beneath the surface to the boundary with the outer core at a depth of 2,800km, is the largest part in volume and mass. The key here is heat and time. Early on, geologists were happy with the idea that the earth was hot inside and was cooling down. But how the heat was transported outwards remained a mystery. In the early part of the 20th century, physicists started to put in place the fundamentals of heat transfer in fluids. Among them was John William Strutt (later Lord Rayleigh), who played a seminal role in developing the theoretical framework for convective instabilities in fluids heated from below. His ideas would become integral in understanding convection in the solid mantle of the earth on geological timescales. One wonders what may have happened if Edwardian geologists had kept abreast of these developments in fluid dynamics. But there was strong opposition from within their own ranks. Despite having shown that convection in a compressible fluid was due to the difference between the adiabatic and actual temperature gradient, the highly influential Cambridge seismologist Harold Jeffreys rejected mantle convection outright. We now know mantle convection to be the underlying cause of plate tectonics.

This book treats the subject of mantle convection in a comprehensive and connected way, moving beyond the earth to consider convection in the terrestrial planets and the icy satellites of Jupiter. Aimed at the graduate market and billed as a research monograph by the authors, each an acknowledged expert in the field, it is a mammoth tome. Understandably some years in the making, it is a splendid effort, set to become the standard reference for years to come. On offer is an up-to-date discussion of the most recent ideas on mantle convection in the earth and planets, from its history to the latest thermodynamic data on earth materials and tomographic studies crucial for constraining numerical and theoretical models. For those wanting a deeper understanding, the dimensional forms of conserved properties and the Navier-Stokes equations in Cartesian, cylindrical and spherical polar coordinates are also given.

Mantle convection differs in several important ways from large-scale geophysical flows in the atmosphere and oceans. The small depth of these layers relative to their horizontal dimensions means that vertical motion, although important, is not dominant. Things are different in the mantle, where plumes (a special type of convection), subduction and return flow are significant features in the vertical plane. As a consequence, mantle flow is truly three-dimensional, but only recently has it been possible to build realistic models that capture this. This state of affairs is due to enhanced computing power with faster processing times, coupled with advances in scientific visualisation techniques. The resulting numerical models are as stunning as they are colourful and allow geoscientists to glimpse for the first time the beauty and complexity arising from our planet's thermodynamic requirement to lose heat.

Nick Petford is reader in geology, Kingston University.