Page 404 - Global Tectonics
P. 404

386   CHAPTER 12




           temperature profiles will be as shown to the right of   assume uniform viscosity throughout the convecting

           the figure. If there is no heating from below the tem-  layer, a parallel-sided rather than a spherical layer,

           perature in the interior of the fluid will be the same   thermal convection alone and no phase changes within


           as that at the base of the fluid layer (Fig. 12.5b). If   the fluid. In the Earth’s mantle it is now thought that
           there is some heating from below, in addition to inter-  the viscosity increases with depth and that buoyancy is
           nal heating (Fig. 12.5c), then the interior of the con-  in part created by compositional variations. As discussed

           vecting fluid will have an intermediate temperature   in Section 12.9 these two factors appear to stabilize
           between cases (a) and (b). This results in a greater   the convective pattern for hundreds of millions of years,
           drop in temperature across the upper boundary layer,   whereas the convective patterns developed in the
           and a lower drop in temperature across the lower   models of Fig. 12.6 are clearly unstable over this period
           boundary layer, compared to case (a). This effect of   of time.
           internal heating, whereby the top boundary layer is
           strengthened and the bottom boundary layer weak-
           ened, may therefore be applicable to the mantle.  12.5.2 Feasibility of
             The effect of internal heating and the lack of a lower
           thermal boundary layer is illustrated in Fig. 12.6. This   mantle convection
           shows the results of two numerical models with param-
           eters appropriate to the mantle. The three frames on   In order to gain insight into the feasibility and nature
           the left relate to a model with heating from below and   of convection within a spherical, rotating Earth, it is
           no internal heating and those on the right to a model   convenient to assume that the mantle approximates a

           with internal heating and no lower boundary layer. In   Newtonian fluid. Although this assumption may be

           the first case one can clearly see cold sinking columns   erroneous, it does allow simple calculations to be made
           and hot rising columns analogous to Fig. 12.5a. In the   on the convective process.
           right hand case only downwellings are apparent and the   The condition for the commencement of thermal
           upwellings are passive and widely distributed (cf. Fig.   convection is controlled by the magnitude of the dimen-
           12.5b).                                      sionless Rayleigh number (R a ), which is defi ned as the
             Although instructive, these models probably do not   ratio of the driving buoyancy forces to the resisting
           accurately simulate convection in the mantle as they   effects of the viscous forces and thermal diffusion.




                        218.3 Ma                      349.6 Ma





                        441.9 Ma                      587.0 Ma





                        536.7 Ma                      738.7 Ma





                        0.0       Temperature    2840.  0.0      Temperature    2840.
           Figure 12.6  Frames from numerical models illustrating (a) convection in a layer heated from below and (b)
           convection in a layer heated internally and with no heat from below (from Davies, 1999. Copyright © Cambridge
           University Press, reproduced with permission).
   399   400   401   402   403   404   405   406   407   408   409