|
Animations
of Thermohaline Doubly-advective Convection in Porous Media
Inequities in temperature are not the only first-order cause of flows
in Nature. With respect to geologic fluids, compositional effects can
play an important role- sometimes even dominating over thermal buoyancy-
in generating buoyancy forces. For example, in magma bodies, small differences
in the dissolved volatile content of H2O and CO2 or small variations
(a few wt%) in silica concentration can produce strong buoyancy effects.
Indeed, there is abundant evidence from large volume silicic volcanic
deposits (e.g., the Cenozoic great ignimbrite sheets in North America)
that gradients in H2O and in silica can effectively stabilize some silicic
systems with respect to convection. The picture in these cases is of
a density stratified cap with silica and H2O increasing upwards over
a large subjacent volume of quasi-homogenized and perhaps convecting
magma.
Another, example of a geological flow involving two sources of buoyancy
corresponds to thermohaline convection in the permeable rocks of the
oceanic and continental crust. In this case, spatial heterogeneity of
the salinity of ambient aqueous fluids as well as vertical and lateral
temperature differences gives rise to buoyancy forces that can drive
porous medium convection. We have studied porous media convection recently
(see publication list) and here present some 2-D animations of the evolution
of the salinity and temperature fields for some representative flows.
The parameters choosen for these simulations are typical of conditions
prevailing in the crust of planet Earth.
Three simulations are presented below. In each simulation, the evolution
of the salinity and tempearture fields are presented. All runs were
carried out on a 64 x 64 grid. Details may be found in the paper Schoofs,
S., Spera, F. J., Hansen, U., Chaotic thermohaline convection in low-porosity
hydrothermal systems. Earth and Planetary Science Letters, v. 174, p.
213-229.
The animations do not play with a constant timestep. For case 1, every 2000th
step represents a frame. For case 2, every 5000th step represents a
frame. The size of a time step is not exactly the same in each case;
a variable step size allows the calculation to be done more quickly
for the same resoltion compared to a constant timestep.
The three examples include:
Case 1: Rr =0.25, Ra =600, Le =100, aL =5x10-5, ar =10; tfinal=0.22
Case 2: Rr =0.40, Ra =600, Le =100, aL =5x10-5, ar =10; tfinal=1.36
Case 3: Rr = 1.0, Ra =600, Le =100, aL =5x10-5, ar =10; tfinal=1.36
*To
play these videos within your browser the latest version of Quicktime
is required.
|
