Geoflow 2 – Geophysics in Space!

How can we image the Earth’s interior when the deepest drillhole is 12 thousand meters deep?

Right, we go into space on the ISS. Ok, honestly this is quite counterintuitive, which is why this experiment immediately caught my attention as well. Today I read the Geoflow 2 experiment that is supposed to shed light on the interior of the Earth. In a very simple form, they took two solid spheres and put a liquid in between and gave it a spin.

However, the sciencedaily article doesn’t cover some important fact. They’re writing ?The liquid, of course, is the mantle?[1]Reference:
European Space Agency (2012, June 25). Geoflow: Space station experiments shed light on conditions deep inside Earth. ScienceDaily.
, since the Earth’s mantle is anything but a liquid, how is this a valid experiment? Any computer simulated convection model of the Earth works with plastic flow. A scientific fact that is confirmed with every earthquake strong enough we measure, because there are certain seismic waves that do not propagate through liquids.

Let’s first look at the setup of the experiment.
As before, they take two solid spheres and put them together. In between they put a liquid. Whereas, the solid spheres are the core and the crust. But now it gets tricky. How do you simulate gravity on a model, when ?real? gravity of the Earth is influencing your model? Yes, this is where you go to your local space agency and better have some good arguments. And they did.
They can regulate the temperature of the model within an accuracy of a tenth of a degree. This is very important for applied different temperature gradients on the model. A high voltage electrical field between the two spheres to simulate gravity. (This only works without ?real? gravity.) Now you can start to rotate the spheres and observe.

Now about this liquid that is supposed to resemble solid rock from the earth’s mantle.
This is where it gets tricky. When you do experiments including fluid dynamics you usually have some large scale phenomenon and need to fit everything into a small-scale model. Now fluid dynamics has some nice tricks. They’re called dimensionless quantities and resemble the combination of some important properties of the earth. I’ll give you an example: The Rayleigh number is a number that describes the relationship of buoyancy and viscosity in a fluid [2]Reference:
Turcotte, D.; Schubert, G. (2002). Geodynamics (2nd ed.). New York: Cambridge University Press
. When this number gets below a threshold heat transfer only occurs in the form of conduction, when the value exceeds a threshold heat transfer occurs in the form of convection. Now we have some good estimates of the Rayleigh number of the mantle and some other dimensionless quantities, if we get the fluid to resemble the dimensionless quantities of the mantle a fluid is a valid substitute for solid rock in small scale.

This image is directly from the experiment and shows you some aspects of the setup. We see, that the thickness of the gap is only 13.5 mm. Which really hints at the scale of this model. Have a look at the voltage! 10 kV is quite a number and creates our potential that simulate gravity. The temperature gradient is only equal or below 17? which is quite small. And rotation is set to be zero. I bet all of you got that from the table as well, but maybe I can shed some light on the latter two columns.
Maybe you realized that I haven’t talked about viscosity yet. The reason is that we have two “viscosities” in this table. The kinematic viscosity of the fluid and the dynamic viscosity ,which should coincide with of the Earth’s mantle. Dynamic viscosity is quite hard to measure but yet so important when regarding fluiddynamic properties [3]Reference:
Landau, L. D. und E. M. Lifschitz: Lehrbuch der theoretischen Physik, Band VI: Hydrodynamik, Akademie Verlag, 1991
Pr is the Prandtl number, a dimensionless qunatity. It describes the relation of kinematic viscosity and thermal diffusivity. It is often provided in property tables of materials as it solely depends on fluid properties [4]Reference:
White, F. M. (2006). Viscous Fluid Flow (3rd. ed.). New York: McGraw-Hill. ISBN 0-07-240231-8.
Ra is the Rayleigh number, which has been described before.
Ta is the Taylor number which links rotational (centrifugal) forces to viscous forces[5]Reference:
B?nard cells and Taylor vortices. Koschmieder, E. L. Cambridge University Press, 1993. p. 234
It is easy to see that this setup is for a stationary system, but I found a presentation concerning a rotating system. The numbers are the same, but you can compare values.Numbers of rotating system

The team has worked several years to get this tiny experiment into space. It’s just as big as a shoe box but to work with all the data they get, they had to update their computer cluster quite a bit.
Let’s see if this experiment will revolutionize our understanding of the Earth’s interior.

Or directly at the source:

via The Earth Story’s Facebook Wall

The following two tabs change content below.
... is a geophysicist by heart. He works at the intersection of machine learning and geoscience. He is the founder of The Way of the Geophysicist and a deep learning enthusiast. Writing mostly about computational geoscience and interesting bits and pieces relevant to post-grad life.

Latest posts by Jesper Dramsch (see all)

Posted in Science and tagged , , , .