Analog modeling

Basic principles

What is analog modeling?

This approach can be useful to answer specific questions, test an hypothesis, or investigate the influence of various parameters in geological processes, because:

To be comparable with nature, analog models must respect scaling principles for lengths, forces and consequently time. Experiments must also be reproductible: a same experiment performed with the same parameters several times must give the same results (else you are obviously missing something/doing something wrong/not controlling all the variables).

A parametric study can be performed by runing a serie of experiments and changing only one parameter at a time time. This allows you to link changes in the results to a precise parameter, and quantify it (e.g. if I increase the density contrast by XX, then the subduction velocity will increase by YY).

Tools of the job & procedures

Many materials and devices can be used for analog modeling in geosciences depending on the scale and type of processes simulated (from deep mantle convection to surface erosion). Materials are chosen so that their mechanical behavior is comparable to the natural rocks they represent at the scale of the experiment. Their rheological properties must therefore be defined prior to the experiments. We adapt the recording devices to the variables we want to follow (velocities, deformation, topography, flow pattern, temperature, stress...). Top and side view photographs are usually taken, and in some cases cross-sections throughout the model can be performed at the end of the experiment. Here are a few examples of materials and devices commonly used.

Sand is a brittle material typically used to simulate the upper crust. Different type of sand can be used, as well as mixed with other materials to adjust density, friction angle, cohesion, etc. Silicons are Newtonian materials that can model the whole lithosphere or specific sub-layers depending on the chosen density and viscosity. Gelatins are elastic materials that can be used to simulate the seismic cycle (crustal scale) or isostatic rebound (lithosphere scale) for example. Parafin (raw material at the top, colored and deformed after experiment at the bottom) are plastic materials used to model the whole lithosphere. Glucose syrup or honey are Newtonian and fitted to simulate the asthenosphere or low viscosity layers within the lithosphere.
Densimeters can measure materials' density from mass measured in the air and in water. Rotational viscometers measures the viscosity of materials by determining the force required to turn the central axis aginst the viscous forces of the fluid. Rheometers can measure viscosity, elastic properties and various other parameters under controled stress or strain and temperature conditions.
Digital camera to take pictures during the experiment. Laser scanners can record the surface topography. All devices are usually connected to a computer that controls the recording.

Examples of experiments

Oceanic and continental subduction

In the following example, oceanic and then continental subduction is simulated with silicone plates representing the whole lithosphere floating over glucose syrup (asthenosphere). The dense oceanic plate drags the lighter continental upper plate leading to continental subduction. In this project, we explored the key parameters controlling the curvature of a mountain belt.

Various parameters were tested including plates' interface strength, plates' rheology and boundary conditions. In SH6 experiment (left side photo), plates' interface is strong, oceanic plates (green) bound a continental indenter (brown). The experiment results in a concave shape of the suture. In SH22 experiment (right side photos) plates' interface is weak and oceanic lithosphere is present only at the front of the subducting plate (already subducted on the central photo, visible in lateral view). The experiments results in a convex shape of the suture bounded by two syntaxes. Looking closely at the displacement field (right photo), we observe variations in the amplitude of the shortening component between the syntaxes and the center of the subduction zone as well as a strong rotational component at the free edge of the upper plate.


Bajolet F., Replumaz A., Lainé R., 2013. Orocline and syntaxes formation during subduction and collision, Tectonics, 32(5), 1529-1546, doi:10.1002/tect.20087.

Collision stage of a mature plateau

In order to explore the dynamics at plateaus' margins, I realised gravity-scaled analog models simulating the shortening of two adjacent lithospheres: a strong one representing the Asian craton, and a weak one representing the Tibetan plateau. The model is layered with sand (brittle crust) and silicones (lower crust and lithospheric mantle) placed on a low-viscosity material (asthenosphere). In free boundary experiments, lateral escape of the model is allowed by placing a neutral (very weak) silicone at one side.

Final stage of a free boundary experiment (top view)

This photo is a top view of an experiment performed with a free boundary, after 25% of shortening. Initially, both high plateau and cratonic lithosphere domains have the same size. A neutral silicone allows lateral escape of the model. The high plateau is preferentially shortened and thickened compared to the cratonic lithosphere. Pervasive deformation in the plateau gradually evolves from constriction with vertical thickening marked by thrusts and tranpression near the confined boundary, to vertical thinning and lateral flow with transtensional faults near the free boundary. On the contrary, the cratonic lithosphere undergoes smaller amount of localized deformation characterized by much moderate lateral flow and strike-slip faults. Extension zones with grabben structures due to lateral escape develop at the boundary between the high plateau and the cratonic lithosphere.

Final stage of a confined experiment (cross-section)

This typical cross-section is realised after 50% of shortening, by freezing the lithospheric part of the model and cut it with a circular saw. We observe the lower crust of the high plateau injecting the lower crust of the cratonic lithosphere along a shear zone emerging as localized thrusts in the cratonic lithosphere. The overflowing channel of lower crust is domed at its extremity and does not break the surface. A filament of high plateau's lower crust at the roof of the channel is brushed backward, highlighting an important back-flow (i.e. toward the plateau) in the middle crust. The cratonic lithospheric mantle has initiated continental subduction. Homogeneous thickening of the plateau is characterized by burial of crustal pop-downs.


Bajolet F., Chardon D., Martinod J., Gapais D., Kermarrec J.J., 2015. Synconvergence flow inside and at the margin of orogenic plateaus: Lithospheric-scale experimental approach, Journal of Geophysical Research, 120(9), 6634-6657, doi:10.1002/JB2015012110.

Where to find experimental labs?

Several laboratories around the world have an experimental plateform. Given the complexity the experimental procedures that often require a long-term development, all laboratories are not modeling everything. Rather, they focus on specific topics/techniques. If your are looking for collaboration and facilities, here are a few labs specialized in geodynamics/tectonics issues: