Where to start on a transient grounding line model

Extension of Elmer in computational glaciology
jvitow
Posts: 2
Joined: 01 Dec 2020, 23:41
Antispam: Yes

Where to start on a transient grounding line model

So I just finished the Elmer Beginners course and have an okay idea of how the components work, but I am a bit lost in starting to create my own model runs.

I want to eventually have a 2D marine-terminating ice stream and I want to model the movement of the grounding line subject to various forces (for example sea level rise).

I understand the basic pieces of what I need to do,
create an extruded mesh with an area of higher resolution near the grounding line
for my bodies, I believe I need 3, the ice itself, an upper free surface and the ocean water.
For boundary conditions, I believe there are 2, the bedrock, the calving front.
for body forces, there is the pressure of the ocean, and of course gravity.

I'm not too worried about temperature right now so for solvers I believe I need a flotation check, as well as a full stokes solver and an upper free surface solver, as well as potentially a few solvers to export data or to read in my initial conditions.

If anyone has any suggestions about how this sounds or knows of any currently existing models which solve this, I'd be grateful for any help!
fgillet
Posts: 46
Joined: 30 Sep 2010, 16:58

Re: Where to start on a transient grounding line model

Hello,
The Tete Rousse example during the beginner course is a good example; the cavity problem is the same as the grounding line (an ice shelf is a big open cavity ).

You also have a 2D example in the elmerice tests (under the elmer sources elmerice/Tests/GL_MISMIP).

With Stokes you have to solve the equations for the 2 free surfaces, top and bottom, because with higher order stresses you may not be exactly at the hydrostatic pressure (while in shallow you only need to solve the equation for one variable, the thickness then use floatation to deduce zb and zs).

You don't have to model the ocean explicitely; its effect is replace by boundary conditions (pressure and melting/accretion).

Before going straight away to the modelling part, my advice is always to first make some simple drawing to represent what you want to model to see the equations to solve and the boundary conditions to apply.

fabien
jvitow
Posts: 2
Joined: 01 Dec 2020, 23:41
Antispam: Yes

Re: Where to start on a transient grounding line model

Hi Fabien, thanks for the suggestion to look at the Stokes example, I've been playing with it for the last while, changed the MATC to LUA, stuff like that. Right now what I'm trying to do is change the direction of the slope, so that the grounding line is on a retrograde slope, below sea level. This seems to cause issues as I get unphysical results. Am I correct in thinking that the way to do this is to use the distance solver and force inland nodes to remain in contact with the bedrock?