grounding line migration in PIG

in-woo

I want to practice grounding line migration part using PIG example.

According to user manual, grounding line migration is solved using md = solve(md, TransientSolutionEnum).

So I add transient model part like this, copying code of 'Greenland transient modeling'

mdSSA_tran.groundingline.migration = 'SubelementMigration';


     disp('Transient running');
    
    ncdata='../Data/Antarctica_5km_withshelves_v0.75.nc';
   
    mdSSA_tran = loadmodel('./Models/PIG.Control_drag');
    
    mdSSA_tran.groundingline.migration = 'SoftMigration';
       
    x1  = ncread(ncdata,'x1');
    y1  = ncread(ncdata,'y1');
    smb = ncread(ncdata,'presprcp');
     
    smb = InterpFromGridToMesh(x1,y1,smb',mdSSA_tran.mesh.x,mdSSA_tran.mesh.y,0);
	smb = smb*mdSSA_tran.materials.rho_freshwater/mdSSA_tran.materials.rho_ice;
	smb = ;
	mdSSA_tran.smb.mass_balance = 
    
    %Time Step
    mdSSA_tran.timestepping.time_step=0.2;
	mdSSA_tran.timestepping.final_time=20;
	mdSSA_tran.settings.output_frequency=1;
    
    
	%Additional options
	mdSSA_tran.inversion.iscontrol=0;
	mdSSA_tran.transient.requested_outputs={'IceVolume','TotalSmb', ...
		                             'SmbMassBalance'};
	mdSSA_tran.verbose=verbose('solution',true,'module',true,'convergence',true);

	%Go solve
	mdSSA_tran.cluster=generic('name',oshostname,'np',2);
	mdSSA_tran=solve(mdSSA_tran,TransientSolutionEnum);

But I got an error like this


model not consistent: requesting grounding line migration, but bathymetry is absent!

how can I change the code "md.groundingline.migration"?

mathieumorlighem

Hi In-Woo,

your changes look good. The problem is that md.geometry.bed is undefined (i.e., NaN) and it defines the bathymetry under floating ice (md.geometry.base = md.geometry.bed for grounded ice and md.geometry.bed<md.geometry.base for floating ice). You can use Bedmap2 data to define md.geometry.bed and then it should run.

Good luck!

Best

Mathieu

in-woo

Thank you for your reply

From what I understand, md.geometry.bed is needed in grounding line migration solver.

After your reply, I just use ncdata of Antarctica geometry data, although I can use the bedmap2 data,

Then, I add the following code line at 'PIG.par' file.


md.geometry.base  = InterpFromGridToMesh(x1,y1,topg,md.mesh.x,md.mesh.y,0);
md.geometry.bed  = InterpFromGridToMesh(x1,y1,topg,md.mesh.x,md.mesh.y,0);

then I got a following consistency error.


model not consistent: bed superior to base on floating ice!

I can't understand following error line...

OR************************************************************

do i have to give a some values in md.basalforcing.floatingice_melting_rate and md.basalforcing.groundedice_melting_rate??

Thank you

heleneseroussi

Dear In-Woo,

The value that you put in md.geometry.bed and md.geometry.base in your models are the same as they come from the same dataset (topg). This is perfect for grounded ice, but for floating ice, you need the bed to be strictly below the base, as bed has to be equal to base on grounded ice and strictly below base on floating ice (where md.mask.groundedice_level_set is negative). So you have to lower the bed for the floating part of you domain.

Values that you put in md.basalforcing.floatingice_melting_rate and md.basalforcing.groundedice_melting_rate are used in the mass transport equation to know how much mass is lost or gained at the base of the ice sheet/shelf.

Thanks,

Helene

in-woo

Oh Thank you

I solve the problem that some estimated md.geometry.base based on the floating ice was under the md.geometry.bed


%ensure hydrostatic equilibrium on ice shelf: 
di=md.materials.rho_ice/md.materials.rho_water;

%Get the node numbers of floating nodes
% 1 : grounded ice
% -1 : floating ice
pos=find(md.mask.groundedice_levelset<0); 

%apply a flotation criterion on the precedingly defined nodes and
%redefine base and thickness accordingly
md.geometry.thickness(pos)=1/(1-di)*md.geometry.surface(pos);
md.geometry.base(pos)=md.geometry.surface(pos)-md.geometry.thickness(pos);

Then, transientsolutionEnum was well performed.

But result was not what I expected that md.results.TransientSolutionEnum. had groundingline result.

It just showed only base elevation. So I have a question about initial parameterization process.

In paramaterization results, 'find(md.mask.groundedice_levelset ==0)' has no 0 values. In md.mask said like this

grounded ice if > 0, grounding line position if = 0, floating ice if < 0

So, is 0 value needed to calculated grounding line migration??

Thank you

heleneseroussi

Hi In-Woo,

md.results.TransientSolution should have most of the properties computed (base, surface, thickness, velocity, grounding line level set, ...).

The grounding line position is theoretically located where md.mask.groundedice_level_set is equal to zero, which does not have to be on a node of the mesh. The grounding line position will fall within an element most of the time. However, it can varies depending on what parameterization you use for the grounding line (in md.groundingline.migration). More details on the different parameterizations implemented in ISSM can be found in Seroussi et al, 2014 (Hydrostatic grounding line parameterization in ice sheet models, see publication page of the ISSM website). I would recommend to use the 'SubelementMigration' or 'SubelementMigration2'.

Best,

Helene