Hydro-Dynamics Coupled Models

bdefleurian

Hei
You actually need to set the friction.coupling flag to 4 to have coupling. Setting the flag to 3 will result in a forced run (i.e. you need to provide the effective pressure to the model). I expect that you are not working on the development branch of ISSM (hydro Nightly runs here are now 9XX)? If you want to couple 333 you will need to run a dynamic solver too, in this test only the hydrology is runnning.

alain

bdefleurian Thanks. You're right and I check the newest version of ISSM 4.17and find it exsited test 9XX. And I try to run the test900-905.m.
Add themd.friction.coupling= 4; md.transient.isstressbalance=1; but its transient solution don't contain my required the velocity field.
I have changed themd=solve(md,'Hydrology');into md=solve(md,'Transient');.
Any questions in my operation?

bdefleurian

Which off the tests are you running? I do not expect any of those tests to succeed in a coupled run straight out of the box, they are not meant to solve the ice dynamics part of the code so there are quite a lot of things missing here to perform a coupled simulation. I am surprised that you get any results at all.

alain

bdefleurian It don't have results except hydrology solution and have some bugs. So I want to know to run it coupled for a refer example.

bdefleurian

What are the error message you are getting?

alain

bdefleurian Hello， I run the test905.m and a example about SquareNoDynHydrologyDCSmbCoupled.
It look like a example of SMB coupled Hydrology. So I add the dynamics part of modle missing.


md=setflowequation(md,'SSA','all');
md=SetIceShelfBC(md);
md.transient.isstressbalance=1;
md.friction.coupling=4;

%Test905 Name: SquareNoDynHydrologyDCSmbCoupled
md=triangle(model(),'../Exp/Square.exp',100000.);
md=setmask(md,'','');
md=parameterize(md,'../Par/SquareNoDyn.par');
md.cluster = generic('name',oshostname(),'np',1);

md=setflowequation(md,'SSA','all');
md=SetIceShelfBC(md);
md.transient.isstressbalance=1;
md.friction.coupling=4;

md.transient.ishydrology=1;
md.transient.issmb=1;
md.hydrology=(hydrologydc);
md.hydrology=initialize(md.hydrology,md);
md.smb=(SMBgradientscomponents);

md.hydrology.isefficientlayer=1;

md.hydrology.sedimentlimit_flag=1;
md.hydrology.sedimentlimit=400.0;
md.hydrology.sediment_transmitivity=3.0*ones(md.mesh.numberofvertices,1);
md.hydrology.mask_thawed_node=ones(md.mesh.numberofvertices,1);

md.hydrology.mask_eplactive_node=zeros(md.mesh.numberofvertices,1);
md.hydrology.epl_conductivity=3.;
md.hydrology.epl_initial_thickness=20;
md.hydrology.epl_colapse_thickness=1.0e-3;
md.hydrology.epl_thick_comp=0;
md.hydrology.epl_max_thickness=1;

md.hydrology.spcsediment_head=NaN*ones(md.mesh.numberofvertices,1);
md.hydrology.spcepl_head=NaN*ones(md.mesh.numberofvertices,1);

md.initialization.sediment_head=zeros(md.mesh.numberofvertices,1);
md.initialization.epl_head=zeros(md.mesh.numberofvertices,1);
md.initialization.epl_thickness=ones(md.mesh.numberofvertices,1);

md.hydrology.steps_per_step=5;
md.smb.steps_per_step=7.;
md.timestepping.time_step=2.;
md.timestepping.final_time=20.0;

smb_step=md.timestepping.time_step/md.smb.steps_per_step;
duration=[md.timestepping.start_time:smb_step:md.timestepping.final_time];

ddf=10.0e-3;
md.smb.accuref=[[0.5 0.5];[0. 2.0]];
md.smb.accualti=0.0;
md.smb.accugrad=0.0;

md.smb.runoffref=0.9*duration*ddf;
md.smb.runoffref=[md.smb.runoffref;duration];
md.smb.runoffalti=0.0;
md.smb.runoffgrad=-6.5e-3*ddf;  %lapse rate *ddf*day per year

md.basalforcings.groundedice_melting_rate=zeros(md.mesh.numberofvertices,1);
md.basalforcings.floatingice_melting_rate=zeros(md.mesh.numberofvertices,1);

md=solve(md,'Transient');

test905
boundary conditions for stressbalance model: spc set as zero
no balancethickness.thickening_rate specified: values set as zero
no hydrology.basal_moulin_input specified: values set as zero
uploading input file and queueing script
launching solution sequence on remote cluster

Ice Sheet System Model (ISSM) version 4.17
(website: http://issm.jpl.nasa.gov contact: issm@jpl.nasa.gov)

??? Error in ==> ./classes/IoModel.cpp:1712
FetchDataToInput error message: "md.friction.effective_pressure" not found in binary file

===================================================================================
= BAD TERMINATION OF ONE OF YOUR APPLICATION PROCESSES
= EXIT CODE: 1
= CLEANING UP REMAINING PROCESSES

= YOU CAN IGNORE THE BELOW CLEANUP MESSAGES

Error using loadresultsfromdisk (line 17)

============================================================
Binary file test905.outbin not found

This typically happens when the run crashed.

Please check for error messages above or in the outlog

Error in loadresultsfromcluster (line 45)
md=loadresultsfromdisk(md,[md.miscellaneous.name '.outbin']);

Error in solve (line 169)
md=loadresultsfromcluster(md);

Error in test905 (line 60)
md=solve(md,'Transient');

It seem like no find the effective presssure. What caused the mistake?

bdefleurian

You need to give an initial effective pressure. This is due to the fact that on the first timestep the effective pressure is needed before the hydrological model runs for the first time.

alain

bdefleurian It's Okay. But it seem like no using coupled rule with hydrology models. Because I comment md.friction.coupling=4; it will run same transient solution. It means two solution of vel have no diffence. I didn't know the inter rules. Or we just use it securely.

bdefleurian

I am not sure that I completely understand your message. Test 905 is however not designed to do some troubleshouting of the hydro-ice dynamic coupling. If you look in more detail at the model you will see that there is no surface slope here, hence no driving stress and no displacement. I would suggest to look into the examples rather than the Nightly runs to design a coupled test. The Nightly runs are here to control the code numeric and not necessarily to produce relevant physical results.

lidaan92

bdefleurian I get it. I need a real glacier to test the hydro-dyn question. It's real that present version issm could solve the simple hydro-dynamic relationship by my above model setup scenes. It doesn't need extra modif the source code cpp file.

bdefleurian

A synthetic glacier would work too just something a bit more realistic than an ice cube. The code as it is works but you will probably need to look more closely at the Boundary Condition that you want to apply. You can check my set-up for the SHMIP experiments that would give you a good start in term of the hydrology model. Be aware that this model was built on a previous version of the code so some debugging might be needed.

alain

bdefleurian Really grateful to you. Best wish to you.