Difference between revisions of "ISMIP-HOM test suite exercise"

From Interactive System for Ice sheet Simulation
Jump to: navigation, search
Line 1: Line 1:
In this exercise, we will test out Glimmer/CISM's higher-order stress balance subroutines by running the model through a few of the [http://homepages.ulb.ac.be/~fpattyn/ismip/ ISMIP-HOM] test suite problems. Specifically, we'll run tests A and C with periodic basal topography (test A) and periodic basal friction (test C) for wavelengths of 160, 80, and 40 km.  
+
In this exercise, we will test out Glimmer/CISM's higher-order stress balance subroutines by running the model through a few of the [http://homepages.ulb.ac.be/~fpattyn/ismip/ ISMIP-HOM] test suite problems. The tests we'll run are for 3d models, so the domain and boundary conditions vary in the ''x'' and ''y'' directions (i.e. in map plane). For test A, the topography varies periodically in ''x'' and ''y'', and for test C, the basal traction varies periodically in ''x'' and ''y''. While the amplitude of the variations is the same for all tests, the wavelength is decreased by a factor of two for each successive test. For λ=160 km, the velocities solutions essentially look like that from a shallow ice model. Halving λ to 80 km, then to 40, 20, 10, and finally 5 km, the higher-order components of the stress balance become successivly more important to the velocity solution.
 +
 
 +
The tests are set up so that the scaled
  
  

Revision as of 09:31, 3 August 2009

In this exercise, we will test out Glimmer/CISM's higher-order stress balance subroutines by running the model through a few of the ISMIP-HOM test suite problems. The tests we'll run are for 3d models, so the domain and boundary conditions vary in the x and y directions (i.e. in map plane). For test A, the topography varies periodically in x and y, and for test C, the basal traction varies periodically in x and y. While the amplitude of the variations is the same for all tests, the wavelength is decreased by a factor of two for each successive test. For λ=160 km, the velocities solutions essentially look like that from a shallow ice model. Halving λ to 80 km, then to 40, 20, 10, and finally 5 km, the higher-order components of the stress balance become successivly more important to the velocity solution.

The tests are set up so that the scaled 



use some python scripts developed by Tim Bocek and Jesse Johnson to configure Glimmer/CISM to


Test A and C for wavelengths of 160, 80, and 40 km (skip shorter wavelength, longer running tests).

Play w/ the grid spacing to see how that affects results?

Can we get a 0-order solution for these as well?