[issm-support] ISSM Questions

[Morlighem, Mathieu (334H-Affiliate)]

Hello Gordan,

question 1:

We use the penalty method to constrain the ice temperature below pressure melting point (Tpmp).
If the ice exceeds Tpmp, we add a penalty to the stiffness matrix and load vector:

kappa x T = kappa x Tpmp

where kappa = kmax x 10^penalty_factor is a large number.
This forces T to be equal to Tpmp.
We can then retrieve the melting rate by computing how much energy was necessary to constrain T:

kappa (T-Tpmp) = L / c  Mb

where L is the latent heat, c is the heat capacity and Mb the melting rate.
The problem with this method is that the melting rate is VERY noisy and can reach values that are not physically realistic. This is due to the fact that no regularization is applied (Mb is not in H1 but L2, if you are familiar with Hilbert and Lebesgue spaces).

One alternative is to use the enthalpy model, which does not use penalties. It is still under development but you could give it a try. Just set md.thermal.isenthalpy=1;
Let us know if it works better.

question 2:

FS is much more sensitive than SIA numerically, and requires to have a mesh that is refined enough, boundary conditions that are consistent etc.
What we do generally is we don't jump from SIA to FS. You should try SSA (MacAyeal) and HO (Blatter/Pattyn) first. Also, most of the time, HO does a great job and it is not necessary to use a FS model.

Do not hesitate if you have more questions

Regards,
Mathieu
--
Mathieu Morlighem

Assistant Project Scientist
Department of Earth System Science
University of California, Irvine
Croul Hall, Irvine, CA 92697-3100
(949) 824-1353 (UCI office)
(818) 354-4134 (JPL office)
(626) 429-5780 (cell)
Mathieu.Morlighem at uci.edu<mailto:Mathieu.Morlighem at uci.edu>

On Jan 21, 2013, at 10:52 AM, Gordan Stuhne <gordan at atmosp.physics.utoronto.ca<mailto:gordan at atmosp.physics.utoronto.ca>> wrote:

Hello,

I am quite interested in your model, and have been working with it
using some of my own grids and data relating to Greenland.
My SIA (Hutter) results basically look rational for 100yr prognostic
runs (using thermal penalty_lock=10 for stability), but I have a few
questions.  Firstly, the basal melting rates reported in the
output files are very high (on the order of m/yr).  This is
presumably computed from latent heat fluxes where the
basal ice becomes temperate, but the details don't seem
to be documented in the code and I am not sure if I am
interpreting it right.

Secondly, I am having trouble trying to run the same case in
FS mode.  I've tried adjusting time-step, viscosity overshoot,
and other parameters with no success.  Is there some
useful rule of thumb for preparing the initial conditions?

Your insight would be greatly appreciated,
Thanks in advance,
Gordan Stuhne,
Dept. of Physics,
University of Toronto

_______________________________________________
issm-support mailing list
issm-support at issm.ess.uci.edu<mailto:issm-support at issm.ess.uci.edu>
http://issm.ess.uci.edu/cgi-bin/mailman/listinfo/issm-support

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://issm.ess.uci.edu/pipermail/issm-support/attachments/20130121/7e0296dc/attachment.html>