GLIDE - The Ice Sheet Model

GLIDE is a 3D finite-difference ice sheet model. GLIDE is the core system of the GLIMMER ice sheet model. It was originally developed by Tony Payne at the University of Bristol, but has since been modified to form the land-ice component of the GENIE project. This system is a collaborative effort to create a standard ice sheet model which can be used by other projects.

Some features of GLIDE:

• The model is written in Fortran95 with some code generated by python scripts. The model is developed to a large extent on Linux systems and works with the NAG and INTEL f95 compilers. It can also be compiled under Windows.
• The model is hosted on NeSCForge and can be accesed via the GLIMMER project pages.
• Model I/O uses the netCDF library and complies with the CF standard for Climate and Forecast Data.
• The model is implemented as a library and can be linked to programs providing a climate driver. Various climate drivers are included with the model.
• The model comes with a visualisation tools written in python using GMT.
• The model is reasonably well documented. The documentation is written in LaTeX.
• Licensed under the GNU General Public Licence.

These pages document my involvement with GLIMMER and GLIDE and serve as a place where I can put some experiments and musings. You can also get a pdf of the current documentation or browse the HTML Documentation online (complain to me when it is too out of date).

News and Musings

GLIMMER, Version Control and Build-Systems (5 December 2005)

A while back I gave a tutorial of how to use version control systems, CVS and subversion, and GNU automake and autoconf in the context of GLIMMER. The tutorial also included some bits on the GLIMMER configuration files and the code generator. You can get the pdf slides or if you want something more compact the handouts of the tutorial. The presentation was produced with the beamer class for LaTeX. Finally, no more powerpoint!

Using exact solutions to verify GLIMMER - Part 2 (7 November 2005)

I've also run some experiments for test D and created some new images. You can find the results here. I consider these tests as successful and will add a module to the GLIMMER cvs repository once I get to go from Ed.

HTML Documentation (4 November 2005)

We can now generate HTML Documentation for GLIMMER. The LaTeX sources are translated using TeX4ht.

Using exact solutions to verify GLIMMER (21 October 2005)

You can find the initial results of the tests proposed by Ed Bueler and others to verify ice sheet models using exact solutions for the isothermal case here.

Basal sliding (11 May 2005)

Trying to figure out why we get very narrow regions of fast flow when basal sliding is switched on. This effect can be clearly seen in my simulations of the past Fennoscandian ice sheet. I have repeated the EISMINT-2 Exp H with varying sliding parameters to investigate this further.

Playing with the non-linear solver (1 December 2004)

I've fixed some bugs with the non-linear solver and played with it by varying the timestep size of the EISMINT-2 Experiment H. The result suggest that it doesn't really matter if the non-linear or linear solver are used, at least for this experiment. Obviously, there might also still be a bug lurking.

Heat flux due to sliding (29 November 2004)

The additional heat flux due to sliding was indeed missing. I've fixed that now. Both Experiment G and Experiment H look quite good now as well.

I've also written a new PyCF program which will compare EISMINT-2 output with the results reported in the EISMINT-2 paper (e.g. for the EISMINT-2 Experiment H). Again, the results look good.

EISMINT-2 (26 November 2004)

The EISMINT-2 tests are now also added and look very promising. However, there seems to be a problem when the ice is sliding: basal temperatures are too low and solution is patchy.

EISMINT-1 (24 November 2004)

I finally managed to reproduce the EISMINT-1 results.