Skip to content

Latest commit

 

History

History
130 lines (95 loc) · 5.29 KB

README.md

File metadata and controls

130 lines (95 loc) · 5.29 KB
Raw

PyISSM

Experimental code on running the Ice-sheet and Sea-level System Model (ISSM) via Python instead of the default Matlab interface. Focusing on Pine Island Glacier (based on this tutorial).

Quickstart

Launch in Pangeo Binder (Interactive jupyter lab environment in the cloud).

Binder

Installation

Start by cloning this repo-url

git clone <repo-url>

Then I recommend using conda to install the dependencies.

cd pyissm
conda env create -f environment.yml

Activate the conda environment first.

conda activate pyissm

Then clone the ISSM svn trunk repository. You will need to have subversion installed.

echo 'anon' | svn checkout --username anon --password-from-stdin -r 25855 https://issm.ess.uci.edu/svn/issm/issm/trunk

After that, you will need to compile some of the dependencies shipped with ISSM.

export ISSM_DIR=/path/to/pyissm/trunk
source $ISSM_DIR/etc/environment.sh && cd $ISSM_DIR/externalpackages/m1qn3 && ./install.sh
source $ISSM_DIR/etc/environment.sh && cd $ISSM_DIR/externalpackages/triangle && ./install-linux.sh

Setup some environment variables and configuration settings, and patch the configure file to remove the 'm' for minimal (Python 3.8 and above only).

export CONDA_DIR=/path/to/.conda/envs/pyissm
cd $ISSM_DIR && source $ISSM_DIR/etc/environment.sh && autoreconf -ivf
cd $ISSM_DIR && \
    ./configure \
    --prefix="$ISSM_DIR" \
    --disable-static \
    --enable-debugging \
    --enable-development \
    --with-numthreads=8 \
    --with-python-version=3.7 \
    --with-python-dir="$CONDA_DIR" \
    --with-python-numpy-dir="$CONDA_DIR/lib/python3.7/site-packages/numpy/core/include/numpy" \
    --with-fortran-lib="-L$CONDA_DIR/lib/gcc/x86_64-conda-linux-gnu/7.5.0/ -lgfortran" \
    --with-mpi-include="$CONDA_DIR/lib/include" \
    --with-mpi-libflags="-L$CONDA_DIR/lib -lmpi -lmpicxx -lmpifort" \
    --with-metis-dir="$CONDA_DIR/lib" \
    --with-scalapack-dir="$CONDA_DIR/lib" \
    --with-mumps-dir="$CONDA_DIR/lib" \
    --with-petsc-dir="$CONDA_DIR" \
    --with-triangle-dir="$ISSM_DIR/externalpackages/triangle/install" \
    --with-m1qn3-dir="$ISSM_DIR/externalpackages/m1qn3/install"
# sed -i 's/-lpython${PYTHON_VERSION}m/-lpython${PYTHON_VERSION}/g' $ISSM_DIR/configure

For completeness, here's the equivalent commands for compiling under MATLAB. The 'with-matlab-dir' filepath needs to be set to where MATLAB is installed.

cd $ISSM_DIR && \
    ./configure \
    --prefix="$ISSM_DIR" \
    --disable-static \
    --enable-debugging \
    --enable-development \
    --with-numthreads=8 \
    --with-matlab-dir="/opt/Matlab/R2019a/" \
    --with-fortran-lib="-L$CONDA_DIR/lib/gcc/x86_64-conda-linux-gnu/7.5.0/ -lgfortran" \
    --with-mpi-include="$CONDA_DIR/lib/include" \
    --with-mpi-libflags="-L$CONDA_DIR/lib -lmpi -lmpicxx -lmpifort" \
    --with-metis-dir="$CONDA_DIR/lib" \
    --with-scalapack-dir="$CONDA_DIR/lib" \
    --with-mumps-dir="$CONDA_DIR/lib" \
    --with-petsc-dir="$CONDA_DIR" \
    --with-triangle-dir="$ISSM_DIR/externalpackages/triangle/install" \
    --with-m1qn3-dir="$ISSM_DIR/externalpackages/m1qn3/install"

Finally compile ISSM!

cd $ISSM_DIR
make --jobs=8
make install

You'll need to export some more environment variables for things to work. Plus add a line into your .bashrc file to apply the settings each time you restart. See also official instructions at https://issm.jpl.nasa.gov/documentation/addpath/

export PYTHONPATH=$ISSM_DIR/bin:$PYTHONPATH
export PYTHONPATH=$ISSM_DIR/lib:$PYTHONPATH
echo "source $ISSM_DIR/etc/environment.sh" >> $HOME/.bashrc

Syncing/Updating to new dependencies

conda env update -f environment.yml
cd $ISSM_DIR && svn update
make clean
# now recompile ISSM following instructions above

Running jupyter lab

conda activate pyissm
python -m ipykernel install --user --name pyissm  # to install conda env properly
jupyter kernelspec list --json                    # see if kernel is installed
jupyter lab &

TODO

  • Use newer dataset inputs (e.g. BedMachine, ALBMAPv1, MEaSUREs Phase Map of Antarctic Ice Velocity)
  • Go from Python 2.7 to Python 3
  • Get reproducible binder build to work (using Dockerfile)
  • PyGMT plots!

Notes

  • ISSM installed in $ISSM_DIR which is /home/user/pyissm/trunk, activated by running source $ISSM_DIR/etc/environment.sh
  • Pine Island Tutorial located in $ISSM_DIR/examples/Pig, and can be copied to /home/user/pyissm/Pig
  • Datasets found under /home/user/pyissm/Data

References

  • Larour, E., Seroussi, H., Morlighem, M., & Rignot, E. (2012). Continental scale, high order, high spatial resolution, ice sheet modeling using the Ice Sheet System Model (ISSM). Journal of Geophysical Research: Earth Surface, 117(F1). https://doi.org/10.1029/2011JF002140