Repository for work related to NASA ROSES award 80NSSC21K0748 (2021-24).
The stability of the Greenland Ice Sheet in a warming climate is a critical societal concern. Predicting Greenland's contribution to sea level remains a challenge as historical simulations of the past decades show limited agreement with observations. In this project, we develop a data assimilation framework that combines sparse observations and the ice sheet model PISM to produce a reanalysis of the state of the Greenland Ice Sheet from 1980 to 2020 using probabilistic filtering methods.
Get pism-ragis source from GitHub:
$ git clone [email protected]:pism/pism-ragis.git
$ cd pism-ragis
Optionally create Conda environment named pism-ragis:
$ conda env create -f environment.yml
$ conda activate pism-ragis
or using Mamba instead:
$ mamba env create -f environment.yml
$ mamba activate pism-ragis
Install pism-ragis:
$ pip install .
We strive to make our workflow reproducible, however, perfect reproducability is hard to achieve. For example, third-party providers of input data sets may change their access patterns, breaking the workflow. PISM needs to be installed on an HPC system with ample disk space (10s of TB).
Please install PISM ![Parallel Ice Sheet Model (PISM)]