DeepIceDrain

Mapping and monitoring deep subglacial water activity in Antarctica using remote sensing and machine learning.

Getting started

Quickstart

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

Usage

Once you've installed properly installed the deepicedrain package (see installation instructions further below), you'll have access to a range of tools for downloading and performing quick calculations on ICESat-2 datasets. The example below shows how to calculate ice surface elevation change on a sample ATL11 dataset between ICESat's Cycle 3 and Cycle 4.

import deepicedrain
import xarray as xr

# Loads a sample ATL11 file from the intake catalog into xarray
atl11_dataset: xr.Dataset = deepicedrain.catalog.test_data.atl11_test_case.read()

# Calculate elevation change in metres from ICESat-2 Cycle 3 to Cycle 4
delta_height: xr.DataArray = deepicedrain.calculate_delta(
      dataset=atl11_dataset, oldcyclenum=3, newcyclenum=4, variable="h_corr"
)

# Quick plot of delta_height along the ICESat-2 track
delta_height.plot()

Installation

Basic

To just try out the scripts, download the environment.yml file from the repository and run the commands below:

cd deepicedrain
conda env create --name deepicedrain --file environment.yml
pip install git+https://github.com/weiji14/deepicedrain.git

Intermediate

To help out with development, start by cloning this repo-url

git clone <repo-url>

Then I recommend using conda to install the non-python binaries. The conda virtual environment will also be created with Python and poetry installed.

cd deepicedrain
conda env create -f environment.yml

Activate the conda environment first.

conda activate deepicedrain

Then install the python libraries listed in the pyproject.toml/poetry.lock file.

poetry install

Finally, double-check that the libraries have been installed.

poetry show

(Optional) Install jupyterlab extensions for interactive bokeh visualizations.

jupyter labextension install @pyviz/jupyterlab_pyviz
jupyter labextension install dask-labextension

jupyter labextension list  # ensure that extensions are installed

Advanced

This is for those who want full reproducibility of the conda environment, and more computing power by using Graphical Processing Units (GPU).

Making an explicit conda-lock file (only needed if creating a new conda environment/refreshing an existing one).

conda env create -f environment.yml
conda list --explicit > environment-linux-64.lock

Creating/Installing a virtual environment from a conda lock file. See also https://docs.conda.io/projects/conda/en/latest/user-guide/tasks/manage-environments.html#building-identical-conda-environments.

conda create --name deepicedrain --file environment-linux-64.lock
conda install --name deepicedrain --file environment-linux-64.lock

If you have a CUDA-capable GPU, you can also install the optional "cuda" packages to accelerate some calculations.

poetry install --extras cuda

Running jupyter lab

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

Related Projects

This work would not be possible without inspiration from the following cool open source projects! Go check them out if you have time.

• ATL11
• ICESAT-2 HackWeek
• icepyx