AWARDS

Automated Workflow for Analysis Ready Data from Satellite imagery

Logo created using Google Gemini

About

Automatic Workflow for Analysis Ready Data from Satellite imagery (AWARDS) to process a large amount of satellite data with high efficiency. It has been developing to support wall-to-wall satellite data processing and analysis ready data (ARD) generation.

License

This repository is licensed under the GNU General Public License v3.0. This means you are free to use, modify, and distribute the code, provided that:

• You include the original copyright notice and a copy of the license.
• You disclose the source code of your modifications.
• Any derivative works or larger projects incorporating this code are also licensed under GPL-3.0 (the "copyleft" requirement).

See the LICENSE file for full details.

Supported data resource

• Sentinel-2 data from Copernicus Data Space Ecosystem (CDSE)
• HLS data from NASA EarthData

Installation

# install via pip
# 1. create conda environment
conda create -n my_env -c conda-forge python=3.10 gdal rios rasterio geopandas fiona shapely pyproj mpi4py

# 2. activate
conda activate my_env

# 3. isntall
pip install awards

# 4. verify
python -c "import awards; print(awards.__version__)"

Development

Conda environment setup

# create the conda environment from yml configuration
# modify the $name in env.yml if needed
conda env create -f env.yml --name awards

Install Fmask modules

• The following modules to generate cloud masks from Fmask algorithms (Zhu et al., 2015) excluded from 'env.yml' file, and should be installed from source
  • py-fmask-ext
  • py-fmask-l2a

Clone these two repositories then run the following scripts to install the modules.

# 1. activate the conda environment
conda activate awards

# 1. install py-fmask-ext
# after cloning the repository, enter the directory and run the following installation:
python ./setup.py build && python ./setup.py install 

# 2. install py-fmask-l2a
# after cloning the repository, enter the directory and run the following installation:
python ./setup.py build && python ./setup.py install 

Install AWARDS in editable mode

pip install -e .

Register the python kernel

python -m ipykernel install --user --name awards --display-name "awards"

# when run the jupyter notebook tutorials, select the "awards" as the python kernel

Run a test

# when run the jupyter notebook tutorials, select the "awards" as the python kernel
# if the following works, then the awards module is installed correctly

import logging
from awards.common import yml_logging

yml_logging.setup_logging()
logger = logging.getLogger(__name__)

logger.info("test...")

Tutorials

Sample data

Sample data can be downloaded fron Zenodo:

• Sentinel-2 ARD and 10-m crop classification: AWARDS Sentinel-2 Analysis Ready Data (ARD) Sample Dataset V1
• HLS ARD: AWARDS Harmonized Landsat Sentinel-2 (HLS) Analysis Ready Data (ARD) Sample Dataset V1

Sentinel-2 ARD generation

• Jupyter notebooks: jupyter/s2
• Python scripts: script/s2
• Bash examples to run the python scripts: bash/s2

HLS ARD generation

• Jupyter notebooks: jupyter/hls
• Python scripts: script/hls
• Bash examples to run the python scripts: bash/hls

Crop classification using Sentinel-2 ARD and machine learning

• Jupyter notebooks: jupyter/crop_classification
• Python scripts: script/crop_classification
• Bash examples to run the python scripts: bash/crop_classification