CoastPy

Python tools for cloud-native coastal analytics.

Installation

You can install coastpy with pip in a Python environment where GDAL and pyproj are already installed.

pip install coastpy

However, if you start from scratch, it's probably easier to install with conda:

conda env create -f https://raw.githubusercontent.com/TUDelft-CITG/coastpy/refs/heads/main/environment.yaml

Data

Data Access Policy Update (Effective August 31, 2025)*

With the conclusion of the CoCliCo project and its associated cloud storage funding, all datasets hosted in the coclico Azure Storage account have been transitioned from public to private access.

To access the data, you will now need a Shared Access Signature (SAS) token. Tokens are available upon reasonable request from Deltares.

The data remains directly accessible from the cloud using the same tools and methods. Please see the tutorials and analytics sections for examples of how to use tools like Python and DuckDB with the required SAS token.

Global Coastal Transect System (GCTS)

Cross-shore coastal transects are essential to coastal monitoring, offering a consistent reference line to measure coastal change, while providing a robust foundation to map coastal characteristics and derive coastal statistics thereof.

The Global Coastal Transect System consists of more than 11 million cross-shore coastal transects uniformly spaced at 100-m intervals alongshore, for all OpenStreetMap coastlines that are longer than 5 kilometers.

# Download all transects located in the United States.
duckdb -c "COPY (SELECT * FROM 'az://coclico.blob.core.windows.net/gcts/release/2024-08-02/*.parquet' AS gcts WHERE gcts.country = 'US') TO 'United_States.parquet' (FORMAT 'PARQUET')"

# Download transects by bounding box.
duckdb -c "COPY (SELECT * FROM 'az://coclico.blob.core.windows.net/gcts/release/2024-08-02/*.parquet' AS gcts WHERE bbox.xmin <= 14.58 AND bbox.ymin <= -22.77 AND bbox.xmax >= 14.27 AND bbox.ymax >= -23.57) TO area_of_interest.parquet (FORMAT 'PARQUET')"

# Or, download the data in bulk using AZ CLI
az storage blob download-batch \
    --destination "./" \
    --source "gcts" \
    --pattern "release/2024-08-02/*.parquet" \
    --account-name coclico

Coastal Grid

The Coastal Grid dataset provides a global tiling system for coastal analytics. It supports scalable data processing workflows by offering coastal tiles at varying zoom levels (5, 6, 7, 8, 9, 10) and buffer sizes (500 m, 1000 m, 2000 m, 5000 m, 10000 m, 15000 m).

Installation & usage instructions

Better installation and usage instructions will come when we build the documentation. For now, to run tutorials, analytics or scripts, you could consider proceeding as follows.

Installation

1. Install Git or GitHub Desktop
2. Clone CoastPy
3. Install Miniforge
4. Open a Miniforge prompt (finder/spotlight)
5. Run the following commands:

# Update conda if you already had miniforge installed
conda update --all --yes

# Create the software environment
conda env create -f https://raw.githubusercontent.com/TUDelft-CITG/coastpy/refs/heads/main/environment.yaml

Usage

1. Open Miniforge
2. Change to the directory where CoastPy was cloned by using cd /path/to/coastpy
3. Activate the software environment by mamba activate coastal
4. Launch Jupyter lab by jupyter lab
5. Navigate to the tutorials folder in Jupyter lab

Citation:

@article{CALKOEN2025106257,
  title     = {Enabling coastal analytics at planetary scale},
  journal   = {Environmental Modelling & Software},
  volume    = {183},
  pages     = {106257},
  year      = {2025},
  issn      = {1364-8152},
  doi       = {https://doi.org/10.1016/j.envsoft.2024.106257},
  url       = {https://www.sciencedirect.com/science/article/pii/S1364815224003189},
}

Contributing

Interested in contributing? Check out the contributing guidelines. Please note that this project is released with a Code of Conduct. By contributing to this project, you agree to abide by its terms.

License

coastpy was created by Floris Calkoen. The software is licensed under the terms of the MIT license. Data licenses are typically CC-BY-4.0, and can be found in the respective STAC collection.