U.S. DOE H2O Wave Hindcast Resource Characterization

A worked example of wave energy resource characterization at U.S. marine energy test sites, using the U.S. DOE WPTO High-Resolution Wave Hindcast. Data access goes through the us-marine-energy-resource package. The analysis covers aggregate statistics, monthly climatology, joint probability distributions, extreme sea state contours, and multi-site comparison, using MHKiT for the wave calculations.

The source of truth is the Quarto document us_doe_h2o_wave_hindcast_resource_characterization.qmd. The rendered notebook and PDF are committed alongside it. The notebook also runs on Kaggle with a companion dataset that provides pre-downloaded data.

Getting started

Requires uv and Quarto, or open the repository in GitHub Codespaces, which installs both and runs make sync for you. The PDF is produced by Typst, which ships inside Quarto, so no LaTeX install is needed anywhere.

make sync            # build the environment and register the Jupyter kernel
make check-toolchain # confirm Quarto, Typst and the kernel are wired up
make render          # render the notebook and PDF

Run make help for the full list of targets. The first render downloads the sea state records and caches them; renders after that read from disk. The cache lives under ~/.mer_wave_cache, or wherever MER_WAVE_CACHE_DIR points — in a Codespace that is /workspaces/.mer_wave_cache, which survives a container rebuild so the download happens once. The size depends on which points are compared and over how many years, so the document measures it rather than guessing: see "What the Records Cost", which reports the per-point download and writes .cache/download_statistics.json. For the four default points over five years it comes to roughly 585 MB, most of it raw S3 chunks under s3_chunks/ that distil into the much smaller per-point records (US_*/) later renders read. Points in the Atlantic domain cost far more, because that domain bundles about twenty times as many nodes into each chunk.

No account or API key is needed. The document reads the published hindcast files on S3 anonymously. An API key from the NLR Developer Network enables the API backend for larger queries; see .env.example.

Data

Source data is the WPTO U.S. Wave Hindcast, produced by the U.S. Department of Energy Water Power Technologies Office and distributed through the AWS Open Data registry at s3://wpto-pds-us-wave. It is a U.S. Government work in the public domain.

License

BSD 3-Clause License. Copyright (c) 2026, Alliance for Energy Innovation, LLC under the terms of Contract DE-AC36-08GO28308. The U.S. Government retains certain rights in this software. See LICENSE.