1 file.-- This software is provided "as is", without warranty of any kind, express or implied, including but not limited to the warranties of merchantability, fitness for a particular purpose, and noninfringement. In no event shall the authors or copyright holders be liable for any claim, damages, or other liability, whether in an action of contract, tort, or otherwise, arising from, out of, or in connection with the software or the use or other dealings in the software, This repository is part of the HF-EOLUS project and constitutes the first step towards its Task 2: obtaining wind fields from oceanic HF-radar data. Task 2 aims to use, evaluate, and develop the extraction of wind data from HF-radar data. In the HF-EOLUS methodology, the first step in wind extraction is obtaining radial metrics from the spectra collected by HF-Radar stations. This repository applies a reproducible, containerized workflow to process high-frequency (HF) radar spectra from INTECMAR’s VILA and PRIO HF-Radar stations located on the Galician shelf (NW-Spain). The primary output is LLUV files containing radial metrics derived from CODAR SeaSonde spectral data. For more details, please visit the repository page on GitHub.

The applied workflow features period-specific configurations, automatic manifest generation, and scalable AWS batch processing (Lambda & S3 Batch Operations) for each station, as detailed in [1]. The underlying analysis of the HF radar spectra is carried out by the SeaSondeR R package [2]. Finally, the processed LLUV datasets are publicly available on Zenodo (PRIO [3] VILA [4]), This work has been funded by the HF-EOLUS project (TED2021-129551B-I00), financed by MICIU/AEI /10.13039/501100011033 and by the European Union NextGenerationEU/PRTR - BDNS 598843 - Component 17 - Investment I3, Peer reviewed

1 file.-- This software is provided "as is", without warranty of any kind, express or implied, including but not limited to the warranties of merchantability, fitness for a particular purpose, and noninfringement. In no event shall the authors or copyright holders be liable for any claim, damages, or other liability, whether in an action of contract, tort, or otherwise, arising from, out of, or in connection with the software or the use or other dealings in the software, This repository implements a complete workflow for batch processing HF-Radar spectra files using the SeaSondeR R package on AWS. For more details, please visit SeaSondeRAWSLambdaDocer on GitHub, This work has been funded by the HF-EOLUS project (TED2021-129551B-I00), financed by MICIU/AEI /10.13039/501100011033 and by the European Union NextGenerationEU/PRTR - BDNS 598843 - Component 17 - Investment I3, Peer reviewed

The goal of SeaSondeR is to provide a comprehensive set of tools for processing and analyzing data from the SeaSonde(R) High-Frequency Radar (HF-Radar) instrument. This package is intended to facilitate the creation of radial metrics files from spectra data. SeaSondeR is an R package developed as an open-source alternative to proprietary tools for processing HF-Radar spectra and generating Radial Metrics. SeaSondeR integrates all key processing steps into a single workflow, reading and processing both spectrum (CS) files and SeaSonde(R) antenna patterns, identifying the first-order spectral region according to CODAR methodology, and applying the MUSIC algorithm to estimate signal arrival directions. Drawing on technical manuals, patent literature, and prior scientific work, SeaSondeR is designed for both experimental and synthetic data, facilitating cloud-based analyses of extensive HF-Radar datasets while promoting free software solutions in operational oceanography and coastal monitoring, This software is provided "as is", without warranty of any kind, express or implied, including but not limited to the warranties of merchantability, fitness for a particular purpose, and noninfringement. In no event shall the authors or copyright holders be liable for any claim, damages, or other liability, whether in an action of contract, tort, or otherwise, arising from, out of, or in connection with the software or the use or other dealings in the software, R package published in CRAN to read CODAR's SeaSonde High-Frequency Radar spectra files, compute radial metrics, and generate plots for spectra and antenna pattern data. For more information ,please visit the package page on CRAN and the package pages on GitHub, This work has been funded by the HF-EOLUS project (TED2021-129551B-I00), financed by MICIU/AEI /10.13039/501100011033 and by the European Union NextGenerationEU/PRTR - BDNS 598843 - Component 17 - Investment I3, Peer reviewed

1 file, HF‑EOLUS Geospatial Processing Tools is a set of CLI tools for transforming and analyzing large sets of point geospatial data into cloud-native formats, generating GeoParquet and STAC metadata. It offers a step-by-step pipeline to go from tabular data to analysis-ready assets (grids, node mapping, aggregation, and cataloging). More details on GitHub: https://github.com/GOFUVI/hf_eolus_geo_tools, This work has been funded by the HF-EOLUS project (TED2021-129551B-I00), financed by MICIU/AEI /10.13039/501100011033 and by the European Union NextGenerationEU/PRTR - BDNS 598843 - Component 17 - Investment I3, Peer reviewed

[Workflow overview]
1. Cloud-native ingestion – Parameterised Lambda + Step Functions workflow that downloads MeteoGalicia grids, applies optional spatial filters and stores structured Parquet assets on S3 along with metadata sidecars.
2. Interpolation engine – Containerised R jobs executed via AWS Batch that build refined grids, tune interpolation hyperparameters, compute diagnostics (bias, RSR, LOOCV) and emit GeoParquet layers plus supporting plots.
3. Catalog publishing – Dockerised STAC builder that packages GeoParquet assets, metadata and visual diagnostics into STAC collections/items ready for downstream discovery platforms.
4. Reproducible automation – A single run_pipeline.sh driver wires together ingestion, processing and catalog publication with a consistent set of environment variables, enabling portable redeployments and verifiable outputs without manual intervention., [Capabilities]
- Scalable interpolation: Supports inverse distance weighting, regression kriging and universal kriging with configurable search radii, neighbourhood sizes and cross-validation folds.
- Diagnostics-first design: Automatically tracks bias, RSR, LOOCV scores and quadrant consistency, persisting both numeric summaries and ready-to-publish plots.
- GeoParquet + STAC outputs: Generates spatially indexed GeoParquet layers alongside STAC metadata so datasets can be catalogued, discovered and downloaded with standard tooling.
- Automation hooks: CLI scripts and infrastructure definitions expose every step for reuse inside CI/CD pipelines, AWS Batch arrays or ad-hoc container executions., [Software availability]
The full source code, infrastructure definitions and documentation are available at https://github.com/GOFUVI/hf-eolus-interpolation-toolkit, The HF-EOLUS Wind Interpolation Toolkit scales the methodology originally developed for MeteoGalicia WRF outputs so that end-to-end wind field interpolation can be orchestrated reproducibly on AWS. The toolkit automates ingestion of NetCDF model data, conversion to hourly Parquet partitions, spatial interpolation (IDW, regression kriging and universal kriging), diagnostic scoring and publication of GeoParquet deliverables via STAC catalogs, ensuring reproducibility from raw inputs to discoverable data products, This work has been funded by the HF-EOLUS project (TED2021-129551B-I00), financed by MICIU/AEI /10.13039/501100011033 and by the European Union NextGenerationEU/PRTR - BDNS 598843 - Component 17 - Investment I3, Peer reviewed

[Workflow overview]
1. Cloud-native ingestion – Parameterised Lambda + Step Functions workflow that downloads MeteoGalicia grids, applies optional spatial filters and stores structured Parquet assets on S3 along with metadata sidecars. 2. Interpolation engine – Containerised R jobs executed via AWS Batch that build refined grids, tune interpolation hyperparameters, compute diagnostics (bias, RSR, LOOCV) and emit GeoParquet layers plus supporting plots. 3. Catalog publishing – Dockerised STAC builder that packages GeoParquet assets, metadata and visual diagnostics into STAC collections/items ready for downstream discovery platforms. 4. Reproducible automation – A single run_pipeline.sh driver wires together ingestion, processing and catalog publication with a consistent set of environment variables, enabling portable redeployments and verifiable outputs without manual intervention., [Capabilities]
- Scalable interpolation: Supports inverse distance weighting, regression kriging and universal kriging with configurable search radii, neighbourhood sizes and cross-validation folds. - Diagnostics-first design: Automatically tracks bias, RSR, LOOCV scores and quadrant consistency, persisting both numeric summaries and ready-to-publish plots. - GeoParquet + STAC outputs: Generates spatially indexed GeoParquet layers alongside STAC metadata so datasets can be catalogued, discovered and downloaded with standard tooling. - Automation hooks: CLI scripts and infrastructure definitions expose every step for reuse inside CI/CD pipelines, AWS Batch arrays or ad-hoc container executions., [What's new]
- Incremental STAC publishing: scripts/11-build_stac_catalog.py and the Docker wrapper now support --incremental, preserving existing assets/Items while appending only new GeoParquet partitions.
- Year-by-year STAC builds: run_build_stac_catalog.sh/11-build_stac_catalog.py add --by-year, optional --years and --limit-years to sync/process only selected years and load just the corresponding Items; the wrapper also raises the Docker nofile ulimit to avoid descriptor exhaustion on large catalogs.
- STAC subsetting utility: New scripts/12-subset_stac_nodes.py (with run_subset_stac_nodes.sh) filters catalogs by node_id and/or polygon and emits a derived STAC plus a single Parquet/GeoParquet.
- Faster subsets on incomplete catalogs: 12-subset_stac_nodes.py accepts --skip-source-items to avoid enumerating every source Item when only a template is needed, enabling subsets even if the source catalog is incomplete.
- Hardened ingestion/interpolation scripts: scripts/08-run_interpolation.sh improves NetCDF download robustness and parameter handling; run_pipeline.sh propagate the new flags and AWS profile/region wiring.
- Wrapper enhancements: run_build_stac_catalog.sh and the subset wrapper reuse a shared Docker image, sync S3 prefixes automatically when needed, and expose AWS profiles/regions consistently., [Software availability]
The full source code, infrastructure definitions and documentation are available at https://github.com/GOFUVI/hf-eolus-interpolation-toolki, The HF-EOLUS Wind Interpolation Toolkit scales the methodology originally developed for MeteoGalicia WRF outputs so that end-to-end wind field interpolation can be orchestrated reproducibly on AWS. The toolkit automates ingestion of NetCDF model data, conversion to hourly Parquet partitions, spatial interpolation (IDW, regression kriging and universal kriging), diagnostic scoring and publication of GeoParquet deliverables via STAC catalogs, ensuring reproducibility from raw inputs to discoverable data products, This work has been funded by the HF-EOLUS project (TED2021-129551B-I00), financed by MICIU/AEI /10.13039/501100011033 and by the European Union NextGenerationEU/PRTR - BDNS 598843 - Component 17 - Investment I3, Peer reviewed

1 file, This documentation describes how the HF-EOLUS project stores large volumes of high-frequency radar Radial Metrics data in the GeoParquet format and organizes them with STAC (SpatioTemporal Asset Catalog) metadata standards. It serves as a guide for technical users interested in data storage and cataloging strategies for geospatial datasets. While HF-Radar observations are the primary focus, the same principles apply to meteorological model outputs processed in the project. Key benefits include efficient, cloud-optimized storage and internationally standardized metadata for interoperability. For more details, please visit the specifications repository on GitHub, This work has been funded by the HF-EOLUS project (TED2021-129551B-I00), financed by MICIU/AEI /10.13039/501100011033 and by the European Union NextGenerationEU/PRTR - BDNS 598843 - Component 17 - Investment I3, Peer reviewed

This release completes the documentation for HF-EOLUS. Every external workflow now has a clear inventory that lists its GeoParquet deliverables, partition layouts, key columns, STAC extensions, and Zenodo DOIs, giving readers a single place to trace each processing line end to end. README and the overview have been expanded so all seven processing lines (plus the MeteoGalicia interpolation branch) are described consistently, with the interpolation workflow explicitly called out as a parallel branch that is not yet wired into the wind-resource chain. The GeoParquet and STAC specifications now open with shared fundamentals and then dive into per-workflow profiles detailing schemas, geometry choices, asset naming, and catalog hierarchies, making it easier to see how the standards are applied in practice. To tie everything together, a cross-document validation checklist verifies that every workflow is mentioned in README/overview and captured in both specification files, ensuring complete traceability, This documentation describes how the HF-EOLUS project stores large volumes of high-frequency radar Radial Metrics data in the GeoParquet format and organizes them with STAC (SpatioTemporal Asset Catalog) metadata standards. It serves as a guide for technical users interested in data storage and cataloging strategies for geospatial datasets. While HF-Radar observations are the primary focus, the same principles apply to meteorological model outputs processed in the project. Key benefits include efficient, cloud-optimized storage and internationally standardized metadata for interoperability. For more details, please visit the specifications repository on GitHub, This work has been funded by the HF-EOLUS project (TED2021-129551B-I00), financed by MICIU/AEI /10.13039/501100011033 and by the European Union NextGenerationEU/PRTR - BDNS 598843 - Component 17 - Investment I3, Peer reviewed

This release includes links to online browsers and viewers for inspecting the catalogs, This documentation describes how the HF-EOLUS project stores large volumes of high-frequency radar Radial Metrics data in the GeoParquet format and organizes them with STAC (SpatioTemporal Asset Catalog) metadata standards. It serves as a guide for technical users interested in data storage and cataloging strategies for geospatial datasets. While HF-Radar observations are the primary focus, the same principles apply to meteorological model outputs processed in the project. Key benefits include efficient, cloud-optimized storage and internationally standardized metadata for interoperability. For more details, please visit the specifications repository on GitHub, This work has been funded by the HF-EOLUS project (TED2021-129551B-I00), financed by MICIU/AEI /10.13039/501100011033 and by the European Union NextGenerationEU/PRTR - BDNS 598843 - Component 17 - Investment I3, Peer reviewed

1 file, Automated pipeline to acquire Sentinel-1 SAR OCN/OWI data, convert it into partitioned GeoParquet, and generate a STAC catalog ready for local or cloud analytics. Includes Dockerized scripts (download_sar.sh, ingest_sar.sh) for reproducible download and ingestion, with optional publishing to Amazon S3 and registration in AWS Athena for SQL querying. For setup, usage, and specifications, please refer to the repository: https://github.com/GOFUVI/hf_eolus_sar_ingestion, This work has been funded by the HF-EOLUS project (TED2021-129551B-I00), financed by MICIU/AEI /10.13039/501100011033 and by the European Union NextGenerationEU/PRTR - BDNS 598843 - Component 17 - Investment I3, Peer reviewed

1 file, Automated pipeline to acquire Sentinel-1 SAR OCN/OWI data, convert it into partitioned GeoParquet, and generate a STAC catalog ready for local or cloud analytics. Includes Dockerized scripts (download_sar.sh, ingest_sar.sh) for reproducible download and ingestion, with optional publishing to Amazon S3 and registration in AWS Athena for SQL querying. For setup, usage, and specifications, please refer to the repository: https://github.com/GOFUVI/hf_eolus_sar_ingestion, This work has been funded by the HF-EOLUS project (TED2021-129551B-I00), financed by MICIU/AEI /10.13039/501100011033 and by the European Union NextGenerationEU/PRTR - BDNS 598843 - Component 17 - Investment I3, Peer reviewed

9 files.-- Access to catalogs.zip https://doi.org/10.5281/zenodo.17131228, HF-EOLUS project task 2 objective is to achieve HF-Radar Wind Inversion through Artificial Neural Networks training. This Zenodo record documents the methodology and results executed in pursuit of that goal. Readers seeking an extended methodological and results compendium can consult README.md, which delineates each pipeline, its configuration, and the evaluation metrics in detail, This work has been funded by the HF-EOLUS project (TED2021-129551B-I00), financed by MICIU/AEI /10.13039/501100011033 and by the European Union NextGenerationEU/PRTR - BDNS 598843 - Component 17 - Investment I3, Peer reviewed

9 files.-- Access to catalogs.zip https://doi.org/10.5281/zenodo.17464583, A buoy wind height correction stage has been incorporated through the dockerised wrapper scripts/aggregation/apply_buoy_wind_height_correction.sh, which applies the neutral logarithmic profile to translate the Vilano buoy winds from their native 3 m measurement height to the 10 m reference used across the SAR products. All affected artificial neural network checkpoints were retrained with the corrected buoy series, and the resulting metrics remain within the previously reported uncertainty envelope—no substantial shifts were detected in RMSE, directional error, or range-classification accuracy, This record documents the methodology and results executed in pursuit of that goal. Readers seeking an extended methodological and results compendium can consult README.md, which delineates each pipeline, its configuration, and the evaluation metrics in detail, This work has been funded by the HF-EOLUS project (TED2021-129551B-I00), financed by MICIU/AEI /10.13039/501100011033 and by the European Union NextGenerationEU/PRTR - BDNS 598843 - Component 17 - Investment I3, Peer reviewed

Data preparation stages pivot the radar spectra around Bragg peaks, validate the geometrical consistency of each station, enrich the features with bearings, distances and maintenance metadata, and standardise schema variations across radar sites. Stratified partitions, cross-domain concatenation, and maintenance-aware normalisation ensure that model training, validation and inference rely on reproducible datasets that capture both spatial variability and temporal stability.

Training routines span hyper-parameter optimisation, final non-cross-validated retraining, and multiple fine-tuning strategies—plain adaptation, L2-SP anchoring, and knowledge distillation—to explore cross-domain transfer between the spatially rich SAR corpus and the buoy's high-fidelity point measurements. Inference jobs cover grid-wide predictions, domain-specific test sets, and maintenance-interval diagnostics, delivering quantitative metrics and provenance-rich artefacts that can be catalogued alongside the input datasets. By packaging the entire pipeline, the repository enables systematic experimentation with HF-radar wind inversion models while providing the documentation necessary for reproducibility and scientific auditing, The HF-EOLUS toolkit consolidates all command-line pipelines and artefact definitions required to retrieve near-surface wind speed and direction from HF-radar observations via artificial neural networks. It couples radar-derived backscatter statistics with two supervisory references: Sentinel-1 Level-2 OCN products aggregated on regular spatial grids and collocated in-situ winds measured by an oceanographic buoy embedded in the same footprint. The workflow respects the frequency-dependent radar operating band by combining gated regression heads with a complementary range-classification branch, thereby preserving the physical constraints of the inversion problem. For comprehensive documentation and reproducibility artefacts, consult the project repository on GitHub, This work has been funded by the HF-EOLUS project (TED2021-129551B-I00), financed by MICIU/AEI /10.13039/501100011033 and by the European Union NextGenerationEU/PRTR - BDNS 598843 - Component 17 - Investment I3, Peer reviewed

Data preparation stages pivot the radar spectra around Bragg peaks, validate the geometrical consistency of each station, enrich the features with bearings, distances and maintenance metadata, and standardise schema variations across radar sites. Stratified partitions, cross-domain concatenation, and maintenance-aware normalisation ensure that model training, validation and inference rely on reproducible datasets that capture both spatial variability and temporal stability.

Training routines span hyper-parameter optimisation, final non-cross-validated retraining, and multiple fine-tuning strategies—plain adaptation, L2-SP anchoring, and knowledge distillation—to explore cross-domain transfer between the spatially rich SAR corpus and the buoy's high-fidelity point measurements. Inference jobs cover grid-wide predictions, domain-specific test sets, and maintenance-interval diagnostics, delivering quantitative metrics and provenance-rich artefacts that can be catalogued alongside the input datasets. By packaging the entire pipeline, the repository enables systematic experimentation with HF-radar wind inversion models while providing the documentation necessary for reproducibility and scientific auditing, [What's New]
A buoy wind height correction utility (scripts/aggregation/apply_buoy_wind_height_correction.sh) now accompanies the pipeline. The helper encapsulates the neutral logarithmic wind-profile adjustment inside a containerised execution environment, accepting configurable source and target heights so deployments can harmonise buoy winds to any reference level without manual scripting. Documentation and the illustrative data-preparation pipeline reflect this addition, ensuring that downstream partitions ingest height-standardised buoy tables while preserving the raw measurements for provenance and auditing workflows, The HF-EOLUS toolkit consolidates all command-line pipelines and artefact definitions required to retrieve near-surface wind speed and direction from HF-radar observations via artificial neural networks. It couples radar-derived backscatter statistics with two supervisory references: Sentinel-1 Level-2 OCN products aggregated on regular spatial grids and collocated in-situ winds measured by an oceanographic buoy embedded in the same footprint. The workflow respects the frequency-dependent radar operating band by combining gated regression heads with a complementary range-classification branch, thereby preserving the physical constraints of the inversion problem. For comprehensive documentation and reproducibility artefacts, consult the project repository on GitHub (https://github.com/GOFUVI/hf-wind-inversion), This work has been funded by the HF-EOLUS project (TED2021-129551B-I00), financed by MICIU/AEI /10.13039/501100011033 and by the European Union NextGenerationEU/PRTR - BDNS 598843 - Component 17 - Investment I3, Peer reviewed

1 file, The HF EOLUS Wind Resource Toolkit packages a reproducible workflow for deriving offshore wind-resource metrics from HF-radar neural-network inference products. It bundles a CLI, reusable Python modules (hf_wind_resource), configuration templates, and Docker-based helpers so that researchers can transform catalogued GeoParquet inputs into power-ready deliverables without rewriting analytics or orchestration code. The toolkit is dataset-agnostic: any STAC snapshot conforming to the documented ANN schema can be processed after updating the catalogue configuration files under config/. Please check the whole repository on GitHub, This work has been funded by the HF-EOLUS project (TED2021-129551B-I00), financed by MICIU/AEI /10.13039/501100011033 and by the European Union NextGenerationEU/PRTR - BDNS 598843 - Component 17 - Investment I3, Peer reviewed

1 file, The HF EOLUS Wind Resource Toolkit packages a reproducible workflow for deriving offshore wind-resource metrics from HF-radar neural-network inference products. It bundles a CLI, reusable Python modules (hf_wind_resource), configuration templates, and Docker-based helpers so that researchers can transform catalogued GeoParquet inputs into power-ready deliverables without rewriting analytics or orchestration code. The toolkit is dataset-agnostic: any STAC snapshot conforming to the documented ANN schema can be processed after updating the catalogue configuration files under config/, This work has been funded by the HF-EOLUS project (TED2021-129551B-I00), financed by MICIU/AEI /10.13039/501100011033 and by the European Union NextGenerationEU/PRTR - BDNS 598843 - Component 17 - Investment I3, Peer reviewed

1 file, HF-EOLUS Puertos del Estado Buoy Ingestion Pipeline converts PdE buoy wind observations into analysis-ready, cloud-optimized GeoParquet with accompanying STAC metadata. It consolidates per-buoy time series (wind speed and direction) into a single GeoParquet file per buoy and publishes a static STAC catalog for easy discovery and interoperability. The approach leverages open standards to enable efficient querying and integration with common geospatial and data science tools. For more details, please see the repository, This work has been funded by the HF-EOLUS project (TED2021-129551B-I00), financed by MICIU/AEI /10.13039/501100011033 and by the European Union NextGenerationEU/PRTR - BDNS 598843 - Component 17 - Investment I3, Peer reviewed

1 file, Tools to ingest and catalog HF‑Radar “Radial Metrics” (CODAR LLUV) into analysis‑ready GeoParquet. The pipeline converts LLUV files into three GeoParquet v1.1 datasets on AWS S3 (echo‑level radial metrics, header metadata, and range summaries), integrates with AWS Glue for distributed parsing and AWS Athena for SQL querying, and optionally generates station‑level STAC Collections/Items linking to the Parquet assets. See the repository README for setup, usage, and technical details, This work has been funded by the HF-EOLUS project (TED2021-129551B-I00), financed by MICIU/AEI /10.13039/501100011033 and by the European Union NextGenerationEU/PRTR - BDNS 598843 - Component 17 - Investment I3, Peer reviewed

1 file, S1OCN is an R package for browsing, downloading, and extracting data from Sentinel‑1 Level‑2 OCN (Ocean) products. For more details, please visit S1OCN on GitHub, This work has been funded by the HF-EOLUS project (TED2021-129551B-I00), financed by MICIU/AEI /10.13039/501100011033 and by the European Union NextGenerationEU/PRTR - BDNS 598843 - Component 17 - Investment I3, Peer reviewed