Vineyard Missing-Area Analysis

A spatial database linking vineyard parcels in France to their AOC (Appellation d'Origine Contrôlée) regulations and to per-parcel missing-vine estimates from a LiDAR-based detection pipeline.

Purpose

The main goal of this project is to create a queryable spatial database that can be used to analyze the results of the vinDAR pipeline. This is done by bringing together three otherwise disconnected sources of information about French vineyards:

• Per-parcel missing-area results from the vinDAR analysis
• EU Protected Designation of Origin (PDO) polygons defining the extent of each appellation
• Hand-compiled AOC regulations describing each appellation's designation level, row/plant spacing rules, and area limits

Together, these allow parcel level missing area estimates to be evaluated against the regulatory context of the appellation each parcel falls within.

Repository structure

├── README.md
├── requirements.txt           # Python dependencies
├── database/
│   └── schema_build.sql       # Builds the full database: tables, constraints, etc.
├── data/                      # Input data (not fully committed — see "Data access")
│   ├── parcel_results.csv     # Output from vinDAR repository
│   ├── EU_PDO.gpkg            # EU PDO (AOC) geometries
│   ├── aoc_regulations.csv    # Hand-compiled AOC regulatory attributes
│   └── AllVDR.gpkg            # Vineyard parcel geometries and attributes (restricted)
└── vindar-database.ipynb      # Runs schema_build.sql, returns a summary DataFrame

Setup

This project uses Python and DuckDB with the spatial extension.

1. Clone the repository:

   git clone <repo-url>
   cd <repo-name>

2. Install dependencies (Python 3.10+ recommended):

   pip install -r requirements.txt

3. Place the required data files in the data/ folder (see Data access below).
4. Open and run vindar-database.ipynb. It runs the SQL script to build the database and returns a summary table.

Data access

Four input files are expected in the data/ folder. Two are included in the repository, one is publicly downloadable, and one is restricted.

File	Source	Status
parcel_results.csv	Output from the vinDAR pipeline	Included in repo
aoc_regulations.csv	Hand-compiled from public AOC specifications	Included in repo
EU_PDO.gpkg	figshare: Candiago et al. 2024	Public — download and place in data/
AllVDR.gpkg	Provided by project partners	Restricted — see alternative below

Public alternative for AllVDR.gpkg: Vineyard parcel geometries can be obtained from the French Registre Parcellaire Graphique (RPG), filtering for vineyard land-use codes. The schema and column names in schema_build.sql assume the structure of AllVDR.gpkg, so minor edits to the query in the SQL script are likely to be required when substituting RPG data.

References & acknowledgements

• vinDAR-MEDS — companion LiDAR-based detection pipeline. https://github.com/vinDAR-MEDS
• EU PDO geometries — Candiago, S., Tscholl, S., Bassani, L., Fraga, H., & Egarter Vigl, L. (2024). Quality wines in Italy and France: a dataset of protected designation of origin specifications [Data set]. figshare. https://doi.org/10.6084/m9.figshare.25393261.v2
• LiDAR source data (underlying the vinDAR detection pipeline) — Institut national de l'information géographique et forestière. (2026). LiDAR HD Point Clouds [Data set]. (Original work published 2024). http://data.europa.eu/88u/dataset/ignf_nuages-de-points-lidar-hd
• DuckDB and the spatial extension. https://duckdb.org/
• Developed for EDS 213: Databases and Data Management, Bren School of Environmental Science & Management, UC Santa Barbara.

License

This project is licensed under CC BY 4.0.