Skip to content

Latest commit

 

History

History
97 lines (72 loc) · 7.56 KB

README.md

File metadata and controls

97 lines (72 loc) · 7.56 KB

Setup

Install the dependencies in a virtual environment:

poetry config virtualenvs.create true
poetry config virtualenvs.in-project true
poetry install --with analytics
source .venv/bin/activate

We also need GDAL to handle some geospatial data processing. We are installing it separated from Poetry due to system-specific dependencies. For Ubuntu/Debian, install GDAL using your system's package manager:

sudo apt update
sudo apt install gdal-bin libgdal-dev

Confirm that GDAL was installed successfully:

gdalinfo --version

To use GDAL in Python, install the matching pip package:

sudo apt install python3.13-dev
pip install gdal==<your GDAL version>

Ensure GDAL is properly installed and can be imported:

from osgeo import gdal

Notebooks

Notebook Description Output file
Exploration notebooks
explore_dist_alert.ipynb Initial exploration of DIST-ALERT, which monitors disturbances in vegetation cover. -
explore_sufosat.ipynb Initial exploration of SUFOSAT's data. -
explore_sufosat_nat2000_ign.ipynb Initial exploration of SUFOSAT, IGN, and Natura 2000 data. -
Data preprocessing notebooks
prepare_sufosat_layer.ipynb Generates a FlatGeobuf vector file containing polygons of clear-cut areas detected in France since January 2024, derived from SUFOSAT data. The polygons include attributes such as start date, end date, duration, number of merged polygons, and area. data/sufosat/sufosat_clear_cuts_2024.fgb
prepare_natura2000.ipynb Work in progress Work in progress
prepare_ign_slope_raster.ipynb Generates a compressed GeoTIFF raster indicating terrain slopes >= 30% across France, based on the IGN BD ALTI digital elevation model. Pixels meeting this condition are assigned a value of 1. Also responsible for vectorizing the raster into a FlatGeobuf file. data/ign/bdalti25/slope_gte_30.tif, data/ign/bdalti25/slope_gte_30.fgb
prepare_cadastre.ipynb Generates FlatGeobuf files that delineate the boundaries of cities (with Etalab as the source) and departments (with OpenStreetMap as the source) across France. data/cadastre/cadastre_france_cities.fgb , data/cadastre/cadastre_france_departments.fgb
Statistic notebooks
stats_abusive_clear_cuts.ipynb Produces a FlatGeobuf file containing SUFOSAT clear-cut geometries (>= 0.5 ha) along with their intersection areas with Natura2000 zones and slopes >= 30%. We also include the associated department and city INSEE code. The filtering of abusive clear-cuts is left to consumers of the file. data/abusive_clear_cuts/abusive_clear_cuts_2024.fgb

How to download the preprocessed data

  • Get your Scaleway credentials from Vaultwarden.

  • Install the AWS CLI. You can also use the Scaleway CLI directly if you prefer. This guide covers the AWS CLI for now, as it is more widely used.

  • Create or update your ~/.aws/config file:

    [profile d4g-s13-brigade-coupes-rases]
region = fr-par
services = d4g-s13-brigade-coupes-rases


[services d4g-s13-brigade-coupes-rases]
s3 =
  endpoint_url = https://s3.fr-par.scw.cloud
s3api =
  endpoint_url = https://s3.fr-par.scw.cloud

  • Create or update your ~/.aws/credentials file:

    [d4g-s13-brigade-coupes-rases]
aws_access_key_id = YOUR_ACCESS_KEY
aws_secret_access_key = YOUR_SECRET_KEY

  • Run the following command to list the files that will be downloaded from S3:

    aws s3 ls s3://brigade-coupe-rase-s3/analytics/data/ --recursive --profile d4g-s13-brigade-coupes-rases

  • Run the following command from the root of your project to sync local files with the S3 data storage:

    aws s3 sync s3://brigade-coupe-rase-s3/analytics/data/ analytics/data/ --exact-timestamps --profile d4g-s13-brigade-coupes-rases