pyvpp 1.0.0

Python package to download and process vegetation products from WEkEO (HR-VPP) and CDSE (Sentinel-2)

PyVPP

Python package to download and preprocess vegetation products from Copernicus services:

• HR-VPP (High Resolution Vegetation Phenology and Productivity) via WEkEO
• Sentinel-2 L2A raw data via CDSE (Copernicus Data Space Ecosystem)

⚠️ IMPORTANT - VPP_Index Dataset Currently Unavailable (January 2025)

The VPP_Index dataset (EO:EEA:DAT:CLMS_HRVPP_VI) currently returns 404 errors on WEkEO.

This dataset, which contained daily vegetation indices (LAI, FAPAR, NDVI, PPI), appears to have been discontinued or moved since 2024.

✅ Datasets that ARE working:

• VPP_ST (Seasonal Trajectories): Contains PPI, QFLAG
• VPP_Pheno (Phenology & Productivity): Contains SOSD, MAXD, EOSD, LENGTH, AMPL, TPROD, etc.

🔧 If you needed VPP_Index:

• For PPI → Use VPP_ST
• For phenology parameters → Use VPP_Pheno
• For LAI, FAPAR, NDVI → Currently not available via PyVPP (see alternatives below)

See DATASETS_AVAILABLE.md for more details.

---

Installation

pip install pyvpp

Quick Start

PyVPP supports two main workflows:

A. HR-VPP Products (WEkEO)

1. Configure your WEkEO credentials

import pyvpp

# Create .hdarc file with your credentials
pyvpp.create_hdarc("your_wekeo_username", "your_wekeo_password")

2. Download HR-VPP data

Phenology parameters (RECOMMENDED - WORKING)

import pyvpp

# Download phenology parameters (annual products)
downloader = pyvpp.wekeo_download(
    dataset='VPP_Pheno',  # ✅ Working
    shape='path/to/your/area.shp',  # or DEIMS ID
    dates=['2020-01-01', '2020-12-31'],
    products=['SOSD', 'MAXD', 'EOSD', 'LENGTH']
)

downloader.run()

Seasonal Trajectories - PPI (RECOMMENDED - WORKING)

import pyvpp

# Download PPI (Plant Phenology Index) every 10 days
downloader = pyvpp.wekeo_download(
    dataset='VPP_ST',  # ✅ Working
    shape='path/to/your/area.shp',
    dates=['2020-01-01', '2020-12-31'],
    products=['PPI', 'QFLAG']
)

downloader.run()

Vegetation Indices (⚠️ CURRENTLY NOT AVAILABLE)

import pyvpp

# ⚠️ WARNING: This dataset is currently returning 404 errors
# VPP_Index (LAI, FAPAR, NDVI) is not available as of January 2025
downloader = pyvpp.wekeo_download(
    dataset='VPP_Index',  # ⚠️ Returns 404 error
    shape='path/to/your/area.shp',
    dates=['2020-01-01', '2020-12-31'],
    products=['LAI', 'FAPAR']  # Not available
)

# This will fail with: 404 Client Error

---

B. Sentinel-2 Raw Data (CDSE)

New in v0.4.0: Download Sentinel-2 L2A products directly from CDSE.

1. Configure CDSE credentials

You have three options:

Option 1: Configuration file (recommended)

mkdir -p ~/.pyvpp
cat > ~/.pyvpp/config.toml << EOF
[cdse]
user = "[e-mail]"
password = "your_password"
EOF

Option 2: Environment variables

export CDSE_USER="[e-mail]"
export CDSE_PASSWORD="your_password"

Option 3: Direct in code

from pyvpp.cdse import load_cdse_credentials
# Credentials passed directly to CDSEDownload

Register at: https://dataspace.copernicus.eu/

2. Download Sentinel-2 data

L2A - Bottom of Atmosphere (Default - Atmospherically Corrected)

from pyvpp import CDSEDownload

# Download Sentinel-2 L2A products (recommended for most use cases)
downloader = CDSEDownload(
    shape='path/to/your/area.shp',  # or DEIMS ID
    dates=('2023-06-01', '2023-06-30'),
    bands=['B04', 'B08'],  # Red, NIR
    outdir='sentinel2_data',
    processing_level='L2A',  # Default: atmospherically corrected
    utm_zone='30T'  # Optional: filter by UTM zone
)

# Run full pipeline: search -> download -> mosaic -> clip
downloader.run()

L1C - Top of Atmosphere (No Atmospheric Correction - for ACOLITE, etc.)

from pyvpp import CDSEDownload

# Download Sentinel-2 L1C products (TOA reflectance)
# Useful for custom atmospheric correction (ACOLITE, Sen2Cor, etc.)
downloader = CDSEDownload(
    shape='path/to/your/area.shp',
    dates=('2023-06-01', '2023-06-30'),
    bands=['B04', 'B08'],
    outdir='sentinel2_L1C',
    processing_level='L1C'  # Top of Atmosphere
)

downloader.run()

Output: one mosaic per requested band per acquisition date, organized in subfolders <outdir>/YYYYMMDD_s2msi/. Files are named YYYYMMDD_s2msi_<alias>.tif (e.g. 20230610_s2msi_b04_red.tif). Each band is delivered in its native resolution (10 m or 20 m); the SCL band keeps its uint8 dtype.

Processing Level Comparison:

• L2A (default): Bottom of Atmosphere - atmospherically corrected, ready for analysis
• L1C: Top of Atmosphere - raw reflectance, use for custom atmospheric correction with tools like ACOLITE

Available Sentinel-2 bands

Reflectance bands (any combination of resolutions can be requested in a single call):

Band	Alias	Resolution	Description
B02	b02_blue	10 m	Blue
B03	b03_green	10 m	Green
B04	b04_red	10 m	Red
B08	b08_nir	10 m	NIR (broad, 842 nm)
B01	b01_coastal	20 m	Coastal aerosol
B05	b05_re1	20 m	Red-edge 1
B06	b06_re2	20 m	Red-edge 2
B07	b07_re3	20 m	Red-edge 3
B8A	b8a_nir	20 m	NIR (narrow, 865 nm)
B11	b11_swir1	20 m	SWIR 1
B12	b12_swir2	20 m	SWIR 2

Cloud / scene-classification mask (L2A only):

Band	Alias	Resolution	Description
SCL	scl	20 m	Sen2Cor Scene Classification Layer (categorical)

Example use cases:

# NDVI calculation
CDSEDownload(shape='area.shp', dates=('2023-01-01', '2023-12-31'), bands=['B04', 'B08'])

# RGB visualization
CDSEDownload(shape='area.shp', dates=('2023-01-01', '2023-12-31'), bands=['B02', 'B03', 'B04'])

# Water mask / turbidity (mixes 10 m + 20 m + categorical SCL)
CDSEDownload(
    shape='area.shp',
    dates=('2023-01-01', '2023-12-31'),
    bands=['B02', 'B03', 'B04', 'B08', 'B11', 'B12', 'SCL'],
)

---

Available Datasets & Products

✅ VPP_Pheno (Vegetation Phenology & Productivity)

Status: Working
Products:

• SOSD (Start of Season Date)
• EOSD (End of Season Date)
• MAXD (Maximum Date)
• MINV (Minimum Value)
• MAXV (Maximum Value)
• AMPL (Amplitude)
• LENGTH (Length of Season)
• LSLOPE (Left Slope)
• RSLOPE (Right Slope)
• SPROD (Seasonal Productivity)
• TPROD (Total Productivity)
• SOSV (Start of Season Value)
• EOSV (End of Season Value)

✅ VPP_ST (Seasonal Trajectories)

Status: Working
Products:

• PPI (Plant Phenology Index)
• QFLAG (Quality Flag)

⚠️ VPP_Index (Vegetation Indices)

Status: NOT AVAILABLE (404 Error as of January 2025)
Products (no longer accessible):

• LAI (Leaf Area Index)
• FAPAR (Fraction of Absorbed Photosynthetically Active Radiation)
• NDVI (Normalized Difference Vegetation Index)
• PPI (now available in VPP_ST)

✅ SLSTR (Sentinel-3 Land Surface Temperature)

Status: Working
Products: SL_2_LST___

Using DEIMS IDs

You can use DEIMS site IDs instead of shapefiles:

downloader = pyvpp.wekeo_download(
    dataset='VPP_Pheno',
    shape='deimsid:https://deims.org/bcbc866c-3f4f-47a8-bbbc-0a93df6de7b2',
    dates=['2020-01-01', '2020-12-31'],
    products=['SOSD', 'MAXD']
)

Step-by-step workflow

PyVPP performs the following operations:

1. Downloads all Sentinel-2 tiles that intersect your area of interest
2. Filters tiles by UTM zone
3. Creates mosaics for each date and product
4. Clips mosaics to your exact boundaries
5. Saves final products as mosaic_YYYYMMDD_PRODUCT_rec.tif
6. Cleans up intermediate files

Advanced Usage

Use credentials directly (without .hdarc)

downloader = pyvpp.wekeo_download(
    dataset='VPP_Pheno',
    shape='area.shp',
    dates=['2020-01-01', '2020-12-31'],
    products=['SOSD'],
    user='your_username',      # Direct credentials
    password='your_password'
)

Clean old .hdarc format

If you have an old .hdarc file (pre-March 2024):

import pyvpp
pyvpp.clean_old_hdarc()  # Removes obsolete 'url:' line

Step-by-step execution

# Execute steps individually
downloader.download()        # Only download
downloader.mosaic_and_clip() # Only mosaic and clip
downloader.clean()           # Only clean intermediate files

Alternatives for LAI, FAPAR, NDVI

Since VPP_Index is currently unavailable, here are alternatives:

1. Google Earth Engine (Recommended)

// Calculate NDVI from Sentinel-2
var s2 = ee.ImageCollection('COPERNICUS/S2_SR_HARMONIZED')
  .filterDate('2020-01-01', '2020-12-31')
  .filterBounds(geometry);

var addNDVI = function(image) {
  var ndvi = image.normalizedDifference(['B8', 'B4']).rename('NDVI');
  return image.addBands(ndvi);
};

2. Sentinel Hub

https://www.sentinel-hub.com/

• Pre-computed LAI, FAPAR, NDVI available

3. Copernicus Global Land Service

https://land.copernicus.eu/global/products/

• 300m resolution (not 10m like HR-VPP)

4. Calculate from Sentinel-2 directly

import rasterio
import numpy as np

# NDVI = (NIR - Red) / (NIR + Red)
with rasterio.open('sentinel2.tif') as src:
    red = src.read(4)  # Band 4
    nir = src.read(8)  # Band 8
    ndvi = (nir - red) / (nir + red)

Troubleshooting

Error 404: Dataset not found

If you get a 404 error with VPP_Index:

requests.exceptions.HTTPError: 404 Client Error: Not Found for url: 
https://gateway.prod.wekeo2.eu/hda-broker/api/v1/datasets/EO:EEA:DAT:CLMS_HRVPP_VI

Solution: Use VPP_Pheno or VPP_ST instead (see examples above).

Authentication failed

# Recreate your .hdarc file
import pyvpp
pyvpp.create_hdarc("username", "password")

No matches found

This usually means:

1. Date range is outside available data
2. Area of interest has no coverage
3. Product name is misspelled

Check the official documentation for data availability.

Requirements

• Python >= 3.9
• requests >= 2.28
• geopandas >= 0.12
• rasterio >= 1.3
• pyproj >= 3.4
• shapely >= 1.8
• deims >= 3.1
• fiona >= 1.8.20
• hda >= 2.18 (for WEkEO only)

Documentation & Support

• HR-VPP Documentation: https://land.copernicus.eu/pan-european/biophysical-parameters/high-resolution-vegetation-phenology-and-productivity
• WEkEO Help Center: https://help.wekeo.eu/
• PyVPP Repository: https://github.com/Digdgeo/PyVPP
• Report Issues: https://github.com/Digdgeo/PyVPP/issues

Citation

If you use PyVPP in your research, please cite:

García-Díaz, D. (2025). PyVPP: Python package for HR-VPP data access and processing.
https://github.com/Digdgeo/PyVPP

License

MIT License - see LICENSE file for details

Contributing

Contributions are welcome! Please feel free to submit a Pull Request.

Changelog

v1.0.0 (Current)

• Per-date output: mosaics organized in YYYYMMDD_s2msi/ subfolders, one per acquisition date
• Landsat-style naming: YYYYMMDD_s2msi_<alias>.tif (e.g. 20230610_s2msi_b04_red.tif)
• Red-edge bands: B05, B06, B07, B8A now fully supported
• Token auto-refresh: prevents 401 Unauthorized on long download sessions
• Resume support: already-extracted SAFE directories are skipped on re-run
• geopandas >= 0.14 required (fixes fiona 1.10+ compatibility)

v0.4.0

• Added CDSE (Copernicus Data Space Ecosystem) support for Sentinel-2 L2A/L1C
• New module pyvpp.cdse with CDSEDownload class
• Multi-resolution support (10 m + 20 m + SCL in a single call)
• Per-band output files in native resolution
• Authentication via config file, env vars, or direct credentials

v0.1.9

• Fixed .hdarc format for HDA API v2.18+
• Added create_hdarc(), clean_old_hdarc(), delete_hdarc() helpers

---

Last updated: May 2026 Status: VPP_ST ✅ | VPP_Pheno ✅ | VPP_Index ⚠️ (unavailable) | CDSE Sentinel-2 ✅