Current and future European potential vegetation types

Potential Natural Vegetation (PNV) estimates for the European continent at 1km grain size.

Source
Zenodo
Published
2025
Region
Europe
Format
multiple formats
Access
Open access
License
cc-by-4.0
Themes
Plants & Vegetation, Biodiversity & Ecosystems

Get the data from Zenodo

About this dataset

This dataset contains Potential Natural Vegetation (PNV) estimates for the European continent at 1km grain size. Estimates are made for six different vegetation types following the MAES Ecosystem classification at level 1. The predictions have been made through an ensemble of Bayesian Habitat distribution models available through the ibis.iSDM package (Jung 2023) . For more information on the methodology, original data and used covariates, please see the accompanying preprint ( Jung 2024 ).

Uploaded are:

The most likely current PNV transition (see screenshot) as categorical raster (and screenshot, see png)

(Classes: 1=Woodland.and.forest | 2=Heathland.and.shrub | 3=Grassland | 4=Sparsely.vegetated.areas | 5=Wetlands | 6=Marine.inlets.and.transitional.waters)

Current PNV estimates as cloud-optimized geoTIFF ("COG") files (.tif)

Future PNV estimates (zipped) for each considered SSP - GCM combination as geoTIFF (.tif).

Variable naming scheme:

Current: "pnv_XX_laea_1km.tif"

where XX represents the vegetation type

Future: Here the hierachical organization scheme of Essential Biodiversity Variables (EBV) is followed where files are separated in folders by

Scenario | metric | entity | time, so for example "SSP126-GFDL-ESM4/suitability_mean/grassland/"

Filenames are labelled by the date (e.g. "2040.tif").

Metrics and layers names and their interpretation:

For current:

"mean" = Average Ensemble posterior prediction

"sd" = Standard deviation of posterior prediction

"q05" = Lower percentile (5%) of posterior prediction

"q50" = Median or 50% percentile of posterior prediction

"q95" = Upper percentile (95%) of posterior prediction

"mode" = Most commonly encountered value of posterior prediction

"cv" = Coefficient of variation of posterior prediction

For future:

"mean" = Average Ensemble posterior prediction

"q05" = Lower percentile (5%) of posterior prediction

"q50" = Median or 50% percentile of posterior prediction

"q95" = Upper percentile (95%) of posterior prediction

---

Data properties:

Shared Socioeconomic Pathways (SSP)

SSP1-2.6, SSP2-4.5, SSP5-8.5

General circulation models (GCMs)

GFDL-ESM4,

IPSL-CM6A-LR,

MPI-ESM1-2-HR,

MRI-ESM2-0,

UKESM1-0-LL

Spatial grain

1 km²

Geographic projection

LAEA

Temporal grain

30 year climatologies

Spatial extent

Continental Europe including Turkey (see screenshot)

Temporal extent

1990 to 2020 (Current), 2020 - 2100 (Future)

Number of variables/entities

7

All files are provided as is and the author takes no responsibility for errors or misuse and misinterpretation.

How the data was collected

This dataset contains Potential Natural Vegetation (PNV) estimates for the European continent at 1km grain size. Estimates are made for six different vegetation types following the MAES Ecosystem classification at level 1. The predictions have been made through an ensemble of Bayesian Habitat distribution models available through the ibis.iSDM package (Jung 2023) . For more information on the methodology, original data and used covariates, please see the accompanying preprint ( Jung 2024 ).

Uploaded are:

The most likely current PNV transition (see screenshot) as categorical raster (and screenshot, see png)

(Classes: 1=Woodland.and.forest | 2=Heathland.and.shrub | 3=Grassland | 4=Sparsely.vegetated.areas | 5=Wetlands | 6=Marine.inlets.and.transitional.waters)

Current PNV estimates as cloud-optimized geoTIFF ("COG") files (.tif)

Future PNV estimates (zipped) for each considered SSP - GCM combination as geoTIFF (.tif).

Variable naming scheme:

Current: "pnv_XX_laea_1km.tif"

where XX represents the vegetation type

Future: Here the hierachical organization scheme of Essential Biodiversity Variables (EBV) is followed where files are separated in folders by

Scenario | metric | entity | time, so for example "SSP126-GFDL-ESM4/suitability_mean/grassland/"

Filenames are labelled by the date (e.g. "2040.tif").

Metrics and layers names and their interpretation:

For current:

"mean" = Average Ensemble posterior prediction

"sd" = Standard deviation of posterior prediction

"q05" = Lower percentile (5%) of…

"q5…

Browse sustainability datasets on SusanHub