EUNIS habitat type F3.1a, predicted habitat suitability - version 1, June 2016

The modelled suitability for the EUNIS habitat type is an indication of where conditions are favourable for the habitat type based on sample plot data (Braun-Blanquet database) and the Maxent software package. The modelled suitability map may be used as a proxy for the geographical distribution of the habitat type. Note however that it is not representing the actual distribution of the habitat type.

Also note that predictions are less reliable due to data deficiency in the eastern part of Europe, and to a lesser extent to the Scandinavian countries.

Geographic restriction for plot observations: n/a

Remarks: -Prediction in eastern part of Europe (Caucasus, Türkiye) uncertain due to lack of data for that area.

Date (Publication): 2016-07-01

Date (Creation): 2016-07-06

Edition: 01

Citation identifier: eea_r_3035_1_km_eunis-hab-f3-1a_p_1940-2011_v01_r00

Status: Obsolete

Point of contact

Organisation name	Individual name	Electronic mail address	Website	Role
European Environment Agency		sdi@eea.europa.eu	http://www.eea.europa.eu	Point of contact
European Environment Agency		sdi@eea.europa.eu		Custodian

Point of contact

No information provided.

Maintenance and update frequency: Unknown

EEA topics

Biodiversity

GEMET - INSPIRE themes, version 1.0

Habitats and biotopes

GEMET

natural area
tundra
terrestrial ecosystem
heathland

Keywords

Place

Europe

Use limitation: EEA standard re-use policy: unless otherwise indicated, re-use of content on the EEA website for commercial or non-commercial purposes is permitted free of charge, provided that the source is acknowledged ( http://www.eea.europa.eu/legal/copyright). Copyright holder: European Environment Agency (EEA).

Access constraints: Other restrictions

Other constraints: no limitations to public access

Spatial representation type: Grid

Distance: 1 km

Language of dataset: English

Character set: UTF8

Topic category

Biota

Begin date: 1940-01-01

End date: 2011-12-31

Coordinate reference system identifier: EPSG:3035

Distribution format

GeoTIFF ( )

OnLine resource

Protocol	Linkage	Name
EEA:FILEPATH	https://sdi.eea.europa.eu/webdav/datastore/public/eea_r_3035_1_km_eunis-hab-f3-1a_p_1940-2011_v01_r00/F3-1a_ed1.tif
WWW:URL	https://sdi.eea.europa.eu/data/cb23d43e-5495-4ab9-861e-1c2d13db33e2	Direct download

Hierarchy level: Dataset

Conformance result

Title: Commission Regulation (EU) No 1089/2010 of 23 November 2010 implementing Directive 2007/2/EC of the European Parliament and of the Council as regards interoperability of spatial data sets and services

Date (Publication): 2010-12-08

Explanation: See the referenced specification

Statement: The database compiled for the Braun-Blanquet project is a compilation of various national and regional vegetation databases. The maintenance of these databases is in principle in the hands of the custodians. However, before uploading the databases into Braun-Blanquet database a quality check is performed by Alterra and Masaryk University. If possible, detected errors are corrected and reported back to the data provider. For the modelling of the habitat suitability map the Maxent software is used ( http://www.cs.princeton.edu/~schapire/maxent/). The grid values in the map represent the probability (ranging from 0-1) that the cell is suitable for the habitat.

The grid file represents the habitat suitability of the EUNIS type. For the modelling the widely used software Maxent for maximum entropy modelling of species’ geographic distributions was used. Maxent is a general-purpose machine-learning method with a simple and precise mathematical formulation, and has a number of aspects that make it well-suited for species distribution modelling when only presence (occurrence) data but not absence data are available (Philips et al. 2006). Because EUNIS habitats have a particular species composition, they are assumed to respond to specific ecological requirements, allowing us to generate correlative estimates of geographic distributions. Modelling habitats that have been floristically defined is a well-known procedure for ecological modelling at local scales, and a promising technique to be applied also at the continental level.

The Maxent method considers presence data (known observations of a given entity) and the so-called background data. Background data comprise a set of points used to describe the environmental variation of the study area according to the available environmental layers. It is assumed that these layers represent well the most important ecological gradients on a European scale. As layers the following environmental parameters have been used: Potential Evapotranspiration, Topsoil pH, Solar radiation, Temperature Seasonality (standard deviation *100), Mean Temperature of Wettest Quarter, Annual Precipitation, Precipitation Seasonality (Coefficient of Variation), Precipitation of Warmest Quarter & Distance to water (rivers, lakes, sea).

Statistical output of the model:

#Training samples: 261

Regularized training gain: 1.451

Unregularized training gain: 1.6818

Iterations: 500

Training AUC: 0.9294

#Test samples: 28

Test gain: 1.6294

Test AUC: 0.9168

AUC Standard Deviation: 0.0248

#Background points: 5261

bio_12_etrs2_ras contribution: 16.9278

bio_15_etrs2_ras contribution: 1.7383

bio_18_etrs2_ras contribution: 3.0896

bio_4_etrs2_ras contribution: 47.2878

bio_8_etrs2_ras contribution: 1.1727

bld_m_sd1_1km_eu_ll contribution: 2.8954

cecsum_m_sd1_1km_eu_ll contribution: 0.1047

clyppt_m_sd1_1km_eu_ll contribution: 0.2259

crfvol_m_sd1_1km_eu_ll contribution: 4.1454

dist2water1km contribution: 0.0476

orcdrc_m_sd1_1km_eu_ll contribution: 0.8552

pet_he_yr contribution: 11.6802

phihox_m_sd1_1km_eu_ll contribution: 0.3306

sltppt_m_sd1_1km_eu_ll contribution: 0.4748

sndppt_m_sd1_1km_eu_ll contribution: 2.8708

solar_1km contribution: 6.1532

bio_12_etrs2_ras permutation importance: 11.098

bio_15_etrs2_ras permutation importance: 1.4671

bio_18_etrs2_ras permutation importance: 3.3948

bio_4_etrs2_ras permutation importance: 39.0025

bio_8_etrs2_ras permutation importance: 0.9225

bld_m_sd1_1km_eu_ll permutation importance: 2.9407

cecsum_m_sd1_1km_eu_ll permutation importance: 0.3514

clyppt_m_sd1_1km_eu_ll permutation importance: 1.0385

crfvol_m_sd1_1km_eu_ll permutation importance: 13.2838

dist2water1km permutation importance: 0.3511

orcdrc_m_sd1_1km_eu_ll permutation importance: 2.4478

pet_he_yr permutation importance: 15.1798

phihox_m_sd1_1km_eu_ll permutation importance: 2.9375

sltppt_m_sd1_1km_eu_ll permutation importance: 2.438

sndppt_m_sd1_1km_eu_ll permutation importance: 0.9287

solar_1km permutation importance: 2.2179

Entropy: 7.1344

Prevalence (average of logistic output over background sites): 0.1155

Fixed cumulative value 1 cumulative threshold: 1

Fixed cumulative value 1 logistic threshold: 0.0186

Fixed cumulative value 1 area: 0.5052

Fixed cumulative value 1 training omission: 0.0115

Fixed cumulative value 1 test omission: 0.0357

Fixed cumulative value 1 binomial probability: 5.91E-07

Fixed cumulative value 5 cumulative threshold: 5

Fixed cumulative value 5 logistic threshold: 0.0954

Fixed cumulative value 5 area: 0.3203

Fixed cumulative value 5 training omission: 0.0307

Fixed cumulative value 5 test omission: 0.0357

Fixed cumulative value 5 binomial probability: 1.40E-13

Fixed cumulative value 10 cumulative threshold: 10

Fixed cumulative value 10 logistic threshold: 0.1694

Fixed cumulative value 10 area: 0.2387

Fixed cumulative value 10 training omission: 0.0651

Fixed cumulative value 10 test omission: 0.1071

Fixed cumulative value 10 binomial probability: 2.35E-16

Minimum training presence cumulative threshold: 0.5143

Minimum training presence logistic threshold: 0.0103

Minimum training presence area: 0.587

Minimum training presence training omission: 0

Minimum training presence test omission: 0.0357

Minimum training presence binomial probability: 2.51E-05

10 percentile training presence cumulative threshold: 18.2675

10 percentile training presence logistic threshold: 0.2609

10 percentile training presence area: 0.1671

10 percentile training presence training omission: 0.0996

10 percentile training presence test omission: 0.1786

10 percentile training presence binomial probability: 8.34E-21

Equal training sensitivity and specificity cumulative threshold: 22.3303

Equal training sensitivity and specificity logistic threshold: 0.3025

Equal training sensitivity and specificity area: 0.1426

Equal training sensitivity and specificity training omission: 0.1418

Equal training sensitivity and specificity test omission: 0.2143

Equal training sensitivity and specificity binomial probability: 1.08E-22

Maximum training sensitivity plus specificity cumulative threshold: 18.2675

Maximum training sensitivity plus specificity logistic threshold: 0.2609

Maximum training sensitivity plus specificity area: 0.1671

Maximum training sensitivity plus specificity training omission: 0.0996

Maximum training sensitivity plus specificity test omission: 0.1786

Maximum training sensitivity plus specificity binomial probability: 8.34E-21

Equal test sensitivity and specificity cumulative threshold: 16.6662

Equal test sensitivity and specificity logistic threshold: 0.2475

Equal test sensitivity and specificity area: 0.1785

Equal test sensitivity and specificity training omission: 0.0958

Equal test sensitivity and specificity test omission: 0.1786

Equal test sensitivity and specificity binomial probability: 3.20E-19

Maximum test sensitivity plus specificity cumulative threshold: 36.3306

Maximum test sensitivity plus specificity logistic threshold: 0.4278

Maximum test sensitivity plus specificity area: 0.0842

Maximum test sensitivity plus specificity training omission: 0.2759

Maximum test sensitivity plus specificity test omission: 0.2143

Maximum test sensitivity plus specificity binomial probability: 4.69E-41

Balance training omission, predicted area and threshold value cumulative threshold: 3.5127

Balance training omission, predicted area and threshold value logistic threshold: 0.072

Balance training omission, predicted area and threshold value area: 0.3592

Balance training omission, predicted area and threshold value training omission: 0.0192

Balance training omission, predicted area and threshold value test omission: 0.0357

Balance training omission, predicted area and threshold value binomial probability: 1.25E-11

Equate entropy of thresholded and original distributions cumulative threshold: 10.0342

Equate entropy of thresholded and original distributions logistic threshold: 0.1695

Equate entropy of thresholded and original distributions area: 0.2384

Equate entropy of thresholded and original distributions training omission: 0.0651

Equate entropy of thresholded and original distributions test omission: 0.1071

Equate entropy of thresholded and original distributions binomial probability: 2.18E-16