EUNIS habitat maps: enhancing thematic and spatial resolution
for Europe through machine learning

J 2025

EUNIS habitat maps: enhancing thematic and spatial resolution for Europe through machine learning

SI-MOUSSI, Sara; Stephan HENNEKENS; Sander MUCHER; De Keersmaecker WANDA; Milan CHYTRÝ et al.

Základní údaje

Originální název

EUNIS habitat maps: enhancing thematic and spatial resolution for Europe through machine learning

Autoři

Vydání

Scientific Data, BERLIN, NATURE PORTFOLIO, 2025, 2052-4463

Další údaje

Jazyk

angličtina

Typ výsledku

Článek v odborném periodiku

Obor

10600 1.6 Biological sciences

Stát vydavatele

Německo

Utajení

není předmětem státního či obchodního tajemství

Odkazy

URL

Impakt faktor

Impact factor: 6.900 v roce 2024

Označené pro přenos do RIV

Ano

Organizační jednotka

Přírodovědecká fakulta

Klíčová slova anglicky

Conservation of Natural Resources; Ecosystem; Europe; Machine Learning; Satellite Imagery

Štítky

14314020, rivok

Příznaky

Mezinárodní význam, Recenzováno

Změněno: 2. 2. 2026 14:42, Mgr. Marie Novosadová Šípková, DiS.

Anotace

V originále

The EUNIS habitat classification is crucial for categorising European habitats, supporting European policy on nature conservation and implementing the Nature Restoration Law. To meet the growing demand for detailed and accurate habitat information, we provide spatial predictions across Europe (EEA39 territory) for 260 EUNIS habitat types at hierarchical level 3, together with independent validation and uncertainty analyses. Using ensemble machine learning models, together with high-resolution satellite imagery and ecologically meaningful climatic, topographic and edaphic variables, we produced a European habitat map indicating the most probable habitat overall at 100-m resolution across Europe. Additionally, we provide information on prediction uncertainty and the most probable habitats at level 3 within each EUNIS level 1 formation. This product is particularly useful for both conservation and restoration purposes. Predictions were cross-validated at European scale using a spatial block cross-validation and evaluated against independent data from France (forests only), the Netherlands and Austria. The maps achieved strong predictive performance, with F1-scores ranging from 0.61 to 0.94 in spatial cross-validation and from 0.33 to 0.95 in external validation datasets with distinct trade-offs in terms of recall and precision across habitat formations. Accuracy improved for rare or localized habitats when considering the top 3 predicted classes.

Citovat

SI-MOUSSI, Sara; Stephan HENNEKENS; Sander MUCHER; De Keersmaecker WANDA; Milan CHYTRÝ; Emiliano AGRILLO; Fabio ATTORRE; Idoia BIURRUN; Gianmaria BONARI; Andraz CARNI; Renata CUSTEREVSKA; Tetiana DZIUBA; Klaus ECKER; Behlul GULER; Ute JANDT; Borja JIMENEZ-ALFARO; Jonathan LENOIR; Jens-Christian SVENNING; Grzegorz SWACHA a Wilfried THUILLER. EUNIS habitat maps: enhancing thematic and spatial resolution for Europe through machine learning. Scientific Data. BERLIN: NATURE PORTFOLIO, 2025, roč. 12, č. 1, s. 1-18. ISSN 2052-4463. Dostupné z: https://doi.org/10.1038/s41597-025-06235-7.

@article{2548657,
   author = {SiandMoussi, Sara and Hennekens, Stephan and Mucher, Sander and Wanda, De Keersmaecker and Chytrý, Milan and Agrillo, Emiliano and Attorre, Fabio and Biurrun, Idoia and Bonari, Gianmaria and Carni, Andraz and Custerevska, Renata and Dziuba, Tetiana and Ecker, Klaus and Guler, Behlul and Jandt, Ute and JimenezandAlfaro, Borja and Lenoir, Jonathan and Svenning, JensandChristian and Swacha, Grzegorz and Thuiller, Wilfried},
   article_location = {BERLIN},
   article_number = {1},
   doi = {https://doi.org/10.1038/s41597-025-06235-7},
   keywords = {Conservation of Natural Resources; Ecosystem; Europe; Machine Learning; Satellite Imagery},
   language = {eng},
   issn = {2052-4463},
   journal = {Scientific Data},
   title = {EUNIS habitat maps: enhancing thematic and spatial resolution for Europe through machine learning},
   url = {https://doi.org/10.1038/s41597-025-06235-7},
   volume = {12},
   year = {2025}
}

TY  - JOUR
ID  - 2548657
AU  - Si-Moussi, Sara - Hennekens, Stephan - Mucher, Sander - Wanda, De Keersmaecker - Chytrý, Milan - Agrillo, Emiliano - Attorre, Fabio - Biurrun, Idoia - Bonari, Gianmaria - Carni, Andraz - Custerevska, Renata - Dziuba, Tetiana - Ecker, Klaus - Guler, Behlul - Jandt, Ute - Jimenez-Alfaro, Borja - Lenoir, Jonathan - Svenning, Jens-Christian - Swacha, Grzegorz - Thuiller, Wilfried
PY  - 2025
TI  - EUNIS habitat maps: enhancing thematic and spatial resolution for Europe through machine learning
JF  - Scientific Data
VL  - 12
IS  - 1
SP  - 1-18
EP  - 1-18
PB  - NATURE PORTFOLIO
SN  - 20524463
KW  - Conservation of Natural Resources
KW  - Ecosystem
KW  - Europe
KW  - Machine Learning
KW  - Satellite Imagery
UR  - https://doi.org/10.1038/s41597-025-06235-7
N2  - The EUNIS habitat classification is crucial for categorising European habitats, supporting European policy on nature conservation and implementing the Nature Restoration Law. To meet the growing demand for detailed and accurate habitat information, we provide spatial predictions across Europe (EEA39 territory) for 260 EUNIS habitat types at hierarchical level 3, together with independent validation and uncertainty analyses. Using ensemble machine learning models, together with high-resolution satellite imagery and ecologically meaningful climatic, topographic and edaphic variables, we produced a European habitat map indicating the most probable habitat overall at 100-m resolution across Europe. Additionally, we provide information on prediction uncertainty and the most probable habitats at level 3 within each EUNIS level 1 formation. This product is particularly useful for both conservation and restoration purposes. Predictions were cross-validated at European scale using a spatial block cross-validation and evaluated against independent data from France (forests only), the Netherlands and Austria. The maps achieved strong predictive performance, with F1-scores ranging from 0.61 to 0.94 in spatial cross-validation and from 0.33 to 0.95 in external validation datasets with distinct trade-offs in terms of recall and precision across habitat formations. Accuracy improved for rare or localized habitats when considering the top 3 predicted classes.
ER  -

SI-MOUSSI, Sara; Stephan HENNEKENS; Sander MUCHER; De Keersmaecker WANDA; Milan CHYTRÝ; Emiliano AGRILLO; Fabio ATTORRE; Idoia BIURRUN; Gianmaria BONARI; Andraz CARNI; Renata CUSTEREVSKA; Tetiana DZIUBA; Klaus ECKER; Behlul GULER; Ute JANDT; Borja JIMENEZ-ALFARO; Jonathan LENOIR; Jens-Christian SVENNING; Grzegorz SWACHA a Wilfried THUILLER. EUNIS habitat maps: enhancing thematic and spatial resolution for Europe through machine learning. \textit{Scientific Data}. BERLIN: NATURE PORTFOLIO, 2025, roč.~12, č.~1, s.~1-18. ISSN~2052-4463. Dostupné z: https://doi.org/10.1038/s41597-025-06235-7.

Přehled o publikaci