Development of a methodological framework for avian species habitat mappping in Greece

The aim of this study is to present the framework developed for providing spatially continuous (wall to wall) avian species habitat maps that integrate functional requirements (breeding, nesting, foraging, wintering) over the Greek Natura 2000 SPA areas at high resolution scale under the LIFE IP 4 NATURA project. Originally, a standardized avian habitat classification and mapping methodology was developed, based on EU and/or international typologies. The proposed activities will align the classification approach with EU Ecosystem Typology, IUCN Global Ecosystem Typology, EUNIS habitat types, Annex I habitat types (Dir. 92/43/EEC), national habitat/habitat type inventories, and relevant policy frameworks. Also, for the specification of the scheme a consultation procedure with key stakeholders will be implemented including Ministry of Environment and Energy (MEEN), Natural Environment & Climate Change Agency (NECCA), experts/ornithologists, and Natura 2000 managers to refine categories for local contexts (e.g., specialized wetland or coastal habitats, etc.). The classification scheme will also establish crosswalks between existing national classifications, the EU Ecosystem Typology and EUNIS/EUR28 habitat types. Targeted visits will also be used to clarify ambiguous habitats and finalize classification descriptors (e.g., farmland habitats with mosaic structures).The Earth Observation (EO) datasets to be employed will include cloud-free, multi-season Sentinel-2 images (covering spring, summer, autumn) to capture habitat phenology and high-very high spatial resolution imagery including Copernicus VHR Image Mosaic 2021. Recent VHR ortho imagery acquired from Natural Environment & Climate Change Agency will also be acquired for extracting habitat maps. Ancillary data as existing habitat maps, topographic layers (DEM), climatic data (temperature, rainfall), Copernicus HRL layers and soil maps will be also gathered and homogenized in a spatial geodatabase. The mapping process will follow an object-based analysis approach using supervised classification methods with machine learning algorithms. The habitat maps will be coupled with avian & functional trait data bird for habitat suitability modelling.