Earth Observation and Biodiversity Big Data for Forest Habitat Types Classification and Mapping

In the light of “Biological Diversity” concept, habitats are cardinal pieces for biodiversity quantitative estimation at a local and global scale. Europe EUNIS (European Nature Information System) is system tool habitat identification assessment. Earth Observation (EO) data, which acquired by satellite sensors, offer new opportunities environmental sciences they revolutionizing methodologies applied. These providing unprecedented insights monitoring evaluating Sustainable Development Goals (SDGs) indicators. This paper shows results novel approach spatially explicit mapping in Italy national scale, using supervised machine learning model (SMLM), through combination vegetation plot database (as response variable), both spectral predictors. The procedure integrates forest data from European Vegetation Archive (EVA), with Sentinel-2 imagery processing (vegetation indices time series, indices, single bands signals) variables (i.e., climatic topographic), to parameterize Random Forests (RF) classifier. obtained classify 24 according III level: 12 broadleaved deciduous (T1), 4 evergreen (T2) eight needleleaved (T3), achieved an overall accuracy 87% II level classes (T1, T2, T3), 76.14% level. highest value was equal 91%, followed 76% 68% forests, respectively. proposed methodology open way increase categories be mapped together their geographical extent, test different semi-supervised algorithms ensemble modelling methods.