- November 16, 2022
- Posted by: EARSC
- Categories: European EO Industry, Members News
Habitat Mapping is a key tool to assess progress towards the European Biodiversity 2030 targets as well as the global Aichi targets and more particular the EU Habitats Directive and the Nature Restoration Law. Many initiatives have already been taken in Europe, ranging from local to national scale. However at European scale, habitat maps remain limited to very coarse spatial resolution of 1 km to 10 km and are often only a prediction of the suitability of the habitat instead of a hard classification.
Recent advances in Machine Learning and Artificial Intelligence, in combination with the high resolution Earth Observation data and derived Remote Sensing-enabled Essential Biodiversity Variables, improve Habitat Mapping which can be directly used for policies as restoration law, pollination initiative, preserving Europe’s natural capital.
- by Bruno Smets 17.10.2022
EXPLORE CONTINENTAL HABITAT MAPPING OVER EUROPE
The latest assessment on the state of the European environment, published by the European Environment Agency (EEA), shows that Europe’s biodiversity continues to decline at an alarming rate, with most protected species and habitats found not to have a good conservation status. It is clear that much more efforts are needed to reverse current trends and to ensure resilient and healthy nature.
To define more efforts or efforts that can have more effect we need to improve our biodiversity and ecosystem monitoring services in order to support our European policy makers. For this we are looking towards new Artificial Intelligence (AI) techniques. By using Neural Networks or Deep Learning methods and satellite-based high-resolution datasets such as the Copernicus High Resolution Vegetation Phenology Product (HR-VPP) we can improve our knowledge on where habitats occur across Europe, a crucial element for improving biodiversity conservation and taking specific actions.
Read more at https://blog.vito.be/remotesensing/habitat-mapping-ai