Making Earth Observation Work for UK Biodiversity: Phase 2 Norfolk Pilot Study
Cameron, Iain; Medcalf, Katie; Turton, Nicki; Bell, Gemma
Environment Systems Ltd, UNITED KINGDOM

A robust system for surveillance of habitat extent and condition is key for the conservation of priority habitats. The Crick framework was developed by Environment Systems on behalf of Defra and JNCC during Phase 1 of the Making Earth Observation Work for UK Biodiversity project. It offers a systematic method to characterise how well habitats can be identified using Earth Observation (EO) data. The ability to distinguish between different habitats in EO data depends upon the spectral signature, phenology and canopy contribution of key plant species together with contextual landscape information. The Crick Framework draws upon this wide range of factors and, together with ecological knowledge, provides a generic classification system where tiers of habitats are grouped by the types of EO and ancillary data required to classify them accurately. This paper presents results from the Norfolk Pilot Study (forming Phase 2 of the project) which is investigating the implementation of the Crick framework for identifying the extent and condition of BAP Priority and Annex I habitats in Norfolk, within the UK. The Norfolk pilot study integrates a wide range of EO data (from landscape-level satellite imagery, such as SPOT-5 and Rapideye, to site specific high resolution multispectral UAV imagery), together with ancillary mapping and elevation data. Priority habitats were classified using the Crick Framework to identify which combinations of EO data should allow for accurate identification, this was then put into practice using an object-oriented ecological rulebase classification methodology. Results so far have demonstrated that a number of priority habitats in this intricate lowland environment can be identified at the broad landscape-scale using satellite data. For Annex I habitats the best results have been found using high resolution digital elevation data from Lidar and UAV photogrammetry, together with very high resolution multispectral imagery targeted at specific sites. The data has also demonstrated the capacity to monitor habitat condition through parameters such as vegetation productivity, wetness/dryness and identification of large single-species stands. Results indicate that EO gives very useful and quantifiable information for looking at extent and condition of specific sites, and can be targeted appropriately using the Crick Framework.