Living England: satellite-based habitat classification. Technical User Guide.
The Living England project, led by Natural England, is a multi-year programme delivering a satellite-derived national habitat layer in support of the Environmental Land Management (ELM) System and the Natural Capital and Ecosystem Assessment (NCEA) Pilot. The project uses a machine learning approach to image classification, developed under the Defra Living Maps project (SD1705 - Kilcoyne et al., 2017). The method first clusters homogeneous areas of habitat into segments using satellite imagery, then assigns each segment to a defined list of habitat classes using Random Forest machine learning algorithm. Phase I of the Living England project established the methodology for the analysis, Phase II achieved an average probability of 78%, and was improved in Phase III with an average probability of 84%. In Phase IV, further improvements to the classification and training data has resulted in an average habitat probability of 88%. The habitat probability reported by Living England represents the degree of agreement between the classification map and field collected data. The probability is output on a per segment basis and will vary between habitats and biogeographic zones. This technical user guide summarises the data and methodologies used to create the Living England habitat map, shown in Figure 1, produced in Phase IV of the project. This guide is to accompany the release of the Phase IV data under an Open Government Licence (OGL) providing users with information on how these modelled predictions were generated and how to interpret the results.