Section 14 CROME

The Crop Map of England (CROME) dataset is a tessellated (hexagonal) vector product derived via remote sensing techniques. It concentrates on classifying areas of crops in England using the Sentinel satellite constellation, a random forest classification and ground-truth data. Ground-truth data were recorded by RPA field inspectors during the spring/summer of each year. Each hexagonal cell was ~0.4 hectares and represents one of 80 classes (55 cereal crops, 15 leguminous crops, 1 energy crop, 3 grassland types, 3 tree classes, water, non-agricultural land and mixed).

Unfortunately, the data are only made available on a county-by-county basis for each year. Requests to the RPA for a bulk download have not yet provided the full dataset. Retrieving and processing all 46 files for each year is feasible but time-consuming, and not deemed efficient use of time when we should be able to get the complete data in bulk, albeit at a later date. A further issue is that the data are trained and evaluated via RPA field inspections which are the same inspections used to validate the IACS data. Since these data are already used in the project, it is not clear whether we would gain any new information from this data set. We have therefore not used CROME data in the current project, but anticipate it being available in the near-future for examination.