A number of data sources are collected or are developed from primary sources of data to input into the crop map model. All of them have a spatial element to them, that is the data collected is known or tagged for a geographic point. All data collected and developed is available either freely to the public or, as is the case with Field Boundaries data, is available under license and on request.
Some summary information on the data used is included in Table 1 below.
|Data||Purpose||Format||Time-frame||Primary source||Secondary source|
|Satellite||To produce zonal statistics from radar backscatter used in the machine learning model||tif||Mar-Oct 2019||JNCC||Ceda Archive|
|Ground Truth||Used to train and test the machine learning model||shp||2019||Rural Payments and Inspections Division (RPID)||N/A|
|Land parcels (field boundaries)||To assign areas within radar images to field areasto produce zonal statistics for individual fields.||shp||2019||Ordnance Survey||N/A|
|Spatial Data Masks||Exclusion of land parcels which intersect with the National Forest Inventory.||shp||2018||National Forest Inventory||N/A|
For any further queries please contact firstname.lastname@example.org