Land cover change: a method for assessing the reliability of land cover changes measured from remotely-sensed data

A method for assessing the reliability of land cover changes measured using data from different dates is presented and tested using a study area in the Cairngorm mountains, Scotland. Two sources of uncertainty are recognised in the change detection process: a) slivers resulting from misalignment of boundaries of land cover polygons and b) false positive changes associated with misclassification error in production of the land cover maps.