Land use changes resulting in the loss of farmland have been the focus of growing public concern and have led to the development of preservation policies worldwide. There is a need to develop adequate tools to monitor such changes in order to inform and improve farmland management policies.
The present study investigates to what extent the periodic updating of Norwegian land cover data (as part of AR5, the Norwegian land capability classification system and dataset) can satisfy the need to monitor farmland decrease. GIS analysis was performed for two municipalities. Changes during
a c.10-year period were analysed and uncertainties in the results explored. The analysis reveals that farmland loss due to reversible processes is far more important than loss through irreversible processes such as development. Uncertainty related to the identified changes is found to be small
enough to leave no doubt about the prevailing change processes in the study area, yet could be of great importance for the most minor change processes. Uncertainties in need of further consideration relate to ambiguous categories, such as regrowth processes. Further modelling of border uncertainty
and classification uncertainty is needed to understand fully the importance of uncertainty.