Dorothee Kolberg

In Norway, agroclimatic zones (ACZs) are a valuable tool for national analyses in subject areas concerning the optimized management of agricultural land resources. However, current Norwegian ACZs have been criticized for having an outdated standard climate normal (1931–1960), a limited representation of the local climatic variation, a lack of important model parameters, and weak methodological documentation. Therefore, this paper presents new ACZs for Norway that address these weaknesses. The most significant methodological updates are the use of the standard climate normal of 1991–2020, additional weather data variables, the downscaling of weather data to 250 m hexagons, and the incorporation of phenological crop models for spring wheat, spring barley, and forage grass. The grass model was calibrated with the number of grass harvests at research stations, while the grain models were calibrated with subsidy claim data. The modeled zones for the three crops were combined into the general ACZs. Example maps of the crop zones and new ACZs for the selected regions and the whole country are presented. The new ACZs are more robust, agronomically relevant, and better aligned with the current climatic conditions in Norway. The deliberate exclusion of factors other than climate ensures the new ACZs’ national comparability and their applicability in policy development, land-use planning, climate adaptation, and agronomic assessments at the national scale.

Peat inversion is a management technique used to reduce emissions and retain carbon in cultivated peatland while allowing for effective forage production. Although maps and land registers document the presence of cultivated peatland that is suitable for peat inversion, these data do not cover all regions of interest. This study explores how an expert system and geostatistical modelling can be used to identify cultivated peatland suitable for peat inversion. The expert system proved to work moderately well for cultivable (but not for cultivated) peatland. Geostatistical modelling, using cultivable peatland as statistical support, gave good results in regions with large, continuous landforms. The results were less accurate in regions with rough, rapidly shifting terrain forms and where peatland was less frequent. The difference could be seen in the range and shape of the semivariograms. Geostatistical modelling can be used to identify cultivated peatland suitable for peat inversion in regions where the semivariogram shows a clear and well-defined spatial autocorrelation structure.