Latitudinal forest advance in northernmost Norway since the early 20th century

Aim: The arctic forest—tundra boundary is expected to advance rapidly northwards in response to climate warming. The purpose of this study was: (1) to analyse region-wide changes in the positions of the latitudinal forest line and tree line since the early 20th century; (2) to calculate rates of northward movement and to estimate uncertainties from different sources of data; (3) to analyse how rates of advance differ between two common tree species, Betula pubescens (downy birch) and Pinus sylvestris (Scots pine); and (4) to place the results in the context of dynamic global vegetation models. Location: The study covers Finnmark county in northern Norway, which has warmed by 1—2 °C since the early 20th century. Methods: Historical forest maps, topographic maps, aerial photographs and satellite imagery were used to delineate region-wide, species-specific and time-specific subarctic forest lines and tree lines. Rates of advance were calculated using two geometrical methods, placing different emphases on the shape and mean position of the lines, but yielding similar results. Results: The analyses revealed an average northward advance of the birch and pine forest lines of 156 and 71 m year -1 , respectively. Both showed considerable spatial and temporal variation. The birch tree line showed the most pronounced advance (340 m year -1 ), whereas the pine tree line showed very limited advance (10 m year -1 ). Data sources contributed an uncertainty of around 10 m year -1 to these rates. Main conclusions: The analysis of historical and recent forest delineation data showed a very restricted advance rate compared to the predictions of dynamic global vegetation models. If these results are representative of the behaviour of the entire circumarctic forest—tundra zone over the present century, they suggest that the modelled prediction of the loss of 40% of the current tundra is a serious overestimate. A stronger focus on factors limiting the response of the forest—tundra ecotone to climate change is needed to refine the output from dynamic global vegetation models.