Assessment of the TREELIM model in predicting present treeline along a longitudinal continentality-maritimity gradient in south-western Norway

ABSTRACT Treelines driven by climatic conditions are characteristic of alpine areas. This study assesses the degree of accuracy with which the TREELIM model predicts present treeline elevation along a regional continentality-maritimity gradient in southwest Norway. The study area consists of 15 sites chosen along this gradient, wherein observed treelines range from nearly sea level to over 1200 m a.s.l. The TREELIM model calculates treeline position based solely on meteorological parameters obtained from gridded climate datasets. The model assumes that tree growth occurs if the length of the growing season (LGS) exceeds a minimum duration and the seasonal mean temperature (SMT) does not fall below a certain value. Six different scenarios were developed in order to determine the effects of changing precipitation and drought within TREELIM. Within these scenarios, three different SMTs were used in order to model the treeline position. Model performance was evaluated by comparing resultant treeline elevation with observed values. Results of experiments using lapse rate-based temperature input data were compared to those using the gridded climate data. Statistical analysis determined that the lapse rate-based temperature scenario removing drought constraints and using a SMT of 6.4°C gave the most accurate prediction of treeline position, and a relatively uniform growing season ranging from 110 to 128 days. Precipitation and snow cover do not significantly affect treeline elevation. The results suggest that TREELIM accurately predicts the treeline in most cases, but is less suited in areas with steeply varying topography.