TEMPEX: Developing spatial layers of climatic temperature extremes: Final report in the BAFU-WSL program “Forests and Climate Change”

Abstract

In this project we calculated eight indices of temperature extremes and a compound index of frost frequency after leaf flush, which was tuned to four different species and calculated in a prognostic mode. The eight indices extracted from downscaled climate data, scaled to NFI plots (1 x 1 km NFI 1 network) in Switzerland for current and projected future climates, included the following: (1) the absolute annual minimum temperature; (2) absolute annual maximum temperature; (3) the annual number of frost days (Tmin < 0°C); (4) the annual number of frozen days (Tmax < 0°C); (5) the annually largest diurnal temperature range; (6) the mean annual diurnal temperature range; (7) the annually longest period of continuous frost days (Tmin < 0°C); (8) the annually longest period of continuous frozen days (Tmax < 0°C). All nine measures generated here indicate that the projected climate trends are affecting the temperature extremes as much as they do the temperature means. The extremes decrease at a rate comparable to temperature means, and the climatic conditions become gradually suitable to species that are less tolerant to low temperature extremes. By the end of the 21st Century, the climate barely reaches conditions represented as the cold limit of Mediterranean species (30-year mean of annual absolute minimum temperatures of ca. -6 °C on the Swiss Plateau). Since the climate is fluctuating quite considerably, and because the trajectory barely reaches this threshold towards the end of the 21st Century, it is not very likely that truly mediterranean species will already find suitable habitats then. Rather, sub-mediterranean and warm-temperate species will be the preferred choice for adaptive and active forest management practice. It remains to be considered that the indices presented here do not include drought, which represents another important constraint for the choice of suitable tree species. Caution is needed when interpreting the presented results. First, we only use data from one scenario (A1B), whereas there is uncertainty as to which of the usually four scenarios used is actually most suitable for describing the future trajectory of the climate. Second, we only use data from three regional climate models (CLM, RCA30, RegCM3) that are all fed by data from the same global circulation model (ECHAM5). This means that only a small variety of possible futures is explored. Finally, we had to request new versions of the obtained data from Meteotest due to errors found during the processing. While we think that most problems are solved now, we are not completely the currently used data is error-free, as we still found some patterns that are counterintuitive. Some of this may originate from the RCM data, though.