Sebastian Lehner, Christoph Matulla, Helfried Scheifinger

An important consequence of climate change is the impact on the seasonal cycle of vegetation flora and fauna. Although it is generally understood that anthropogenic mechanisms play a major role in the warming trend of the climate and that the timing of such phases, especially spring timing events, depends largely on the temperature, the link has yet to be quantitatively shown for different kind of areas on a regional scale, due to high intrinsic noise.
In this study, an end-to-end analysis to external forcings was carried out, linking the earlier timing of phenological spring timing events to changed climate conditions (increasing temperature) and this change in the environment to anthropogenic forcing, for the region of interest: Germany.
Besides a large ensemble originating from six different General Circulation Models (GCMs), driven with various kinds of forcings, the E-OBS data set was used as observational reference for near-surface air temperature and the PEP725 for phenological observations. The latter contains over 100 different phenological phases, from which - after quality checking - 12 phases are evaluated.
To overcome the scale differences, a quantile mapping bias correction approach was used to downscale the GCM data. The generation of simulated phenological time series was done with a temperature-day-sum model, which had to be calibrated and optimised beforehand.
The signal-to-noise ratio was increased by generating samples of 50-year running trends, which make up the basis for the statistical evaluation. U-values from the Mann-Whitney U-Test represent the foundation, on which the null hypothesis for detection, that the observed changes can be explained with naturally forced climate conditions only, was tested. Based thereupon, a newly defined consistency factor was constructed, which allowed the assessment of anthropogenic impact on phenological phases. Observed changes in phenological phases were successfully attributed to anthropogenic impacts.