Article:

Lischke, H., Guisan, A., Fischlin, A., Williams, J. & Bugmann, H., 1998.  
        Vegetation responses to climate change in the Alps: modeling studies.  
        In: Cebon, P., Dahinden, U., Davies, H.C., Imboden, D.M. & Jaeger, C.C.  
        (eds.), Views from the Alps: regional perspectives on climate change.  
        MIT Press, Boston, Massachusetts a.o., pp.  309-350.

Abstract:

Model Validation

Validation of dynamic forest models requires the combination of different approaches, which all together test the transient and equilibrium behavior of the model under a sufficiently large spectrum of conditions.

From the case study of model validation using the Soppen-see pollen proxy data we concluded that paleoecological validation of forest ecosystem models offers on one hand promising potentials, yet suffers at the same time from strong limitations.

The potential of paleoecological model validation is due to the following reasons: Paleoecological proxy data are the only data sources that offer long-enough time series of unmanaged forest dynamics. Moreover, with paleoecological data we can validate forests under a climate that changed at various rates, so increasing our confidence that the model behaves realistically under scenarios of future climatic change.

There are still strong limitations to a paleoecological model validation, as for instance became evident from the many uncertainties outlined in the present study. A ÒfailureÓ of the validation may be due to errors in the ecosystem model, e.g. wrong parameter values, as well as due to errors in the reconstruction of the past climate or the past vegetation. Further uncertainties remain because of the unknown nature of the pollen source area in the past.

This study also revealed the need for improvements and possible directions of progress:

First, an independent, high quality record of the past climatic input conditions is as much needed as the good record to be used for the actual comparisons with the model behavior. Second, a process-oriented model of pollen production, transport, and sedimentation appears to be worth-developing and worth-studying. It could be used to replace the transfer functions often used to reconstruct past vegetation compositions. Thirdly, models of tree migration are highly needed, since migration potentially plays a key role in the transient forest responses to climatic change, especially during phases of strong climatic changes such as the end of the last ice age and as projected for the following century.

The Fate of Forests in an Alpine Region

Due to current data limitations respectively not yet fully developed methods of generating bioclimatic scenarios at any ecological interesting site, our results remain restricted to the few presented case studies. Yet, some general, tentative conclusions could be drawn:

While some forests may in the Alps slightly profit from the envisaged climatic changes or will not be affected at all, some may suffer drastically. Hence, no uniform, simple response of the mountain forests to a climatic change is to be expected. Biggest sensitivities were found in areas which border a continental climate already in the present and in sub-alpine conditions. All these conditions can be traced back to the fact that the studied locations are within mountains. Especially when these findings are contrasted with what we learned about the peripheral forests at lower altitudes, we concluded the following: Mountain forests are among those forests which are especially sensitive to climatic change.

Full text 2.8 MB