B43K-07
From Dynamic Global Vegetation Modelling to Real-World regional and local Application

Thursday, 17 December 2015: 15:10
2006 (Moscone West)
Joerg Steinkamp1,2, Matthew Forrest3, Katharina Kamm4, Marion Leiblein-Wild3, Adrian Pachzelt3, Christian Werner3 and Thomas Hickler3,5, (1)BiK-F, Frankfurt Am Main, Germany, (2)Senckenberg, Frankfurt, Germany, (3)Senckenberg Biodiversity and Climate Research Centre (BiK-F), Frankfurt/Main, Germany, (4)Goethe-University Frankfurt am Main, Department of Ecology, Evolution and Diversity, Frankfurt/Main, Germany, (5)Institute of Physical Geography, Goethe-University Frankfurt am Main, Frankfurt/Main, Germany
Abstract:
Dynamic (global) vegetation models (DGVM) can be applied to any spatial resolution on the local, national, continental and global scale given suitable climatic and geographic input forcing data. LPJ-GUESS, the main DGVM applied in our research group, uses the plant functional type (PFT) concept in the global setup with typically about 10-20 tree PFTs (subdivided into tropical, temperate and boreal) and two herbaceous PFTs by default.

When modelling smaller spatial extents, such as continental (e.g. Europe/North America) national domains, or individual sites (e.g. Frankfurt, Germany), i.e. the scale of decision making, it becomes necessary to refine the PFT representation, the model initialization and validation and, in some case, to include additional processes.

I will present examples of LPJ-GUESS applications at the continental to local scale performed by our working group including i.) a European simulation representing the main tree species and Mediterranean shrubs, ii.) a climate impact study for Turkey, iii.) coupled dynamic large grazer-vegetation modelling across Africa and, iv.) modelling an allergenic and in Europe invasive shrub (Ambrosia artemisiifolia), iv.) simulating water usage by an oak-pine forest stand near Frankfurt, and v.) stand specific differences in modelling at the FACE sites. Finally, I will present some thoughts on how to advance the models in terms of more detailed and realistic PFT or species parameterizations accounting for adaptive functional trait responses also within species.