MARINE GEOLOGY AND GLACIOLOGY
|
|
Dynamic European Climate-VEGetation Impacts and Interactions - Models |
|
|
Contact: T. Giesecke tg@geus.dk |
MODELS
DVMs Driven by GCM Output
Aim:
To model palaeovegetation dynamics using GCM generated palaeoclimate.
The climate simulated by GCMs will be used to force DVMs in order to simulate European vegetation. We will use the available outputs from PMIP (Joussaume 1999) to force three vegetation models with which the group has expertise: IBIS (Kucharik et al, 2000), LPJ-GUESS (Smith et al. 2001, Sitch et al. 2003) and CARAIB (Warnant 1994; Otto et al. 2002). We will choose the runs that gave the best results over Europe to force the vegetation models and also use new GCM output in collaboration with LSCE, Gif-sur-Yvette. We will also apply a climate-correction technique to both correct for climate model biases and interpolate the simulated climate to the fine spatial resolution needed to simulate the vegetation.
Click on image to download in full size, JPG-file (340 KB)
Data-Model Comparison
Aim:
Assess model performance by comparison of modelled palaeovegetaion with two independent proxies - pollen and plant macrofossils.
Comparison of DVM-simulated with the observed vegetation will use two independent methods: the PFT/biome method (Prentice et al., 1996) and the Bioclimatic Affinity Groups (BAG) method (Laurent et al. in press). A developing migration module within LPJ will be tested against
palaeodata. Data-model divergence could indicate model weakness or the extent of cultural
impact. We will assess the extent of cultural impact using pollen indicators and independent archaeological data.
DVMs Driven by Pollen-derived palaeoclimate
Aim:
To model palaeovegetation dynamics using palaeoclimate generated from pollen data.
Holocene pollen data, when mapped over large areas, indicate the effects of climate forcing. The bioclimatic variables derived from pollen data will be used to drive three DVMs (LPJ-GUESS, IBIS and CARAIB) for short time-series simulations covering periods of rapid vegetation change. The purpose of these runs is to explore the wider ecosystem consequences of the changes in vegetation and investigate the feedbacks to the climate system using new estimates of palaeoclimate.The focus will be on the dynamics of the broadleaved-coniferous ecotones that occur in Scandinavia, and mountainous areas in south-east Europe.
Comparison of 3 DVM Performance
Aim:
To compare sensitivities of the DVMs to different factors and assess robustness of output.
Methodology will be designed to compare the three DVMs and assess their performance according to quantitative criteria (Cramer et al. 2001). For this comparison, experiments using exactly the same climatic forcings (GCM outputs or pollen-derived climate) and soil characteristics will be used.
Project Integration
Aim::
To finalise the most robust set of palaeoclimate data; finalise DVM modifications and improvements; generate predictions for future ecosystem dynamics using coupled GCM-DVM simulations.
A generalised comparison of all model output with pollen and plant macrofossil datasets will be made. We will make predictions for the future regarding hydrology, carbon allocation, albedo and woody plant biodiversity through coupled DVM-GCM runs. Two of the DVMs and IBIS are directly coupled to GCMs (Delire et al. 2002,2003). With fully coupled climate-biosphere models, the vegetation cover responds synchronously to changes in climate and these vegetation changes feed back directly to the climate. These coupled GCM-DVMs thus allow simulation of transition periods and investigation of the importance of vegetation-climate interactions. A series of coupled GCM-DVM simulations will be made to investigate the vegetation and climate interaction over Europe around 11000 years BP, 9000 years BP and 1000 years BP, periods that were not covered by PMIP. The consequences for model reliability when run in coupled mode for future predictions will be assessed.
For more information see:
Edited by:
Niels E. Poulsen, GEUS -
nep@geus.dk
