Research project SR/00/135 (Research action SR)
Vegetation is a major carbon sink and is as such a key component of the international response to climate change, caused by the build-up of greenhouse gases in the atmosphere. However, anthropogenic disturbances like deforestation or fires are the primary mechanism that changes ecosystems from carbon sinks to sources, and are hardly included in the current carbon modelling approaches. Moreover, in tropical regions, theseasonal/interannual variability of carbon fluxes is still uncertain. In the context of climate change and mitigationpolicies like REDD, it is particularly important to be able to quantify and forecast the vegetation dynamics andcarbon fluxes.
Objective
In this purpose, the overall objectives of this research is to dynamically assimilate the land surface characterisation obtained from long SPOT-VEGETATION time series (e.g. plant functional type (PFT), phenology, Leaf Area Index (LAI), land cover change) into the ORCHIDEE global vegetation model, which simulate vegetation dynamics and carbon balance, in order to improve the forecast of the terrestrial Carbon cycle in tropical regions under different anthropogenic forcings. Such approach will allow us to determine whether the African terrestrial carbon balance will remain a net sink or could become a carbon source by the end of the century, according to different climate-change and deforestation scenarios. The challenge of this research is to bridge the gap between the land cover and the land surface model communities.
Method
POT-VGT time series will be used to derive a global land cover map depicting the LCCS classifier values and the PFTs, as well as the vegetation canopy phenology thanks to vegetation indices. This information at global scale is crucial for the land surface parameterisation of the ORCHIDEE model. After adaptation of the assimilation procedure, the model parameters will be optimized firstly for the Amazon basin for which both satellite and fluxnet data are available. An uncertainty analysis on the optimised parameter values will be performed and the model results will be compared to field measurements. Then, the knowledge gained from the simulations over the Amazon will be used for Central Africa where no field data are available. Thanks to Landsat images and a multivariate analysis identifying the driving forces, a spatially explicit deforestation and degradation model will be design, using Cellular Automat, for Central Africa.
This land cover change model coupled with the optimized ORCHIDEE will allow us to predict terrestrial carbon stock evolution and carbon fluxes. Errors estimations and results will be discussed in the light of mitigation policies.
Result
The first products of the research are a 1-km global reference data set characterizing the spectral response to vegetation canopy seasonality of each representative land cover for the main ecoregions throughout the world, and the optimized and validated ORCHIDEE model for Amazon and Central Africa. Then, a Cellular Automat land use land cover change model deforestation and degradation designed for Central Africa will generate land cover simulations based on actual observed conditions and extended in time and space through land cover change drivers. This model, coupled to optimized ORCHIDEE model, will provide quantitative estimates of the carbon stocks and fluxes in Central African forests relevant to scientists and decision-makers, for the current situation and predictions under 9 different scenarios for climate change and deforestation. It will allow us to quantify the potential impact of climate change and anthropogenic disturbances like deforestation on that fragile region as well as the potential effects of mitigation policy like the REDD. Indeed, results will be analysed in the light of the recent REDD effort to tackle deforestation in developing countries. Moreover, thanks to a great partnership between very complementary teams, the Belgian expertise will be enhanced and its networking reinforced. Finally, this research will contribute to bridge the gap between the Carbon community and Earth Observation community.
