D6 - Remote sensing based modelling of continuous environmental variables for ecological and socio-economic scenarios

Abstract
During the 3rd phase the results of the experimental studies of the 2nd phase will be used to work towards the integration of algorithms and methods into operational workflows for the computation of land cover variables from satellite remote sensing data in tropical regions. Within this framework, emphasis will be on the (near real-time) mapping of land cover heterogeneity and variability of land cover variables over space and time in the context of climate change (ENSO) as well as ecosystem biodiversity and productivity.

Summary
Anthropogenic land cover degradation and conversion as well as long term climate change are the main factors driving changes in ecosystems. Satellite based estimations of vegetation status comprise an essential data source for the spatially explicit assessment and modelling of ecosystem dynamics at the local, regional and national level. The planned work for the 3rd phase contributes to this crucial topic and examines the following tasks:

(1) The vegetation productivity of tropical rain forests constitutes a key variable in tropical terrestrial ecosystem valuation. This productivity may be affected by ENSO-related interannual climate changes. The main research activity of the sub-project will be the remote sensing based computation of biophysical ecosystem variables (fraction of vegetation cover, albedo, FPAR) for the modelling of the spatial and temporal dimension of ecosystem productivity in a tropical landscapes. These variables will help to explain the dynamics of natural and agroforest ecosystems and will contribute to the spatially explicit modelling of ecosystem fluxes. Thus, this research activity is closely linked to sub-project B1. The parameterization is based on a time-series analysis of medium resolution satellite data (MERIS, MODIS; Spot-VEGETATION, Envisat/ASAR) and meteorological data. The data will serve as input variables for the modelling of energy and mass fluxes using coupled SVAT/SPM-3D and 3-D ABL models. Further to this, a simple satellite based vegetation photosynthesis model (VPM) for estimating climate related dynamics of vegetation productivity in a humid tropical evergreen forest will be tested and validated against the models of B1.

(2) Aboveground biomass stocks are a key parameter in assessing the economic value and conservation of land surfaces as well as for the estimation of energy and mass fluxes in terrestrial ecosystems. The above ground biomass is closely related to the characteristics of structural vegetation parameters. Modelling of canopy structure parameters will be supported based on high and medium resolution optical and SAR satellite data (ALOS-PALSAR; TerraSAR-X).

(3) The knowledge of the spatial distribution of landscape patterns and land cover change is a prerequisite for the modelling of socio-economic as well as ecological scenarios within multi-disciplinary projects. Hence, the routinely mapping of land cover type and land cover change at the local and regional level will be supported by D6. This will be enabled using available operational SAR as well as optical satellite data and through the integration of enhanced data fusion and analysis methods.

Overall, D6 contributes to all three foci. The generation of spatially explicit quantitative land cover variables constitutes a mandatory prerequisite for an improved modelling of socioeconomic and ecological scenarios (focus 1), for the assessment and modelling of ENSO- related changes of vegetation parameters (focus 3) and for the spatially explicit modelling of land use intensity gradients and ecosystem services at multiple scales (focus 2). The land cover type mapping and land cover change analysis is required by all sub-projects that work on spatially explicit modelling of economical or ecological variables.