Carbon sequestration, biodiversity and social structures in Southern Amazonia: models and implementation of carbon-optimized land management strategies

Climate change will increase precipitation variability – i. e. extreme events like droughts will occur more often also in the tropics and mean temperature will ultimately increase. Land use intensification is associated with (a) losses of ecosystem services like the loss of natural vegetation and associated ecosystem functions in the global and regional climate system, an (b) increasing releases of greenhouse gases (GHG), and (c) the reduction of livelihoods. This project aims at providing interdisciplinary solutions for these problems. Three regions along the land use frontier of Southern Amazonia were selected: Southern Pará: most active deforestation; Northern Mato Grosso: young soy bean production; Central Mato Grosso: established cultivation (>20 years) and adapted mechanized cropping (e.g. no till). Analyses focus on soil carbon (C) turnover, climate, ecosystem functions and socio-economic processes. Simulation models will be combined as software packages to support the decision-making process based on field and acquired data, including a step-by-step up-scaling from local to landscape and regional scale. All research and implementation activities include direct involvement of the stakeholders. Furthermore, joint field experiments for improving C storage and ecosystem functions will be performed in tight cooperation with an NGO founded by the farmers' organization of Mato Grosso. A combined computer-based decision support platform will be developed, including simulation models to run region-specific impacts of different scenarios of land use options and climate change on GHG and C cycling. This will be a highly valuable tool for regional planning authorities. From the scenario calculations simplified versions (e.g. emission factors) will be made available as an easy-to-use decision support system for individual stakeholders. Results will be communicated directly to stakeholders, by human capacity building, and by promoting financially feasible, C-optimized land use techniques throughout tropical areas with similar conditions.
The target of the subproject on “Landscape scale land cover analysis and geodata management” is to analyse landscape scale LULCC to support decision-making for an optimized land management. We develop and apply a landscape-wide analysis approach integrating remote sensing and spatial modelling techniques to gain knowledge on how to mitigate existing and prevent future land use conflicts:
(1) Development of remote sensing-based analysis schemes to derive land use at high resolution (e.g. Landsat data) with regional coverage over the last 25 years
(2) Adaptation of machine learning and time series algorithms to cope with large datasets
(3) Development and application of remote sensing based indicators for assessing landscape patterns and its links with carbon sequestration potential at landscape level
(4) Spatially explicit modelling on landscape level to identify drivers and hot spots of change at the landscape scale
(5) Spatially explicit scenario-building of land use types according to different regional to sub-continental storylines
(6) Development of a spatial data infrastructure for the whole CarBioCial project including data management and web-based technologies for distributed data access in Germany and Brazil