The agricultural sector is among the most climate-sensitive sectors, because changes in temperature, precipitation and carbon dioxide concentrations of the atmosphere directly affect the growth conditions of agricultural crops. Farmers will adapt to changes, which will have substantial consequences for the structure and functions of agricultural landscapes such as water balance, biological diversity and many others. Added to that, this may also affect the climate itself. Because the reaction to changes is essentially local, adaptation strategies need to be developed taking into account regionally varying ecological, economic and societal factors and implemented on a regional scale. Current model systems, though, are not good enough to accurately simulate climate change effects on a regional scale as they do not consider relevant processes, especially the interactions between land use and climate, to a sufficient extent. It is the joint objective of the Research Unit to improve process understanding and knowledge of feedbacks between land surface and atmosphere by combining integrated modelling with hitherto unequalled spatial resolution, intensive field measurements and controlled experiments. This will enable us to produce projections of landscape development and potential adaptation strategies until 2030. For this purpose, high-resolution climate, land use, and crop growth models will be coupled with socio-economic models forming an advanced land system model. By means of specific sensor systems and remote sensing techniques, data on energy and matter fluxes between the soil-plant system and the atmosphere will be collected in two study regions in Southwest Germany. Crops will be exposed to future CO2 and climate conditions in climatic chambers in order to assess the effects on crop yield and yield quality. Data obtained from field measurements and climatic chamber experiments as well as from socio-economic investigations in the study regions will serve to improve model components and to validate the novel land system model.