Numerous climate futures are now available fromglobal climate models. Translation of climate data such as precipitation and temperatures into ecologicallymeaningful outputs formanagers and planners is the next frontier. We describe a model-based open platform to assess vulnerabilities of agricultural systems to climate change on pixel-wise data. The platformincludes a simulationmodeling engine and is suited towork with NetCDF format of input and output files. In a case study covering a region (Auvergne) in theMassif Central of France, the platformis configured to characterize climate (occurrence of arid conditions in historical and projected climate records), soils and human management, and is then used to assess the vulnerability to climate change of grassland productivity (downscaled to a fine scale). We demonstrate how using climate time series, and process-based simulations vulnerabilities can be defined at fine spatial scales relevant to farmers and land managers, and can be incorporated into management frameworks.