Abstract: Numerous climate futures are now available from global climate models. Translation of climate data such as precipitation and temperatures into ecologically meaningful outputs for managers 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 platform includes a simulation modeling engine and is suited to work with NetCDF format of input and output files. In a case study covering a region (Auvergne) in the Massif Central of France, the platform is 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. (C) 2015 Elsevier B.V. All rights reserved.