Environmental scientists need to make predictions that are increasingly cross-disciplinary, bringing together observations and models in both physical and biological systems, and visualising the results. Observations can be from multiple platforms, and there are often many competing models that could be used. At the same time, catchment managers and policy makers face a challenging future trying to ensure a wide range of ecosystem and hydrological services are delivered from increasingly constrained budgets whilst complying with a range of regulation requirements. There is also a greater requirement for transparency and access to data and making regulatory decision making processes visible to the public. The Environmental Virtual Observatory Pilot project (EVOp) is a new initiative from the UK Natural Environment Research Council (NERC) designed to explore new tools and approaches to support these challenges. The long term vision of the Environmental Virtual Observatory is to: - Make environmental data more visible and accessible to a wide range of scientists and potential users including for public good applications; - Provide tools to facilitate the integrated analysis of data to give greater access to added knowledge and expert analysis and to visualisation of the results; - Develop new, added-value knowledge from public and private sector data assets to help tackle environmental challenges. The EVO will exploit cloud computing to give a shared working space for data, models and analysis tools; in this two year pilot project we will develop five local and national exemplars to demonstrate and test the opportunities and constraints from such an approach. The question-based exemplars being developed are focused on (i) management options for flooding and diffuse pollution at local and national scales, (ii) approaches for transferring hydrological models for both flooding and drought from data rich to data poor areas and (iii) defining the uncertainty bounds of current climate change predictions on change in soil carbon at a global scale. By developing exemplars focussed on some major environmental questions at a local, national and global scale we are able to directly test issues such as data assimilation, adapting and linking models to work in a cloud environment, and portal design for a wide range of end-users. New international standards for model exchange and exchange of analysis tools are desirable to supplement the emerging data exchange standards, and the EVOp will make a contribution here. A working prototype portal will be delivered in December 2012 that examines these issues, for a possible next phase.