The NERC pilot Environmental Virtual Observatory (EVO) project is now well underway. It is, at this stage, only a pilot or proof-of-concept project but has ambitions to use cloud technology to provide a platform for observational data and model results across scales, across applications and across users from policy makers to interaction with the public (and not forgetting that it might also be useful to scientists in developing and testing methods and models by making observational data more readily available).
“Storybooks” for how different types of application might work within the EVO are already being tested with different levels of stakeholders and some of these are relevant to CCN, including the impact of land management on flood runoff generation and nutrient transfers. As yet these use relatively simple models to explore what might happen under different future scenarios. This allows stakeholders such as farmers to explore what impact their management actions might have on the downstream catchment.
This is relatively simple to set up for the pilot project as a demonstration of the potential functionality of the EVO but there is also an interesting science question that then arises about how far such predictions can be made robust for different conditions around the country. There are, after all, relatively few detailed experimental studies of land management impacts at field or small catchment scales, and those have proven somewhat difficult to model. The impacts might indeed depend on the particular hydrological conditions during the period of the experiments.
So there will be uncertainty about representing the impacts of such changes in these data rich sites, but most of the landscape will be data poor. How far therefore should the knowledge exchange process in the EVO platform reflect these uncertainties, and how far should they be hidden from users so as not to confuse or lose the message? The EVO pilot project has about a year to run and so there will be only limited time to explore such issues. If the full project goes ahead, however, such questions will be crucial to how the platform should be used. It is a real opportunity to explore some of the uncertainties and consequent risk impacts associated with both observations and model outputs.
