Model fusion at the British Geological Survey: experiences and future trends

Abstract

The British Geological Survey (BGS) is developing integrated environmental models to address the grand challenges that face society. Here we describe the BGS vision for an Environmental Modelling Platform (BGS 2009), that will allow integrated models to be built and describe case studies of emerging models in the United Kingdom. This Environmental Modelling Platform will be founded on the data and information that BGS holds. This will have to be made as accessible and interoperable as possible to both the academic and stakeholder decision making community. The geological models that have been built in an adhoc way over the last 5-10 years will be encompassed in a National Geological Model which will be multi-scaled, beginning with onshore United Kingdom and eventually including the offshore continental shelf. The future will be characterised by the routine delivery of 3D model products from a multi-scaled and scalable 3D geological model of the UK which can be dynamically updated. The deployment of this model will generate further significant requirements across the Information and Knowledge Exchange spectrum, from applications development (database, GIS, web and mobile device), data management, information product development, to delivery to a growing number of publics and stakeholders.There is now a growing realisation in the environmental and social sciences that to address the grand challenges that face the world a whole system approach is required. These challenges including climate change, natural resource and energy security and environment vulnerability raise multi-and transdisciplinary issues that require integrated understanding and analysis. Not only must we model the whole physical Earth system, bringing together climate, ecological, hydrological, hydrogeological, and geological models to name but a few, we must link them to socio-economic models. Model fusion may well be the only adequate way to provide the necessary coupled processes framework through which predictions and planning or management decisions can most appropriately be made.A scoping study (Giles et al 2010) assessed the current situation and made some preliminary recommendations in order to create a more integrated and semantically harmonized future in environmental modelling. The only viable option is a 'linked models' approach which enables models to pass parameters between each other at runtime. This solution can bring together the best and most appropriate scientific models and allows the various scientific disciplines to continue the development of their current models. This linking approach is also relevant for integrating models that have been largely built and optimised individually, with appropriate configuration. The European Union has funded multi-national, multi-disciplinary research into 'linked modelling', using the Open Model Interchange (OpenMI) standard. This software used, in conjunction with critical underpinning activities such as data management, semantics and ontologies, understanding of uncertainty and vi...

Introduction to integrated environmental modelling to solve real world problems: methods, vision and challenges

“…For example, a common theme in the experience of the British Geological Survey (Peach et al 2016) and the Environment Agency in the UK (Farrell et al 2016) is the requirement to be able to utilize and link existing models that have often required significant investment in resources to create, and so to establish methods of linking together and adapting existing modelling systems (Sutherland et al 2014). A key area of research highlighted by several authors is the building upon a well-developed data and information framework in which appropriate standards of data quality and semantic interoperability are maintained (Sutherland et al 2014;Laxton 2016).…”

“…Whilst the resulting model is often more widely reported, the underlying data and information structure sometimes receives less attention in terms of organizational investment but forms the basis upon which successful integrated modelling is built. The need for linking mechanisms that cross discipline, and also national, boundaries has been particularly highlighted by Moore & Hughes (2016) and Peach et al (2016). Sutherland et al (2014) review various approaches to hydraulic modelling and discuss the historical development of linked modelling, exploring the implications of fusion and emphasizing that the fusion of models involves not only linking them together, but also providing easier access to information about the models and software tools to facilitate the linking process.…”