@the 10th iEMSs conference, Modelling for Environmental Sustainability, 6-10th July 2020, Brussels. https://iemss2020.com/workshops/
Organizers:
- Gary Polhill (The James Hutton Institute, UK),
- Doug Salt (The James Hutton Institute, UK),
- Matt Hare (The James Hutton Institute, UK) Kit Macleod (The James Hutton Institute, UK),
- Ioannis Athanasiadis (Wageningen University & Research, The Netherlands),
- Ferdinando Villa (Basque Centre for Climate Change, Spain)
- Bruce Edmonds (Manchester Metropolitan University, UK),
- Geo Kocheril (University of Kassel, Germany),
- Andrew Bell (New York University, USA)
Time and time again, as interdisciplinary teams of academics (and sometimes non-academics) are brought together to address an urgent systemic problem, the issue of knowledge and data integration arises. Despite several years of work on various integrated modelling toolkits and workflows, proposals are still submitted with the “Work Package 5” problem: a package of proposed work timetabled for the last six months of the project, usually named “integration”, that is given the unenviable task of tying all the other work together. Integrated modelling is a subset of this problem, but brings the issue into the sharpest focus, because writing software requires you to be specific and detailed: syntax errors, segmentation faults and NullPointerExceptions cannot be elided, brushed-aside or wished away. The software either works, or it doesn’t. If the project co-ordinator is lucky, a naive computer scientist will claim they can automate the integration with a fancy new tool they’ve been working on, and everyone else breathes a sigh of relief, because all their knowledge, data and legacy code is going to be reusable through a “magic wand”, and they can get on with the work they have always done (Athanasiadis 2017).
Integration First is about putting knowledge, data and software integration at the heart of interand trans-disciplinary projects aiming to address systemic issues in coevolving social and environmental systems, rather than adding it as an afterthought. It is about recognizing that such integration needs to be embedded into all tasks, and given time and energy by all participants, and that cultural change and new leadership is needed (Gordon et al. 2019). It also emphasizes that integration is itself a research endeavour, needing science and funding to support it, and to create new methods and tools (Bammer 2017). It is axiomatic to the concept of Integration First that integration can be done badly even if the software compiles and runs, with any resulting model giving potentially misleading information to policymaking processes (Voinov and Shugart 2013).
This workshop will use various exercises to elicit and then discuss the research culture issues that have led to the integration ‘Groundhog Day’ in social-environmental systems modelling, and to consider what roles software might have in facilitating change and supporting integration. We anticipate a position paper to be prepared as an outcome of the workshop.