From ad-hoc modelling to strategic infrastructure: A manifesto for model management

Models are playing an increasingly prominent role in watershed management, and environmental management more generally. To successfully utilize model-based tools for governing water resources, modelling timelines must match decision making timelines, and modelling costs must fall within budget constraints. Clarity on management options for modelling processes, and effective strategies, are likey to improve outcomes. This paper provides a first conceptualisation of model management and lays out its scope. We define management of numerical models (MNM) as governance, operational support, and administration of modelling, and argue that it is a universal activity that is crucial but often overlooked in organizations that rely on modelling. The paper lays out the leverage points available to a model manager, based on a review of model management practices in several fields, highlights lessons learned, and opportunities for further improvement as model management becomes a mainstream concern in both research and practice. •Emerging explicit knowledge around managing numerical modelling (MNM) processes.•MNM improves return on investment in model-supported decision infrastructure.•Leverage points are procedural guidelines, cyberinfrastructure, knowledge management.•Blue-print approaches to MNM are not expected to be successful.•Consideration of context important - an agency's roles, capacity, and resources.