Key points are not available for this paper at this time.
Understanding the oceans requires accurate monitoring and suitable test environments for conducting hypothesis-driven simulations. Marine digital twins can be part of the solution to foster knowledge on the way to a sustainable blue economy. Herein we adapt the current literature on digital twins (DTs) in the industrial domain to the marine domain and reinterpret the dimensions of DTs to build a marine digital twin (MDT). These dimensions are the physical entity, the data, the services, and the virtual models. The latter represents the core aspect of DTs that bridges between the other dimensions. It consists of graphical models, physical models and especially methodologies of artificial intelligence as well as machine learning algorithms. The presented dimensions provide a conceptual view to identify missing components when building MDTs and provide researchers and developers guidance for developing MDTs. To exemplify these dimensions, we present two of our projects where MDTs are developed: NorthSat-X and AI4DTE. In NorthSat-X essential climate variables of the North Sea are incorporated into a MDT. Moreover, in AI4DTE a digital prototype is developed to monitor anomalies of offshore wind farms. At the end, we identified that MDTs face three main challenges that are subject of the paper's discussion: scalability, interoperability, and explainability.
Miedtank et al. (Thu,) studied this question.