We demonstrated that the functional relationships within a complex multi-physics problem could be effectively represented as a network. In such a network, nodes represent parameters, values, and equations of the underlying physical models, while edges depict their connections. Thereby, the type and amount of available information significantly influence how uncertainty can be quantified and propagated, as well as the form in which this information is presented to decision-makers.

In a current research endeavor, we explore how network analytic methods could provide a more robust and efficient approach to characterize and propagate both aleatory and epistemic uncertainties in complex systems because these methods allow for a detailed representation of the complex interdependencies and dynamic interactions within the system. Network models explicitly incorporate the underlying physics and causal relationships, leading to more transparent and interpretable results. Furthermore, network science provides a comprehensive framework that captures both the detailed, localized, microscopic effects of uncertainties and their broader, system-wide macroscopic implications. 

Furthermore, we are advancing the state-of-the-art by facilitating long-term sustainability, which requires a shift from reactive to proactive planning. We aim to develop novel methods to support short-term management and operational decisions, as well as long-term strategic planning. Thereby, the ability to link models from different disciplines is crucial to quantify the risk and resilience of complex systems (e.g. against natural hazards) as well as for decision-makers which need to predict future scenarios to identify sustainable policies.