The natural and built environment form a complex system, comprised of many interrelated subsystems, each interacting in multiple nexus. Manifestations of these interactions can be seen in complex events. Climate change, natural disasters, military conflicts, pandemics, and other such events require accurate preparation, preparedness, and response planning, in a fast, ever changing context. With exascale ($$10^{18}$$
10
18
floating point operations per second) computational levels reached, computing power gives us the capability to model and simulate complex scenarios. This capability gives decision makers tools to game possibilities and enact preparatory and preventative measures to build resilience. There has been a trend of development of decision support, risk assessment, and operational forecasting systems to address this issue, aggregating diverse data sources onto unified platforms. Nonetheless, the majority of such platforms focus on the aggregation of just data and not models, and remain in silos of disciplines. What is needed to prepare and plan for disruptive events is a move towards decision support based on holistic, integrated, model-based analysis. While modeling individual systems has been done for many years, modeling in holistic analysis presents additional challenges. This paper presents an overview of the challenges and advances present for a move to a model-based holistic analysis, and an evaluation of some platforms currently in development and operation. The present work signals gaps in research to be addressed. Finally, we formulate base requirements for the development of systems to perform holistic model-based analysis, and discuss future plans for the development of such a platform.