Climate change threatens to impact many systems critical to human welfare – including energy, water, food and agriculture. These systems also depend on one another in complex ways, introducing the potential for cascading or compound risk. To understand the impact climate change will have on human systems, it is important to account for both interconnected sectoral impacts as well as underlying uncertainty. The challenge lies in including both in a single study; there are computational and practical tradeoffs in including additional complexity. This study lays out some of the key strategies and challenges for uncertainty analysis in the literature on climate impacts to water and energy systems. We find that uncertainty arises at each step of the modeling process – emissions scenarios, climate projections, downscaling results, and impact modeling. Many strategies for uncertainty inclusion involve running additional models and scenarios, introducing additional complexity. As a result, many studies choose to limit one dimension – sectoral or uncertainty – in order to more completely address the other.

Recognizing these challenges and tradeoffs, we make several recommendations for better inclusion and communication of uncertainty in multi-sector dynamics studies. Specifically, we provide distinct approaches for improved decision-making and more general modeling. For the latter, we frame a generalized, simplified modelling approach to better evaluate the many forms of structural, parametric and deep uncertainty explored in this study. This approach can complement more detailed modeling and provide insights on the key relationships and uncertainties.