Evaluating the consistency of inferred multi-actor vulnerabilities to agricultural water shortages through the use of rival framings

This presentation was originally presented at the virtual EWRI Conference, June 2021. 

Water resources planning for the future has long been challenged by the presence of deep uncertainty in how hydroclimatic, societal, and other factors will evolve. Bottom-up planning frameworks attempt to overcome this challenge by employing exploratory modeling, which assesses water resource systems under a broad range of plausible states of the world. Even though generating plausible scenarios in this manner helps avoid myopic planning, their implicit assumptions might have decision-relevant consequences by skewing which factors are identified as most important and our perception of system vulnerability.
This study illuminates these implications in a sub-basin of the Colorado River, supporting several hundred agricultural and other water users. As competition for water resources increases, Colorado State planning efforts carefully consider how future climate conditions and growing demands will manifest through the framework of prior appropriation and shape water allocation in the basin. Our analysis explores how employing different experimental designs to analyze the impacts of changing hydrologic and water demand conditions shapes our interpretation of water supply vulnerability for individual water users. Specifically, we consider the following narrative descriptions: water availability decreases as a result of a drier climate and irrigation demand increases in response, and, alternatively, that water availability decreases as a result of a drier climate so that the increasing stress has some water users leasing their rights to other uses and, as result, irrigation demands decrease.
Each experimental design is paired with magnitude-varying sensitivity analysis and scenario discovery to assess the changing factors’ importance in shaping water shortages experienced by individual users in this basin. Our results show that these different assumptions behind each design can indeed bias the conclusions drawn on the users’ robustness and sensitivity to different factors. These differences can be highly consequential as they can be used to inform water conservation policy in the state.