Water mangers face significant challenges in planning for the coming decades as previously stationary aspects of the regional hydroclimate shift in response to global climate change. Providing scientific insights that enable appropriate use of regional hydroclimate projections for planning is a non-trivial problem. The system of data, models, and methods used to produce regional hydroclimate projections is subject to multiple interacting uncertainties and biases, including uncertainties that arise from general circulation models, re-analysis data products, regional climate models, hydrologic models, and statistical downscaling methods. Moreover, many components of this system were not designed with the information needs of water managers in mind.

To address this problem and provide actionable insights into the sources of uncertainty present in regional hydroclimate data products, Project Hyperion has undertaken a stakeholder engagement process in four case study water basins across the US. Teams of water managers and scientists are interacting in a structured manner to identify decision-relevant metrics of model performance. These metrics are in turn being used to drive scientific investigations to uncover the sources of uncertainty in these quantities. Thus far, we have found that identification of climate phenomena of interest to stakeholders is relatively easy, but translating these into specific quantifiable metrics and prioritizing metrics is more challenging. Iterative feedback among scientists and stakeholders has proven critical in resolving these challenges, as has the roles played by boundary spanners who understand and can speak to the perspectives of multiple professional communities. Here we describe the structured format of our engagement process and the lessons learned so far, as we aim to improve the decision-relevance of hydroclimate projections through a collaborative process.