Scenarios are widely used tools in both assessing the potential impacts of future uncertain conditions and in informing adaptive management. Yet, traditional “top down” scenario approaches face criticisms when they rely on small numbers of prespecified scenarios that can inadvertently exclude consequential dynamics, extremes, and diverse stakeholder impacts. “Bottom up” exploratory approaches, on the other hand, advocate for the use of ensembles that investigate large numbers of hypothetical futures to discover the ones most consequential to a system and its stakeholders. These methods introduce more rigor into how potential uncertainty is explored, but struggle with conveying actionable information and guiding follow-on adaptive actions. This talk introduces the FRamework for Narrative Storylines and Impact Classification (FRNSIC; pronounced “forensic”) bridging the gap between these two approaches. FRNSIC uses hierarchical classification sets to discover scenario storylines within a broad exploratory ensemble, that summarize both critical impacts and the consequential system dynamics that produce them. We demonstrate FRNSIC on the Upper Colorado River Basin, focusing on decadal drought conditions and their impacts. We show how scenario storylines discovered using FRNSIC can be used to investigate potential consequences to the basin’s water users as well as the basin’s downstream deliveries to other states.