Increasing computational capacity afforded by new supercomputers combined with the ability to expend model resources in a geographically targeted region with variable-resolution meshes now affords the ability to simulate societally-impactful extreme weather events through a climate lens, termed storylines. These model configurations allow for an unprecedented look at hazards that require fine spatial resolution while maintaining a global, integrated, modeling framework.

In this talk, we exploit such configurations to use two state-of-the-art global Earth system models -- the Department of Energy's Energy Exascale Earth System Model (E3SM) and the National Center for Atmospheric Research's Community Earth System Model (CESM) -- to simulate two very different atmospheric extremes. In the first example, we reforecast the California 1997 New Year's flood under a variety of climate conditions at 14km, 7km, and 4km grid spacing in E3SM. The flood was initiated by a rain-on-snow flood event associated with an atmospheric river and highlights complex, non-linear interactions between model components and shows numerous benefits associated with higher spatial resolution and improved orographic representation. We also show results from 3km non-hydrostatic simulations of "synthetic" hurricanes extracted from a 25km climate ensemble using the Model for Prediction Across Scales (MPAS) within CESM. This technique permits tail events from larger, lower-resolution climate ensembles to be dynamically downscaled in a unified modeling framework, providing stakeholder relevant information for risk assessment. We discuss value added by such simulations and also highlight the potential for such simulations for use as stress tests within local decision-making frameworks.