Skip to main content
U.S. flag

An official website of the United States government

Using Earth System Models in a hierarchy of models to more effectively address climate and environmental security applications

Presentation Date
Wednesday, January 11, 2023 at 9:45am - Wednesday, January 11, 2023 at 10:00am
Location
Colorado Convention Center - 210/212
Authors

Author

Abstract

Earth System Models (ESMs) include a significant level of physical complexity to faithfully represent most physical processes and must carefully balance resolution requirements with available computational resources. However, climate and environmental security problems often require very rapid answers, at very small scales, making use of ESMs difficult. While models like the DOE’s Energy Exascale Earth System Model, which features variable resolution in every model component, expand the possible national security applications for ESMs, its resolution is still limited. E3SM is unlikely to be able to analyze regions at resolutions below a few kilometers. Further, given the time it takes for ESM simulations, even on the most advanced computational resources, approaches allowing for rapid turnaround are needed to address climate and environmental security problems. Instead of solely relying on ESMs, we must instead consider them as one of a wide array of tools available for answering pressing climate and environmental security questions. Herein we describe three environmental security applications that illustrate the value of a model hierarchy. In the first, we leverage the full Coupled Model Intercomparison (CMIP6) archive to analyze changes in Arctic shipping lanes as the climate changes, with the present-day climate used as a metric for downselecting models used in the analysis. We then define a risk index that combines sea ice coverage with vessel characteristics (e.g. hull strength). In the second, we present a workflow that combines statistically derived or highresmip simulated hurricanes that are used to drive a single layer ocean model to accurately and quickly predict coastal inundation at very fine scales (O(10m)), something not possible within ESMs alone. In the final case, we examine the importance of robust methods to initialize ESMs from longer running CMIP6 simulations or data assimilated products for environmental security applications, using the Arctic as an initial use case.

Funding Program Area(s)