Climate, Variability, and Climate Sensitivity of “Middle Atmosphere” Chemistry Configurations of the Community Earth System Model Version 2, Whole Atmosphere Community Climate Model Version 6 (CESM2(WACCM6))

Abstract

Simulating whole atmosphere dynamics, chemistry, and physics is computationally expensive. It can require high vertical resolution throughout the middle and upper atmosphere, as well as a comprehensive chemistry and aerosol scheme coupled to radiation physics. An unintentional outcome of the development of one of the most sophisticated and hence computationally expensive model configurations is that it often excludes a broad community of users with limited computational resources. Here, we analyze two configurations of the Community Earth System Model Version 2, Whole Atmosphere Community Climate Model Version 6 (CESM2(WACCM6)) with simplified “middle atmosphere” chemistry at nominal 1 and 2° horizontal resolutions. Using observations, a reanalysis, and direct model comparisons, we find that these configurations generally reproduce the climate, variability, and climate sensitivity of the 1° nominal horizontal resolution configuration with comprehensive chemistry. While the background stratospheric aerosol optical depth is elevated in the middle atmosphere configurations as compared to the comprehensive chemistry configuration, it is comparable among all configurations during volcanic eruptions. For any purposes other than those needing an accurate representation of tropospheric organic chemistry and secondary organic aerosols, these simplified chemistry configurations deliver reliable simulations of the whole atmosphere that require 35% and 86% fewer computational resources at nominal 1 and 2° horizontal resolution, respectively.

Key Points

There are differences in stratospheric aerosol optical depth between comprehensive and simplified chemistry configurations

Simplifying the chemistry scheme generally has smaller global impacts than coarsening the horizontal resolution

All configurations have similar climate sensitivities and responses to forcings

Plain Language Summary

Modeling the entire atmosphere, or at least from the surface to an altitude of 140 km (87 miles), takes a lot of computer resources. Simulating 1 year can require the equivalent of thousands of personal computers running for 1 day, for example, which is only realistic for researchers with access to a supercomputer. There are many people who would like to simulate the whole atmosphere to study climate change, space weather, and extreme events, but even with access to a supercomputer, it is still computationally expensive to run. We examined a whole atmosphere model with a simpler chemistry scheme, and at a lower horizontal resolution, to see if it still reproduces major features of climate and climate change. The two configurations perform similarly to the high resolution simulation with complex chemistry, with some minor and understandable differences. Anyone looking to simulate the whole atmosphere, using fewer computational resources, can do so confidently using the described model configurations, as long as they are aware of some of the deficiencies.