Interactive Stratospheric Ozone (O3v2) Module for E3SMv2
Accurate simulation of past climate evolution and projections of future climate relies on the correct representation of the greenhouse gases, including ozone. Simulating climate change driven by ozone is challenging for chemistry-transport modeling because ozone has two chemically distinct regions (stratosphere versus troposphere) with a very sharp interface at the tropopause. Running a detailed atmospheric chemistry model for ozone, including both stratospheric and tropospheric chemical regimes, within a climate model is costly, often prohibitively so. The U.S. Department of Energy's (DOE) Energy Exascale Earth System Model version 1 (E3SMv1) implemented chemistry-climate interactions through stratospheric ozone by incorporating linearized chemistry (Linoz v2), but prescribed tropospheric ozone based on climatology data, resulting in unphysical ozone distributions about the tropopause and uncertain climate impact.
The implication of the findings in this study is that the O3v2 module generally improves the model performance of ozone metrics and provides a computationally efficient way for Earth system models to include the climate impact from the interactive stratospheric ozone.
Stratospheric ozone affects climate directly as the predominant heat source in the stratosphere and indirectly through chemical reactions controlling other greenhouse gases. The U.S. Department of Energy's Energy Exascale Earth System Model version 1 (E3SMv1) implemented a new ozone chemistry module that improves the simulation of the sharp tropopause gradients, replacing a version based partly on long-term average climatologies that poorly represented heating rates in the lowermost stratosphere. The new O3v2 module extends seamlessly into the troposphere and preserves the naturally sharp cross-tropopause gradient, with 20 %–40 % less ozone in this region. Additionally, O3v2 enables the diagnosis of stratosphere-troposphere exchange flux of ozone, a key budget term lacking in E3SMv1. Here, we evaluate key features in ozone abundance and other closely related quantities in atmosphere-only E3SMv1 simulations driven by observed sea surface temperatures (SSTs, years 1990–2014), comparing them with satellite observations of ozone and also with the University of California, Irvine chemistry transport model (UCI CTM) using the same stratospheric chemistry scheme but driven by European Centre forecast fields for the same period. In terms of stratospheric column ozone, O3v2 shows reduced mean bias and improved northern midlatitude variability, but it is not quite as good as the UCI CTM. As expected, SST-forced E3SMv1 simulations cannot synchronize with observed quasi-biennial oscillations (QBOs), but they do show the typical QBO pattern seen in column ozone. This new O3v2 E3SMv1 model mostly retains the same climate state and climate sensitivity as the previous version, and we recommend its use for other climate models that still use ozone climatologies.