Process-oriented diagnostics of dynamical coupling between the troposphere and stratosphere in Earth System Models

Stratosphere-troposphere coupling has pronounced effects on climate variability on intraseasonal to decadal timescales. Although climate models have steadily improved their representation of the stratosphere in recent decades, they still struggle to realistically simulate interactions between the stratosphere and the underlying troposphere. This is problematic because the responses of the climate system to anthropogenic and natural forcings, as well as internal climate variability such as ENSO or the QBO, are partly mediated through the stratosphere. Hence, high-fidelity representation of stratosphere-troposphere coupling is not only critical for understanding observed climate variability and trends (e.g., ozone recovery), but also for constraining future climate projections (e.g, the future polar vortex response). In this light, we have developed a series of “process-oriented diagnostics” (PODs) that can be used to assess the fidelity of key dynamical stratosphere-troposphere coupling processes in climate models. We are using these PODs, which are publicly available through the open-source Model Diagnostics Task Force Framework, to systematically benchmark and evaluate the impact of different model characteristics and configurations on the representation of stratosphere-troposphere coupling in models participating in CMIP6 over the historical record. This presentation will summarize these PODs, and show some preliminary results from their application to CMIP6 models.