The utility of indirect measures of the lower stratospheric residual circulation

The stratospheric residual circulation cannot be directly measured, hence various observable proxies have been used to indirectly quantify changes in the residual circulation on various timescales. However, the extent to which these proxies are successful in capturing the behaviour of the residual circulation is an open question. Here, we use an ensemble of Chemistry-Climate Model Initiative (CCMI) hindcast simulations from the Community Earth System Model version 1 Whole Atmosphere Community Climate Model (CESM1-WACCM) to compare observation-based proxies with direct measures of the residual circulation in a self-consistent manner. The three proxies studied are measures of the contrast in lower stratospheric temperatures between the tropics and poles, and ozone and water vapour concentrations in the tropical lower stratosphere. The temperature-based measure exhibits robust correlations with tropical lower stratospheric upwelling on interannual timescales, and a good year-round correlation (r = 0.73) between their monthly trends during the post-1998 ozone recovery era. We find that tropical mean ozone at 50 hPa has a maximum correlation with tropical upwelling at 70 hPa with a lag of 2 months. After accounting for this lag, ozone closely mirrors tropical upwelling variability on seasonal and interannual timescales as well as for long-term trends, especially for the ozone recovery period. On interannual timescales particularly, both the tropical mean ozone and temperature-based indices are strongly (anti-)correlated with tropical upwelling (r ~ 0.9), indicating these are suitable proxies for the residual circulation in CESM1-WACCM on this timescale. In terms of multi-year trends, tropical ozone shows the highest anti-correlation across months with tropical upwelling (r = -0.82) followed by the temperature-based index. The correlations of monthly trends are consistently smaller during the ozone depletion era (1979−1997) than during the era of ozone recovery (post 1998). The results indicate that both temperature and ozone based measures are suitable proxies for the residual circulation when tested in a self-consistent model framework.