Improving stratospheric transport trend analysis based on SF6 and CO2 measurements

In this study we reexamine nearly four decades of in situ balloon-based stratospheric observations of SF6 and CO2 with an idealized model and reanalysis products. We use new techniques to account for the spatial and temporal inhomogeneity of the sparse balloon profiles and to calculate stratospheric mean ages of air more consistently from the observations with the idealized model. By doing so we are able to more clearly show and account for the variability of mean age of air throughout the bulk of the depth of the stratosphere. From an idealized model guided by the observations, we identify variability in the mean age due to the seasonal cycle of stratospheric transport, the quasi-biennial oscillation in tropical zonal winds, major volcanic eruptions, and linear trends that vary significantly with altitude. We calculate a negative mean age trend in the lowest 5 km of the stratosphere that agrees within uncertainties with a trend calculated from a set of chemistry climate model mean ages in this layer. The mean age trends reverse sign in the middle and upper stratosphere and are in agreement with a previous positive trend estimate using the same observational data set, although we have substantially reduced the uncertainty on the trend. Our analysis shows that a long time series of in situ profile measurements of trace gases such as SF6 and CO2 can be a unique and useful indicator of stratospheric circulation variability on a range of time scales and an important contributor to help validate the stratospheric portion of global chemistry climate models. However, with only SF6 and CO2 measurements, the competing effects on mean age between mean circulation and mixing (tropical entrainment) are not uniquely separable.