On the Determination of Age of Air Trends from Atmospheric Trace Species

Abstract

Abstract Trace chemical species have been used in numerical models to calculate the age of air (AOA), which is a measure of the strength of the mean meridional circulation. The trend in the AOA has also been computed and found to be negative in simulations where greenhouse gases increase with time, which is consistent with the acceleration of the mean meridional circulation calculated under these conditions. This modeling result has been tested recently using observations of SF6, a very long lived species whose atmospheric concentration has increased rapidly over the last half century, and of CO2, which is also very long lived and increasing with time. Surprisingly, the AOA estimated from these gases exhibits no significant trend over the period 1975–2005. Here the Whole Atmosphere Community Climate Model (WACCM) is used to derive estimates of the AOA from SF6 and CO2 over the period 1965–2006. The calculated AOA yields trends that are smaller than the trend derived from a synthetic, linearly growing tracer, even after accounting for the nonlinear growth rates of SF6 and CO2. A simplified global transport model and analytical arguments are used to show that this follows from the variable growth rate of these species. It is also shown that, when AOA is sampled sparsely as in the observations, the resulting trends have very large error bars and are statistically undistinguishable from zero. These results suggest that trends in the AOA are difficult to estimate unambiguously except for well-sampled tracers that increase linearly and uniformly. While such tracers can be defined in numerical models, there are no naturally occurring species that exhibit such idealized behavior.