The isentropic diffusion coefficient (Kyy) characterizes large-scale atmospheric mixing processes. Chemical constituents allow us to directly estimate Kyy from observational data sets. This study investigates general aspects of the mixing processes from a systematic survey of the common features and constituent dependency of Kyy . Analysis with chemical constituent data obtained from a global chemical transport model demonstrates that long-lived species whose chemical lifetimes ( ) are several years long have very common seasonal and latitudinal variations of Kyy, which is considered to represent actual atmospheric mixing. Kyy, estimated from chemical species with shorter than several weeks, becomes considerably greater than that from long-lived species and indicates the significance of nonlinear effects. Meridional transport analysis also investigates that shorter results in greater spatial constituent variations and larger contributions of the eddy compared to the mean motions in the constituent transport flux.
IntroductionLarge-scale isentropic eddy mixing plays an important role in determining constituent distribution (e.g., Plumb 2007). Isentropic mixing due to planetary wave breaking causes significant mixing in the stratospheric "surf zone" (McIntyre and Palmer 1984), while synopticscale disturbance, caused by baroclinic waves, disperses air along isentropes in the extratropical troposphere (e.g., Kida 1980). These mixings increase the constituent gradients over isentropic surfaces at the edge of the mixing region, but decrease them within the mixing region (e.g., Juckes and McIntyre 1987;Miyazaki and Iwasaki 2008).The eddy diffusion coefficient has been widely used to characterize mixing processes. The adiabatic component of the eddy diffusion coefficient, Kyy, reflects wave motions of the atmosphere since the eddy transport is almost parallel to the isentropic surface, particularly in the stratosphere. Kyy has also been used to parameterize the Rossby wave mixing in two-dimensional chemical transport models (CTMs, e.g., Garcia and Solomon 1983). Kyy is generally diagnosed using the potential vorticity (PV) flux (Newman et al. 1988) since the PV acts as a conservative passive tracer for adiabatic and frictionless flow. Magnitude of the diffusion coefficient also can be measured by ideal passive tracer calculations. However, it is difficult to directly estimate the PV flux from any observation, and calculations of the PV flux or an ideal passive tracer can suffer from numerical error. Instead of using the PV flux or the ideal passive tracer, some studies have used the distributions of specific long-lived species such as N2O and CH4 to investigate atmospheric transport and mixing processes (e.g., Lingenfelser and Grose 2002). Although long-lived species allow us to directly estimate Kyy from observational data sets, the sensitivity of Kyy to chemical species has not been clearly understood.In this study, we investigate the common features and constituent dependency of Kyy to facilitate understandin...