Previous: Ext. Abst. Next: Experimental descent rates inferred from N2O, O3, and HNO3 Up: Ext. Abst.
1. Introduction:
Various theoretical and experimental studies have been made in recent years to determine seasonal rates of descent within the polar vortices during winter. Vortex descent has the important effect of altering profiles of mixing ratio versus altitude for various species, and therefore altering the speed and equilibrium points for chemical reactions between species, due to their (frequently) strong dependence on pressure and temperature. Moreover, descent of HNO3, H2O, and H2SO4 into the very low temperature region that forms by mid winter in the 15-25 km altitude range leads to large scale gravitational removal of these species from the lower stratosphere, following their condensation into Polar Stratospheric Cloud (PSC) particles. This has an even more profound effect on stratospheric vortex chemistry.
An early measure of descent in the lower stratosphere was made by Parrish et al., [1988] who observed a dramatic flattening of the mixing ratio profile of N2O over McMurdo Station, Antarctica, by the end of the polar winter, when compared with earlier summer measurements. ER-2 aircraft, ATMOS, and several UARS instruments have added a number of other data sets from which to derive vortex descent rates, though these do not extend beyond ~80º S, and frequently do not reach that far. The edge region of the vortex is generally thought to descent more rapidly than the core. Quasi-inert tracers such as N2O and CH4 are usual chosen for experimental descent measurement.
Backmeister, et al. [1995], Strahan, et al. [1996], and Allen, et al. [2000], among others, have variously noted that different tracers yield different apparent descent rates, and have explained this as a result of mixing across the vortex boundary in the presence of different gradients for different species (see below). To date, these investigations have been limited to analysis of data between ~60-80º S. In what follows, we show that similar results hold true to the center of the vortex core, and not merely its outer regions, and we provide new views of vortex descent.
Diabatic descent has been investigated theoretically by several groups. For instance, Rosenfield et al. [1994] have estimated diabatic descent rates from various altitudes within the polar vortices by calculating diabatic cooling from UKMO assimilated temperature fields for 1992 and 1993, considering the vortices as isolated systems without cross-boundary transport. More recently, Rosenfield et al. [2000] have recalculated polar diabatic descent rates using back trajectory calculations on air parcels over periods up to seven months in an effort to determine average origins and thermal history of air reaching the vortex core by mid-winter. They conclude that descent predicted by this technique is always slower than descent rates calculated on the basis of an ‘isolated’ vortex, because horizontal motion and variation in heating (cooling) rates with latitude and longitude reduce the time-integrated rates of cooling. The use of multi-month back trajectories, even though used only to establish average origins and motions, may bias results however, and it is probably best to take the original [1994] and more recent [2000] calculations of Rosenfield et al. to represent approximate extremes within which real vortex behavior may lie.
Working with ground-based mm-wave measurements of O3 taken for most of a year over the South Pole in 1993, we investigated the use of O3 as a tracer of stratospheric vortex core descent in a paper by Cheng et al., [1996]. Although O3 is essentially inert chemically in the middle-atmospheric winter darkness, it was found that descent traced by constant mixing ratio contours of O3 versus time showed physically incorrect slower descent in the upper stratosphere than in the middle stratosphere. We revisit that observation here, extending the observations to the austral winter of 1999 and to apparent descent rates of HNO3 as well as O3 and N2O versus altitude, inferred from profiles retrieved by deconvolution from pressure-broadened mm-wave emission lines.