• Introduction
• Methodology
• Results of the 990922 case
• Conclusions
• Further work

FIGURES

• Table 1
• Figure 1
• Figure 2

1. Introduction

During the 30 days of APE-GAIA campaign 13 ozone soundings, 42 pTU-soundings and 25 total ozone measurements were made at Marambio (64¡23' S, 56¡72' W). Ozone soundings were timed to match the overpass of Geophysica, a high altitude atmospheric measurement laboratory, so that as much as possible the same air mass was measured.

Strong filamentation can be seen in the ozone profiles from Marambio. In this study the origin of these laminae from four soundings were investigated by using trajectory analysis in conjunction with an analysis of vortex situation and trace gas measurements from Geophysica.

2. Methodology

Ten of the thirteen ozone soundings of Marambio during APE-GAIA show some layering. Trajectories were calculated isentropically by means of a global trajectory model by G. Redaelli and R. Dragani at the University of l«Aquila. In the tables potential vorticities (PV) of the filaments at the sounding and after seven days backward advection are compared with the each other and to the PV values of the core of the vortex edge of the corresponding isentrophic surface (theta). PV unit used here is PVU (=10**-6*K*m**2*kg**-1s**-1).

PV of the vortex edge on each surface was determined to be the mean PV value at the location of maximum gradient along 14 longitudes around the globe between the latitudes of -40.0¡ and -82.5¡ . The mean value was calculated separately for each seven days period.

Winds from ECMWF were retrieved and PV was calculated and drawn together with position of the air parcel on the specific thetas in a 2.5¡ x 2.5¡ grid. Thus the trajectories were compared also to the position and the overall development of the vortex. The trajectories of the ozone filaments and Geophysica were compared in order to find a best possible match and a possibility to use in situ trace gas measurements in the filament analysis. One match, that of 23.09.1999 sounding on 413 K, was found.

In the study the basic assumption is that the filaments of low ozone originate from the chemically perturbed region inside the vortex where as the high ozone filaments originate outside the vortex or from outer side of the edge region. This difference in origin should in turn be seen in differences of PV and tracers. If not, the cause of the filamentation is else than horizontal mixing of air through the vortex edge. In the Southern Hemisphere the PV values inside the vortex are typically very low compared to the PV values in surf zone. Between the two largely isolated air masses there is a transition zone where PV changes rapidly. Thus, low values of PV indicate origin inside the vortex.

3. Results of the 990922 case

In the sounding of 990922 there is two ozone minima on 520 K and 405 K thetas and a deep layer of air rich of ozone in between (PLATE 1.).

PV from the trajectory analysis is presented in Table 990922. The initial value of PV (PVs) on 405 K theta is -21 PVU which is slightly greater than the PV value after 7 days backward advection (PVt) (ö22 PVU). Both are anyway smaller than the PV value of the edge (PVe)(ö18 PVU). On 520 K theta the difference is slightly greater and the sign of the change is the same implying of intrusion from inside the vortex into the edge region.

In the high ozone thetas of 469 K and 425 K the difference of PVt to PVs is opposite to the low ozone thetas implying transport from opposite direction.

In Figure 1. the vortex situation during sounding and the vortex situation after seven days backward advection is presented. From there it seems like the air parcel has crossed the PV isolines and moved from more inside vortex towards edge region. The same applies to 520 K (not shown). On 469 K theta the air parcel seem to stay between the same PV isolines (not shown) where as on 425 K theta it seem to have moved from edge towards inside of vortex, see Figure 2.

On 585 K theta PVt is much greater than the PVe and the difference of PVt to PVs is positive implying intrusion from outside in to the vortex edge region. On 585 K theta PVt is much greater than the PVe and the difference of PVt to PVs is positive implying intrusion from outside in to the vortex edge region. When the 585 K trajectory is plotted on the corresponding vortex situation (not shown) no such phenomena is seen.

4. Conclusions

In this study it filaments of high and low ozone from four ozone soundings that were performed when inner edge of the vortex edge region was above Marambio was studied. It was assumed that the low ozone filaments originated from the core rather than the edge region of the vortex and vice versa for the high ozone filaments and that it could be seen in the behaviour of the trajectories and PV of the air parcels from these filaments. In some cases, as especially for 990922, this assumption seemed to be correct. On the other hand there were discrepancies that might have caused either by the methodology or by the possibility that unlike assumed the origin of the filaments were not as straightforward. Methodological cause might be that there is no real correlation between PV changes and mixing of air on isentropic surfaces.

5. Further work

- theoretical considerations of the method

- inclusion of the rest of the soundings to the analysis

- statistical analysis of filaments of Marambio ozone soundings 1988-2000

- modelling of the filaments