EVIDENCE FOR AN OZONE HOLE PERTURBATION AT 30-DEGREES SOUTH

A relatively strong stratospheric column ozone decrease was observed i n the south of Brazil (29.5 degrees S) in 1993, at the end of October. This ozone decrease was observed when the normal behavior of the ozon e column at low latitudes, in Brazil, reaches its yearly maximum, so t hat a decrease of ozone during this time period is unexpected. The loc al observations were made by two different measurement techniques. Two independent ground-based Brewer spectrophotometers documented strong column ozone decreases in the south of Brazil, in 1993. The vertical d istribution of ozone was observed with ozone-soundings, showing a unif orm ozone decrease at all heights in the stratosphere, and very low oz one in the lower stratosphere, which has been shown to be a characteri stic of Antarctic ozone in Spring. TOMS ozone data, representing a thi rd observational technique, averaged over small latitude-longitude ban ds, correlate very well with the local observations, which leaves no d oubt that the observed ozone decreases in the south of Brazil during O ctober 1993 were real. Also, the observed ozone decreases may be consi dered large, since for comparison, the local seasonal variation is at most of the order of 30 Dobson Units (DU), whereas one of the October decreases measured about 60 DU. There seems to be no physical or chemi cal mechanism at low latitudes, that could account for such a large an d fast perturbation in the stratosphere. On the other hand, inspection of the TOMS total ozone data maps, on the days of the above observati ons, show a distinct link between the Antarctic ozone hole latitudes, reaching out to the north in a curved path, and touching tropical lati tudes over a narrow belt. Trajectory analyses confirm that for days of low ozone in Santa. Maria, Brazil, the air masses at 20 and 25 km hei ght have an Antarctic origin.