Origin of extreme ozone minima at middle to high northern latitudes

Abstract

Extreme ozone minima represent localized and temporally brief (several days) reductions in column ozone amounts below some chosen absolute level. Although such minima at middle to high northern latitudes are known to be primarily dynamical in origin, a remaining issue is whether heterogeneous chemical loss processes may also contribute significantly to their formation. A case in point is the record low 165 Dobson units (DU) minimum occurring on November 30, 1999, when temperatures near 30 hPa at the location of the minimum were lower than the threshold for the formation of type I polar stratospheric clouds (PSC). An examination of Polar Ozone and Aerosol Measurement III data for times surrounding the event shows that PSCs were indeed present in the vicinity where the ozone minimum was observed. However, archived data show that a similar extreme minimum of 167 DU with characteristics comparable to those of the November 30, 1999, minimum occurred on October 30, 1985, when no PSCs were present. An ensemble of 71 extreme ozone minima with amplitudes under 215 DU exhibit a nearly linear relationship between ozone minimum deviations from the zonal mean and corresponding 30-hPa temperature deviations. Such a relationship is predicted by analytic transport models which assume that vertical motions (i.e., upwelling) are responsible for the ozone minima. Temperature deviations near 30-hPa were unusually large for both the November 30, 1999, and the October 30, 1985, events, implying unusually rapid upward transport for these events. All 71 minima occur in regions where deviations from the zonal mean of 330 K potential vorticity are negative, implying an additional contribution to their formation by quasi-horizontal transport. The timescale for column ozone reductions during extreme ozone minima events is also determined and found to be at least 20 times more rapid than expected from known chemical loss processes. The data are therefore most consistent with a purely dynamical origin for extreme ozone minima in general and the November 30, 1999, event in particular. As was shown by earlier work, the basic dynamical process involves a combination of isentropic transport of ozone-poor air from the tropical upper troposphere and rapid upwelling over upper tropospheric anticyclonic disturbances resulting from poleward Rossby wave breaking events.