Summary
The Southern Hemisphere mid-latitudes, particularly the area around south-eastern Australia and New Zealand, has been suggested as one of the regions where ozone recovery might first be detected, due to the combination of a strong impact of stratospheric halogen on ozone and relatively low dynamic variability (Weatherhead et al. 2000). Here, we present total column ozone data from four long-term Dobson sites located in Australia and New Zealand: Brisbane (153 E 27 S), Perth (116 E 32 S), Melbourne (145 E 38 S) and Lauder (170 E 45 S). Brisbane and Melbourne are operated by the Bureau of Meteorology while Perth and Lauder are operated by NOAA but hosted by the Bureau and NIWA respectively. All four time-series are traceable to the World Standard Dobson through a documented sequence of regular intercomparisons. Ozone values in August to October, when the maximum in the annual cycle occurs, are well-correlated from year to year across all four sites, with close to half the variance explained by the phase of the QBO. In February to April, when the minimum of the annual cycle occurs, ozone values are generally quite close together across the sites. After applying a three-year running mean to reduce the impact of the QBO and random dynamic influences, Melbourne and Lauder both display negative trends before the mid 1990s and positive trends afterwards, in both autumn (February-March-April) and spring (August-September-October). The relative magnitude of the decreasing and increasing ozone trends are consistent within uncertainties with the ratio of the increasing and decreasing trend in EESC before and after this time. The decrease in Perth is similar to Melbourne and Lauder at both times of year but has only been followed by an increase in spring and not autumn. We do not detect meaningful trends at Brisbane. The significance of these trends are calculated using the standard CUSUM method.
Related projects
