AbstractThe main aims of this project are: (a) to find the basic climatology of ozone in the Australian region using ground-based Dobson observation records from all Australian stations for different periods, (b) to improve the homonegeity of the data-series records in (a) with the aid of coincident satellite TOMS ozone measurements, by computing and applying TOMS-Dobson total ozone bias to the Dobson ozone values, and (c) by applying statistical regression analysis, to calculate ozone trend using the original data-sets in (a) ("short method") as well using the bias-corrected data-sets in (b) ("long method"). and (d) to improve previous studies by use of data from more stations and having longer durations, and also use satellite data.
In the "short method", the monthly means of total ozone obtained from the Dobson spectrophotometer observations in all Australian stations (Darwin, Cairns, Brisbane, Perth, Melbourne, Hobart and Macquarie Island) were used to estimate their ozone trends by applying the regression analysis with intervention technique. The results of this analysis for a few different sub-periods between 1956 and 1992 (listed in Tables 4-7, 4-8 and 4-10 in chapter 4) show that the ozone content had decreased for all the stations, except for an inconclusive small increase at Macquarie Island. Statistically, the most reliable results from this single-data set estimation (or the "short method") is for Melbourne (both for 1956-1975 and 1976-1992) and Brisbane (especially for 1976-1992), all to within 99% confidence limit. Pro-rata combining of Melbourne and Brisbane results (as done in Tables 4-10(a)) for the two main periods (1956-1975 and 1976-1992) produces the following trends (in % per year) for 1 956-1992: Melbourne: (-0.34 ± 0.05), Brisbane: (430 ± 0.07). Th representative trend for Australian mid-latitude region is (-0.30 ± 0.07) % per year for 1956-1992 by the "short method". The next best trend values are for other stations which have shorter periods (1976-1992):--Perth (-0.15 ± 0.10), Hobart (-0.34 ± 0.13) and Macquarie Island: (+0.10 ± 0.13). The least reliable ones are for Cairns (4 13 ± 0.13) and Darwin (-0. 10 ± 0. 13), both having shortest record spans.
The "long method", which is devised to improve the "short method" and previous trend estimation. It uses both the Dobson and satellite-derived TOMS (Total Ozone Mapping Spectrometer) data by comparing the coincident (November 1978 to September 1988) daily readings of both data-sets to obtain monthly-bias values, for use as corrections to the TOMS (or Dobson) original monthly means. Again, it uses similar statistical regression method. The trend results of this "long method" (as listed in Table 4-9) are most reliable for Melbourne: (445 +0.08) and Brisbane: (-0.68 ±0.12), giving trend for Australian mid-latitude region of (457 ± 0.10) % per year for 1978-1988. This compares well with the results of other researchers. At other stations:— Perth: (-0.04 ±0. 16), Hobart: (-0.96 +0.22) and Macquarie Island: (+0. 10 ±0.13), less confidence can be placed in the results. Thus the trends are again all negative, although Macquarie Island value is not conclusive. The standard errors and the results of both the I- and P-statistics infer that the bias-corrected technique (the "long method") is more accurate and more suited to the models used. In the long method, a shorter span-length (i.e. 1978-1988) of slightly less than one sunspot cycle) was used. This had tended to reduce the accuracy of the results as compared to when a longer span period had been used. So, the full advantage of the "long method" has yet to be realised, when longer spans of satellite data are available in the future.
In addition to the trend estimate, a substantial part of this report was devoted to investigate the basic climatology and statistics of atmospheric ozone in the Australian region, including its horizontal, vertical and temporal variations. The results are mostly summarised in tabulated and graphical forms, such as shown in Appendices A to D.
Finally in this project, an attempt is made to explain the main features and anomalies found in the various results in terms of both natural and human factors, with emphasis placed on the effects of meteorological and geophysical influences on ozone in the atmosphere.
|Date of Award||1994|