The Antarctic ozone hole during 2014

As reported in Dameris and Godin-Beekmann et al. (2014), the Antarctic ozone hole has continued to appear each spring since its first detectable appearance in 1979. Due to the still highly elevated levels of ozone-depleting substances (ODS) and their slow rate of decrease in the atmosphere, the severity of the ozone hole from year to year is currently determined primarily by interannual variations in temperature and dynamics. Nonetheless, Dameris and Godin-Beekmann et al. (2014) were able to estimate an underlying increase in Antarctic ozone from 2000 to 2014 of between 10 and 25 DU, while noting that the variability had been somewhat greater in the last decade than had been observed in the 1990s.

It has also been increasingly realised that Antarctic ozone depletion is very likely to have been the dominant driver of the observed changes over recent decades in the southern hemisphere tropospheric circulation in summer, including an increase in the strength of the Southern Annular Mode (SAM), an expansion of the Hadley Cell and a poleward shift of the mid-latitude maximum of precipitation (Arblaster and Gillett et al. 2014).

In this paper, we provide a description of the overall level of Antarctic ozone depletion in 2014 and the relationship with prevailing meteorological conditions using a range of Australian data and analyses including measurements and analyses by the Commonwealth Scientific and Industrial Research Organisation (CSIRO) Oceans and Atmosphere – Climate Science Centre, ozone measurements obtained by the Australian Antarctic Division (AAD) and the Bureau of Meteorology (BoM), and Antarctic ultraviolet measurements from the Australian Radiation Protection and Nuclear Safety Agency (ARPANSA) biometer network. Other data from satellite missions and ground-based instruments are also presented. This work complements analyses of previous Antarctic ozone holes reported by Tully et al. (2008, 2011) and Klekociuk et al. (2011, 2014a, 2014b, 2015), and other analyses of Antarctic atmospheric conditions and ozone depletion during 2014 provided by the World Meteorological Organisation Antarctic Ozone Bulletins (http://www.wmo.int/pages/prog/arep/gaw/ozone/index.html, accessed 30 April 2020), upper air summaries of the National Climate Data Center (NCDC; http://www.ncdc.noaa.gov/sotc/upper-air, accessed 30 April 2020) and by Blunden and Arndt (2015; http://www.ncdc.noaa.gov/bams-state-of-the-climate, accessed 30 April 2020).

As in previous reports in this series, we use total column ozone measurements from satellite instruments to obtain metrics of the Antarctic ozone hole (see Klekociuk et al. (2015) for details). Here we use data processed with the version 8.5 Total Ozone Mapping Spectrometer (TOMS) algorithm from the TOMS series of satellite instruments, the Ozone Monitoring Instrument (OMI) on the Aura satellite and the Ozone Mapping Profiler Suite (OMPS) on the Suomi National Polar-orbiting Partnership satellite.

Table 1 contains the ranking for the 35 ozone holes adequately observed by satellite instruments since 1979 using eight metrics that provide different measures of the extent of ozone depletion in each year (see the notes accompanying the table for the definition of each metric). The first seven metrics in Table 1 measure various aspects of the maximum area and depth of the ozone hole; the 2014 ozone hole was ranked between 16th and 21st in terms of severity across these metrics, showing quantitative similarity with ozone holes of the late 1980s, early 1990s and some recent years, particularly 2010 and 2013.

Table 1.  Ranked Antarctic ozone hole metrics obtained from TOMS–OMI satellite data
TOMS data are used from 1979 to 2004, OMI data are used from 2005 onwards. Data for 2014 are highlighted in bold. NOTE: in previous papers in this series (Tully et al. 2008, 2011; Klekociuk et al. 2011, 2014a, 2014b, 2015), the Antarctic ozone hole metrics for the year 2005 were an average of both TOMS and OMI data, and the 2005 rankings in those papers will differ to those quoted below. Rank 1 is the lowest ozone minimum, greatest area, greatest ozone loss, etc.; Rank 2 is the second lowest ozone minimum, etc. There was a gap in TOMS coverage during the growth of the 1994 ozone hole; metrics for some parameters for that year are, therefore, undetermined and are left blank. There were no relevant TOMS measurements in 1995. Metric definitions: (1) Maximum 15-day averaged area: the largest value (in each year) of the daily ozone hole area averaged using a 15-day sliding time interval. (2) Daily maximum area: the maximum daily value of the ozone hole area. (3) Minimum 15-day averaged total column ozone: the minimum of the 15-day averaged column ozone amount observed south of 35°S. (4) Daily minimum total column ozone: the minimum of the daily column ozone amount observed south of 35°S. This metric effectively measures the ‘depth’ of the ozone hole. (5) Daily minimum average total column ozone: the minimum of the daily column ozone amount averaged within the ozone hole. This metric effectively measures the ‘average depth’ of the ozone hole. (6) Maximum daily ozone mass deficit: the maximum value of the daily total ozone mass deficit within the ozone hole. This metric effectively measures the combined area and depth of the ozone hole. (7) Integrated ozone mass deficit: the integrated (total) daily ozone mass deficit for the entire ozone hole season. This metric effectively measures the overall severity of ozone depletion. (8) Breakdown date: the final date at which the daily maximum area (metric 2) falls below 0.5 million km². Note that the metrics use 220 DU as the threshold in total column ozone to define the location and occurrence of the ozone hole. OMI, Ozone Monitoring Instrument; TOMS, Total Ozone Mapping Spectrometer

Figure 1 shows time-series of the ozone hole area, minimum polar total column ozone and total ozone deficit within the ozone hole over the latter half of each year from 2009 to 2014. It can be seen that in 2014, the evolution of the ozone hole was generally similar to the long-term mean (white line, which is the average over 1979–2013). Specific notable features in the time-series for Fig. 1a, c are the brief reductions in area and mass deficit during the last few days of August and again around September 25 and the middle of October. These features coincided with disturbances of the polar vortex associated with episodes of poleward heat transport as discussed in Appendix 1 (in association with Figs. A1.2—A1.4). The vortex experienced a relatively undisturbed period at the beginning of October which led to a marked week-long increase in ozone hole area and deficit. The timing of the dissipation of the hole at the beginning of December was similar to the 2009 ozone hole, a few weeks later than that for 2013 which was unusually early and a few weeks earlier than the ozone holes in 2010 and 2011 which were very persistent.

Fig. 1.  Estimated daily (a) ozone hole area, (b) ozone hole depth and (c) ozone mass deficit based on Ozone Monitoring Instrument (OMI) satellite data for 2009–14 and Ozone Mapping Profiler Suite (OMPS) satellite data for 2014. The shaded region and white line show the range and mean respectively over 1979–2013.

A further notable feature are the two relatively low total column values apparent in the OMI instrument data of Fig. 1b towards the end of August; these are on the 23rd and 26th and have values of 127 and 130 DU respectively. No measurements were available from the OMPS instrument on the 23rd, but on the 26th, the minimum value observed was 165 DU. SAOZ (Systeme d’Analyse par Observation Zenithale) total column ozone measurements at Rothera (67.6°S 68.1°W; close to where the OMI minima was recorded) shows 131 DU on the 23rd, and 172 on the 26th, which broadly confirm the OMI figures. The difference between the OMI and OMPS measurements on the 26th highlights that caution is needed interpreting this metric during the formation of the ozone hole, as parts of the polar cap are still in darkness and cannot be measured by these instruments, and effects due to scattering from Polar Stratospheric Clouds at low sun angles can influence the measurements.

A prominent feature of daily maps of total column ozone obtained by satellite instruments during September and October was the ridge of high ozone concentration to the south of Australia (see Krummel et al. 2015). The ridge develops each spring due to transport effects associated with the circulation pattern set up by a wave-1 quasi-stationary planetary wave in the lower stratosphere. Episodes of disturbance in the planetary wave field during these months resulted in the ozone hole being displaced off the pole towards the Atlantic Ocean, and at times up to a third to a half of the Antarctic continent was outside of the ozone hole. During these periods the polar vortex (and hence the ozone hole) became quite elongated, resulting in the tip of South America being on the edge of or within the ozone hole on 14–17 September, 6–7 10 October and 12–14 November.

Between the above-mentioned periods of high ozone, the polar vortex and ozone hole were relatively symmetrical, with the Australian Antarctic stations of Mawson, Davis and Casey being on the edge of or completely inside the ozone hole on 6–7 September, 20–21 September, 26–30 September and 1 October (the date of the peak ozone hole area for 2014). There were also several periods when the stations were outside of the ozone hole: 15, 22–23 September and 16–18 October. The Australian sub-Antarctic station at Macquarie Island spent most of the 2014 ozone hole season under the ozone ridge (see later).

Yearly values from Table 1 for maximum ozone hole area (15-day average) are also presented as a time-series in Fig. 2a, together with the estimated level of Antarctic Equivalent Effective Stratospheric Chlorine (EESC; orange line) (Fraser et al. 2014; Klekociuk et al. 2015), which is a measure of the potential for chemical-ozone depletion in the lower stratosphere. The 2014 value was characteristic of those observed in the early 1990s, and lower than the majority of the intervening years 1993–2013, suggestive of the beginning of ozone hole recovery. However, considerable meteorological variability is also evident, superimposed on the longer term trend, most apparent in the reduced ozone hole area in the dynamically disturbed years of 1986, 1988, 2002, 2004, 2010 and 2012 (Klekociuk et al. 2014b).

Fig. 2.  (a) Maximum ozone hole area (area within the 220 DU contour) using a 15-day moving average during the ozone hole season, based on Total Ozone Mapping Spectrometer (TOMS) data (green) and Ozone Monitoring Instrument (OMI) data (purple) for all available years. The orange line is obtained from a linear regression to Antarctic Equivalent Effective Stratospheric Chlorine (EESC-A) as described in the text. The error bars represent the range of the ozone hole size in the 15-day average window. (b) October monthly mean total column ozone values for Halley station for 1957–2014 (green points and line; BAS 2019) and regression to EESC (orange line) from Fraser et al. (2014) using a mean age of air of 5 years. Orange dashed line is the 95% confidence interval of the regression. (c) Residual formed by subtracting the EESC regression from the October monthly data shown in (b), expressed as a percentage of the monthly average total column amount.

As a general guide, the averages of the metrics in Table 1 for the peak ozone hole period of 1996–2001 can be compared to the average of the most recent 6-year period (2009–14) ozone hole metrics, for simple indications of ozone hole recovery (keeping in mind these metrics are affected by the meteorological conditions). The 1996–2001 mean of the 15-day ozone hole area was (25.6 ± 2.0) × 10⁶ km², whereas the 2009–14 mean was (22.5 ± 2.0) × 10⁶ km², which is indicative of the commencement of possible ozone recovery, but not statistically significant as the 1σ uncertainties overlap. The 1996–2001 mean of the 15-day ozone hole minima was 100 ± 5 DU, whereas the 2009–14 mean was 119 ± 13 DU. There is a strong suggestion that ozone is recovering in this metric with the uncertainties no longer overlapping (at 1σ level). The 1996–2001 mean of the integrated ozone mass deficit was 2180 ± 230 Mt, whereas the 2009–14 mean was 1380 ± 510 Mt, which also suggests the commencement of ozone recovery (uncertainties no longer overlapping at 1σ).

A longer term data record of Antarctic ozone depletion is available from ground-based Dobson spectrophotometer measurements at the British Antarctic Survey’s Halley station (75.6°S and 26.2°W) in Antarctica. Fig. 2b shows the mean October total column ozone obtained at Halley from 1957 to 2014, again with EESC levels.

The Halley total column ozone value for 2014 shown in Fig. 2b of 148 DU was less than that measured in 2012 (197 DU) and 2013 (177 DU), although the value was within 1 standard deviation (σ) of the long-term total column value expected on the basis of regression to EESC (Fig. 2c). Ignoring the dynamically disturbed years of 2002 and 2004 (Klekociuk et al. 2015), the mean October total column ozone value at Halley over 2009–14 (166 ± 18 DU) is higher than that over 1996–2001 (141 ± 4 DU); this increase is statistically significant at the 98% confidence limit based on the Student’s t-test for differences of mean with unequal variances. If we remove 2002 and 2004, the remaining years (1996–2014) show significant ozone growth (recovery) of 1.5 ± 1.2 (2σ) DU/year. For the period 1993–2014, the ozone growth is 1.8 ± 0.8 (2σ) DU/year, although the early 1990s data may be low due to the impact of the Mt Pinatubo and Mt Hudson eruptions, rather than reflecting the minimum of EESC which occurred around the year 2000.

In Fig. 2b, the dynamically disturbed years of 1986, 1988, 2002, 2004, 2010 and 2012 (Klekociuk et al. 2014b) show obvious positive values of the anomaly. The period 1991–95 shows negative anomalies, and these years were potentially affected by enhanced ozone depletion following the volcanic eruptions of Mt Pinatubo (June 1991; Philippines) and Mt Hudson (August 1991; Chile) (Deshler et al. 1992; Aquila et al. 2013). However, the years 1983–85, 1987 and 1989 show negative anomalies that are close to or outside the 95% confidence interval of the fitted regression. For these years, we do not regard that instrumental factors nor underestimation of EESC could alone account for the depth of these anomalies; however, these and other factors deserve more detailed consideration. We note that volcanic effects from the El Chichón eruption in early 1982 are not regarded as having significant negative impacts on Antarctic ozone (Angell et al. 1985; Angell 1997). Further, October vortex mean temperatures were relatively cold during the period from 1987 to 1998.

It is also worth noting the qualitative agreement between Fig. 2a and b, indicating broad agreement between different measures of Antarctic ozone hole depletion.

To complement satellite measurements of total ozone during 2014 shown in Fig. 1, the time-height development of the anomaly in ozone density over the southern polar cap (poleward of 60°S) from September 2014 to February 2015 is summarised in Fig. 3. This figure uses vertical profiles of stratospheric ozone number density obtained from the Optical Spectrograph and Infra-Red Imager System (OSIRIS) instrument on the Odin satellite (see Klekociuk et al. (2015) for details), and the anomaly (expressed as a percentage deviation) has been evaluated using the climatological mean over corresponding days between 2003 and 2014.

Fig. 3.  Time-height cross section of Optical Spectrograph and Infra-Red Imager System (OSIRIS) daily zonal mean stratospheric ozone number density anomaly averaged over latitudes 60–83°S for September to February 2014–15. The anomaly at each height is evaluated relative to the corresponding days for 2003–14 climatological period. White gaps indicate time periods when no measurements were performed. Hatching is shown where the anomaly is above (below) the 90th (10th) percentiles over the climatological period.

Evident in Fig. 3 are regions of anomalously low ozone (hatched blue regions) above ~17-km altitude in September which gave way to regions of anomalously enhanced ozone below ~21-km altitude in October and November. The timing of the increase in ozone generally coincides with the reduction in the severity of ozone hole metrics shown in Fig. 1, particularly the sharp increase in area and deficit in early October and decrease in mid-October. Note also in Fig. 3 that reduced ozone generally persisted at the lowermost altitudes of the figure from December to February of 2015.

Annual values of the vortex-average rate-of-change of ozone mixing ratio as a function of temperature, averaged over days 200–260 (19 July–17 September in nonleap years), for isentropic levels of 450 K (~18 km height) and 850 K (~31 km height), are shown in Fig. 4a and 4b respectively. These figures essentially summarise characteristics during the period when the ozone hole is generally growing. The values are obtained from Aura Microwave Limb Sounder (MLS) version 3.3 data as described in Appendix 1.

Fig. 4.  Vortex-average ozone rate of change versus temperature, averaged between days 200 and 260, on isentropic surfaces of (a) 450 and (b) 850 K obtained from Microwave Limb Sounder v3.3 swath measurements. The vertical and horizontal bars span ± one standard error in the mean of the deseasonalised daily measurements. The base period used to deseasonalise the daily values is 2004–13. The relevant year is indicated to the upper left of each value.

On the 450 K isentrope (Fig. 4a), the ozone rate-of-change is positively correlated with temperature (R = 0.79, significant at the 95% confidence limit), with an enhanced ozone loss rate occurring at lower temperatures. Recent years of relatively late ozone hole development, 2010 (Klekociuk et al. 2011) and 2012 (Klekociuk et al. 2015), lie to the upper right in the figure. The growth of the ozone hole is potentially influenced by the amount of chemical processing that has taken place within the vortex over the winter (which is enhanced at lower temperatures by greater polar stratospheric cloud volume), and the amount of the vortex that is illuminated by sunlight after the end of the polar night (which depends on the size and symmetry of the vortex). As both 2010 and 2012 were years in which the polar vortex was relatively disturbed in spring, Fig. 4a suggests that the influence of the warmer temperatures on reducing heterogeneous reactions is a more important factor during the ozone hole growth phase than distortion of the vortex, which would be expected to increase the ozone loss rate by enhancing photolytic destruction as the equatorward fringes of the vortex come under greater solar illumination.

On the 850 K isentrope (Fig. 4b), the correlation between the ozone rate-of-change and temperature is negative (R = −0.78, also significant at the 95% confidence limit), with 2010 and 2012 appearing as anomalous years. At this level, ozone loss is primarily by gas-phase processes which are more efficient at higher temperatures. Overall, the behaviour of 2014 in Fig. 4a, b appears typical in comparison to years in which the polar vortex during late winter and early spring was relatively undisturbed by dynamical activity. As has been discussed with regard to Figs. 1 and 3, in 2014 episodes of poleward transport of heat occurred in late September and the middle of October (see also Fig. A1.2), somewhat later than the day 200–260 period analysed here.

In 2014, the program of ozonesonde measurements at Australia’s Davis station in Antarctica (68.6°S, 78.0°E) continued. This involved launches at approximately weekly intervals from mid-May to late December, and one flight per month for February to April. Figure 5 shows the 12–20 km partial column ozone amount for 2014 in comparison with earlier measurements. Overall, the evolution of the partial column values did not diverge greatly from the average pattern of previous years but did show generally less variability than earlier years during the spring (particularly when compared with 2012 and 2013), and most notably, no values of this partial column less than 20 DU were observed in contrast to many previous years. The steady decline of partial column ozone at Davis between days 200 and 260, at a range in the middle of the climatological spread, is reflective of the rate of decrease in the overall vortex shown in Fig. 4a.

Fig. 5.  Time-series of partial column ozone for the height interval 12–20 km obtained from ozonesonde measurements at Davis, Antarctica (68.6°S and 78.0°E). Shown are data for all years of measurement, with data for 2014 highlighted in black. The grey line is a climatological mean from Fortuin and Kelder (1998) interpolated to the location of Davis. Note that at Davis, this height range is almost exclusively above the lapse rate tropopause and generally below the burst height of the ozonesonde balloons in winter.

A feature of Fig. 5 is the generally higher interannual variability that is apparent in the winter (June–August, day-of-year 152–243) compared with summer (December–February, 335–59) and autumn (March–May, 60–151). In Fig. 6a, we show the average winter (day-of-year 152–243 in nonleap years) against the standardised SAM index for the corresponding period (Marshall 2003). The period covered is at the time of year when the partial column values do not exhibit a strong seasonal trend, being generally while the vortex is in darkness and thus before ozone undergoes strong depletion. A weighted linear regression of the yearly partial column values (y) against SAM yields the equation

Fig. 6.  (a) Average annual June–August 12–20 km partial column ozone from Davis ozonesonde measurements versus the standardised surface Southern Annular Mode (SAM) index for June–August from Marshall (2003). The value for 2014 is highlighted in red. Vertical bars span ± one standard error in the mean. The line shows the best fit linear weighted linear regression (weighted by the reciprocal of the standard error of each value). (b) Residual partial column ozone, given by the difference between the partial column ozone measurements and the linear regression in (a) versus CSIRO 5-year-lagged Equivalent Effective Stratospheric Chlorine (EESC). The vertical bars span ± one standard error in the residuals. Values progress in year from right (2003) to left (2014, highlighted in red) and the line shows the best fit weighted linear regression.

where values in parentheses are twice the standard error. The fraction of the variance in the partial column values explained by this relationship is R² = 0.84.

As described in Appenzeller et al. (2000), a positive correlation is seen between total ozone and the North Atlantic Oscillation (NAO) in winter months in the Northern Hemisphere. These authors showed that the observed behaviour is due to the raising (lowering) of the polar tropopause by the negative (positive) phase of the NAO which, in turn, effectively decreases (increases) the thickness of the stratosphere and hence decreases (increases) the overburden of ozone. A similar mechanism was proposed by Fogt et al. (2009) to explain the winter correlation between model simulations of total column ozone over the South Pole and SAM. We, therefore, examined the relationship of the Davis partial column ozone values with the mean June–August pressure at the lapse-rate tropopause obtained from the ozonesonde measurements between 2003 and 2014. A positive correlation between the tropopause pressure and SAM was found with R = 0.43 (i.e. a higher tropopause when SAM is negative, supporting the mechanism described by Appenzeller et al. (2000) which is significant at the 91% confidence limit. The relatively low significance of the correlation may reflect that the tropopause pressure over Davis is influenced by synoptic-scale meteorological variability that, while affecting height of the tropopause, is unconnected with SAM variations. The correlation between surface pressure and SAM was found to be R = −0.78, whereas correlation between partial column ozone and surface pressure gave R= −0.6; both of these correlations are significant at the 95% confidence limit and lend further support to the hypothesis that large-scale pressure changes provide a dominant influence on ozone in the winter lower stratosphere above Davis.

We note that Miyagawa et al. (2014), from the analysis of Umkehr measurements at Syowa (69.0°S, 39.6°E) from 1977 to 2011 that were necessarily restricted to the months from August to March, find only a weak positive (negative) correlation between ozone in the lower stratosphere and SAM for August during years of high (low) solar activity (see their Fig. 6a). This difference with the response seen at Davis, which was observed over almost a full solar cycle, may reflect zonal asymmetries in the spatial influence of the SAM on the upper troposphere near the Antarctic coast (e.g. fig. 4 of Thompson et al. 2005). Although a large fraction of the interannual variability in the partial column ozone is removed by subtracting the linear regression with SAM, the resulting residuals do not show a significant linear relationship with other relevant ozone-influencing factors, such as EESC (Fig. 6b), June–August standardised quasi-biennial oscillation (QBO) index (e.g. the 30 hPa QBO index discussed in Appendix 1; not shown) or polar cap (65–90°S) temperature anomalies in the lower stratosphere (e.g. anomalies for the levels discussed in Fig. A1.1 of Appendix 1 evaluated for the 1979–2002 base period, not shown).

The BoM carries out long-term high quality measurements of total column ozone at sub-Antarctic Macquarie Island (54.5° S, 158.9° E) using the Dobson spectrophotometer, continuing a program dating back to 1957. Observations for 2014 are shown in Fig. 7 (red) compared to the 1987–2013 range. In contrast to Dobson observations from Halley shown in Fig. 2, Macquarie Island generally lies outside the boundary of the polar vortex in spring and sees significant fluctuations in daily total ozone due to variations in the shape and position of the vortex.

Fig. 7.  Total column ozone observations made by Dobson spectrophotometer at Macquarie Island in 2014 (red), compared to the 1987–2013 range (The blue line shows the smoothed daily median, the dark blue band the 30th—70th percentile range and the light blue band the 10th—90th percentile range).

As referred to earlier, during most of October and November, total ozone was at or above the climatological mean range at Macquarie Island; however, a number of fluctuations were also observed due to the periods of disturbance to the polar vortex noted earlier.

A number of days with very high ozone levels were recorded through late September and early October. The daily average for 25 September 2014 (day-of-year 268) was 458 DU, followed by 450 DU on 26 September. The daily average for 8 October (day-of-year 281) was 443.5 DU. On these occasions, the ozone hole had become elongated due to dynamical disturbance along an axis located away from Macquarie Island.

In contrast, in late August, Macquarie Island was subject to the influence of the polar vortex, and low daily ozone averages were measured of 309.9 DU on 22 August (day-of-year 234) and 309.7 DU on 28 August (day-of-year 240).

An episode of low ozone was also observed in early December with a daily average 276 DU on 7 December (day-of-year 341); however, this event appears attributable to transport from mid-latitudes rather than polar influence.

Measurements of biologically effective solar ultraviolet (UV) radiation continued in 2014 at Casey (66.3°S, 110.5°E), Mawson (67.6°S, 62.9°E) and Davis (68.6°S, 78.0°E) in Antarctica. Details on the instrumentation and methods used are provided by Tully et al. (2008) and Klekociuk et al. (2015).

In Fig. 8, measurements of the UV Index at Casey, Mawson and Davis from July to December 2014 are compared with daily total column ozone measurements near local noon from the OMI instrument. There was a period of relatively high UV Index at the three stations in the first half of December as the ozone hole was dissipating. The general level of the UV Index at Mawson can be seen to be higher than at the other two stations in Fig. 8, despite Mawson being intermediate in latitude between Davis, the southernmost of the stations and Casey. This is further apparent in Fig. 9 which shows the September–December mean UV Index for the three stations from 2007 to 2014. The mean levels in these months during 2014 were higher at Davis and Mawson than in 2012 and 2013 when there was relatively less ozone depletion over Antarctica (Fig. 1c). The higher UV exposure at Mawson is potentially the result of the asymmetry of the polar vortex, which tends have its inner edge further from the Mawson coast than for the other stations. Additionally, the surface ice albedo at Mawson is higher which is due primarily to the greater extent and persistence of land-fast sea ice than at the other stations, and this will tend to enhance UV levels through backscattering effects.

Fig. 8.  Total column ozone in Dobson units ( left axis) from Ozone Monitoring Instrument (OMI) measurements and daily ground-based UV Index measurements ( right axis) during 2013 for (a) Casey (66.3°S and 110.5°E) and (b) Mawson (67.6°S and 62.9°E) and (c) Davis (68.6°S and 78.0°E). The 220 DU ozone hole threshold is also marked ( left axis).

Fig. 9.  Ultraviolet (UV) Index averaged over September to December for 2007–14 from daily measurements at Casey, Davis and Mawson.

We have examined meteorological conditions and ozone concentrations in the Antarctic atmosphere during 2014 using a variety of data sources, including meteorological assimilations, satellite remote sensing measurements and ground-based instruments and ozonesondes.

The polar vortex was only moderately disturbed and was of a size and duration very close to the 1992–2013 mean for most of the season. Stratospheric 65–90°S temperatures were slightly cooler than the long-term mean in July, August and September, and until late September, ozone depletion was significant and the ozone hole relatively large. A dynamical disturbance in late September and another in mid-October resulted in temperatures being slightly above the mean in October which prevented more severe ozone loss from taking place. As a result, all metrics of the ozone hole reported in this work were close to their median values, ranked between 16th and 21st of the 35 years assessed. In many respects, the 2014 ozone hole resembled those of the early 1990s, with less severe depletion measured than during the 1996–2001 peak period. Given meteorological conditions were close to average, the modest ranking of the metrics is suggestive that the ozone hole is now responding to declining levels of ODS in the stratosphere (as indicated by the decline in EESC).

A comparison of the historical ozone hole metrics from the peak period of 1996–2001 to the 2009–14 period suggests that ozone hole recovery may have commenced based on the 15-day average ozone minima and integrated ozone mass deficit metrics using simple statistics, but the decline in the 15-day average ozone hole area is not yet statistically significant.