The event that influenced the ozone decrease in southern Brazil in October 2016 was first predicted by a forecast model available at SULFLUX (2018). According to the climatology for the season, when these OSE events occurred (Chubachi, 1984), the poor ozone air mass was expected to arrive around 21 October in southern Brazil. Based on the model prediction, a sounding balloon was launched on 21 October in order to measure the ozone concentration profile over Santa Maria (29.68^∘ S, 53.80^∘ W), southern Brazil, on this specific day. The Hybrid Single Particle Lagrangian Integrated Trajectory (HYSPLIT) model from NOAA, which is one of the first dispersion models used to trace air mass trajectories (Glenn et al., 2017), also confirmed this event, and the result showed the retroactive trajectory towards the south of South America (Fig. 1).

Figure 2TIMED/SABER (blue circle) and AURA/MLS (red circle) sounding locations from 19 to 23 October 2016. On 21 October the balloon sounding (red triangle, launch; orange triangle, apex) and the GPS-RO sounding (green circle).

Download

The sounding balloon was launched in Santa Maria and ascended approximately 31 km with eastward displacement from the launch spot (Fig. 2). The methodology used for the launch followed a pattern used in Natal-RN (5.79^∘ S, 35.20^∘ W) in a study by Witte et al. (2017). The measurements collected by the sounding balloon used here were ozone and temperature.

In addition to the ozone sounding data, ground-based measurements included the total ozone column. The Brewer Spectrophotometer was used to identify the large depletion in the ozone content. It was located at the Southern Space Observatory (OES/INPE) in São Martinho da Serra/RS-Brazil (29.53^∘ S, 53.85^∘ W), which is 30 km from Santa Maria. Ozone and temperature profiles were obtained from the TIMED/SABER and AURA/MLS satellites. OMI satellite images on board the ERS-2 satellite were also used. All instruments will be closely described below.

Figure 3Ozone partial pressure profile as obtained by the ozonesonde (black), AURA/MLS (red), TIMED/SABER (blue) and climatology from ozonesonde balloons campaigns between 1996 and 1998 (grey) (following Guarnieri et al., 2004).

Download

The Brewer Spectrophotometer was designed for spectral irradiation measurements in the ultraviolet (UV-B) range of the solar spectrum. This apparatus makes measurements at five wavelengths, 306.3, 310.1, 313.5, 316.8, and 320.1 nm, which allows the deduction of the total column of the atmospheric gases, such as ozone (O₃) and other atmospheric compounds. This ground-based device can make as many as 40 measurements per day depending on the weather conditions. On 21 October 2016 (295 on the Julian calendar), the Brewer performed 15 measurements, which are considered valid according to the operation of the instrument (Vaz Peres et al., 2017) as well as the days before and after the launch.

NASA's Thermosphere Ionosphere Mesosphere Energetics and Dynamics (TIMED) satellite has a circular orbit of 625 km at 74.1^∘ from the Equator (Cooper, 2004). The data obtained from this satellite allow the need to study the influence of solar radiation and human activities in the upper layers of the atmosphere (from 60 to 180 km). The radiosonde cannot reach this region and this satellite is the only one that can provide global measurements of this wide range of altitude. The sounding of the atmosphere was done using broadband emission radiometry (SABER), which is one of the instruments on board the TIMED satellite. This device is a multi-spectral radiometer that measures daily vertical profiles through near-infrared measurements between 1.27 and 17 µmm, with a typical accuracy above 5 % in most channels and a vertical resolution of about 2 km (Bageston et al., 2011). The data collected through the TIMED/SABER satellite are available from altitudes between 10 and 120 km.

Another satellite used in the present study was AURA, which is part of NASA's Earth Observing System (EOS). The Aura orbit is Sun-synchronous at an altitude of 705 km with 98^∘ inclination. One of the instruments on board AURA is the Microwave Limb Sounder (MLS), which measures the emission of thermal microwaves from different regions and scans the Earth every 24.7 s (French and Mulligan, 2010). The AURA/MLS satellite measures the atmospheric composition, such as temperature, humidity and ozone. In addition, it is also possible to obtain measurements during the day and night around the globe by means of the AURA/MLS satellite, which acquires data even in the presence of glacial clouds and aerosol from infrared, visible and ultraviolet measurements (AURA, 2018).

Figure 4The same as Fig. 3 but for the temperature and the additional GPS-RO profile (green).

Download

The temperature was also compared with data obtained from the Cosmic Satellite Constellation which utilize the radio occultation (RO) technique to retrieve the temperature profiles (Foelsche et al., 2006). The RO technique uses the effect of the radio signal refraction in the atmosphere when the satellites of the Global Navigation Satellite System (GNSS) are occult for the receivers on the ground. Through this technique it is possible to retrieve, besides temperature, other important variables, such as the atmospheric density, water vapour and densities of several constituents (GNSS) (COSMIC, 2018).

The last instrument used in this study was the Ozone Monitoring Instrument (OMI) on board the ERS-2/NASA satellite. It was launched to replace the Earth Probe Satellite (TOMS-EP), which ended its activities in 2015 (Antón et al., 2009). The OMI instrument records data from the total ozone column and other atmospheric parameters related to ozone chemistry and climate, such as NO₂ and SO₂. This instrument can also distinguish different types of aerosols, including smoke, dust and sulfates, in addition to measuring pressure and cloud cover and creating a hyperspectral image in the visible and ultraviolet spectrum (OMI, 2018).

The measurements obtained by the AURA/MLS and TIMED/SABER satellites are in mixing ratio, although the balloon measurements are in partial pressure. Thus, it was necessary to convert the satellite data unit from mixing ratio to partial pressure for a reliable comparison between the two sets of data.

This paper presents a full multi-instrumental analysis of a strong ozone secondary effect over southern Brazil and Uruguay between 21 and 23 October 2016. The ozone and temperature profiles observed during the occurrence of this ozone depletion event were compared to the climatology for Santa Maria between 1996 and 1998 (Guarnieri et al., 2004).

If the measurements of the instrument were not simultaneous, it could imply poor data correlation. However, the instruments in this study did not perform simultaneous measurements at the same location and on/at the same day/time. Nevertheless, the results were efficient in identifying poor ozone air mass displacement and analysis of this event.

Figure 5Total ozone column as measured by the Brewer Spectrophotometer at the Southern Space Observatory (SSO/INPE), São Martinho da Serra/RS-Brazil from 18 to 24 October 2016.

Download

Figure 6Time evolution of the total ozone column as observed by the OMI-ERS2 satellite from 19 to 22 October 2016. Available at OMI (2018). The effect of the ozone hole over South America, especially over Uruguay and southern Brazil, which is the region highlighted in the light blue plume.

Download

According to the instruments presented above, data were collected in southern Brazil and Uruguay. The locations of each ozone and temperature profile collected by the satellites and sounding balloon are in Fig. 2. The TIMED/SABER satellite, which is denoted by a blue circle, passed over the regions of interest on 19 October in Uruguay and in the coastal region of Rio Grande do Sul on 23 October. The AURA soundings, which are denoted on the map by a red circle, passed over Uruguay on 20 October and southern Brazil on 21 October. The ozonesonde was launched on 21 October in Santa Maria (red triangle) and reached 31 km height east of the launching site (orange triangle). In addition to these instruments, a closet temperature profile, which was retrieved from GPS radio occultation (green circle on the map), was used for the temperature profile and obtained by radiosonde for a reliable comparison since the two profiles were very close to each other.

Ozone profiles are shown in Fig. 3 for 19, 20, 21 and 23 October. There were no satellites passing over the regions on 22 October. The vertical dashed lines delimited the peak positions from the climatology and observations at around 24 km height. The ozone profile was obtained over the Uruguayan region (near the coast) at around 14:00 UT on 19 October by the AURA/MLS satellite (red profile), which is depicted in Fig. 3a. This shows a regular behaviour (main peak around 24 km height) that is very similar to the climatology for the state of Rio Grande do Sul (grey profile).

On 20 October the TIMED/SABER satellite (blue profile) carried out measurements at about 110 km to the north of the previous day over Uruguay. These sounding data reveal a considerable decrease in ozone concentration when compared to the climatology (Fig. 3b). In fact, this is the most prominent decrease registered during the observation period, with an opposite peak to the climatology, which was at around 24.5 km height with the same order of magnitude (∼ 120 µhPa). This intense inversion of the layer exposes evidence of an influence of the ozone hole. On 21 October (Fig. 3c), which is the day the ozonesonde was launched, the poor ozone air mass was identified by both the balloon (in Santa Maria) and the AURA/MLS satellite (∼ 130 km to the east of Santa Maria). These two instruments uncovered a significant decrease in the ozone profile at around 24 km height. The balloon profile (black points) presented a larger decrease when compared to the AURA/MLS profile (in red), although it presented positive agreement between the two profiles throughout the coincident altitudes measured. Despite the measurements not being performed at the exact same place, both the balloon and the AURA/MLS satellite presented very similar ozone profiles. These profiles displayed less intense reduction than the previous day over Uruguay and approximately 60 µhPa below the climatology peak (half of the previous day's reduction). Nevertheless, this apparatus unveiled a reversal of the ozone layer at around 24.5 km height and showed the influence of the ozone hole over Santa Maria.

As previously mentioned, there were no measurements for the ozone and temperature profiles on 22 October. However, the Brewer Spectrophotometer registered that the total ozone content was still low on this specific day (Fig. 5). The ozone for 23 October recovered considerably according to the TIMED/SABER satellite data retrieved over the coast of the state of Rio Grande do Sul (Fig. 3d). Although the ozone profile had a considerable reduction on this day (∼ 45 µhPa) at the nominal ozone peak in relation to climatology, the ozone layer was recovering, with its main peak at about 20 km height with a value higher than the climatology maximum.

The temperature response to the high ozone depletion over Uruguay and southern Brazil, which occurred mainly on 20 and 21 October, warrants special attention. A sequence of temperature profiles from 19 to 23 October is presented in Fig. 4 in the same way as in Fig. 3. An ordinary temperature profile that is very similar to the climatology is detailed in Fig. 4a with a tropopause height at around 16 km, which is in agreement with data shown in Fig. 3a, where no record of a significant decrease in the ozone layer was identified. Notwithstanding, with the arrival of a polar air mass over Uruguay on 20 October the temperature profile showed a double tropopause and one inversion layer between these two pauses (∼ 23 km height), being the absolute minimum around 25 km height. These phenomena were strongly linked to the effect of the ozone hole over Uruguay observed on this day. The upper minimum of temperature (at about 25 km) was due to the minimum of ozone registered at this height. The temperature inversion (∼ 23 km), on the other hand, was a consequence of energy deposition (from above) in the residual ozone below the ozone minimum. From 22 to 21 km the ozone continued to decrease with another minimum (∼ 21 km) being registered, which is related to the lower tropopause. The temperature profile above 25 km followed its expected behaviour.

On 21 October (Fig. 4c), the balloon, AURA/MLS and radio occultation (GPS-RO) measurements presented the tropopause right after the tropopause (∼ 19 km) presented by climatology, which is a consequence of ozone content reduction. The temperature measurements from the balloon and GPS-RO satellite showed practically identical values from ∼ 3 km height to the tropopause, even though the balloon and GPS-RO obtained their profiles at quite distinct local times (2.5 h apart) but were nevertheless very close to each other. Furthermore, it is possible to identify an inversion layer around 22 km height, although in this case the main tropopause is located ∼ 5 km below the maximum in ozone concentration, more specifically at ∼ 19 km (from the balloon) or 8 km (from GPS-RO and AURA/MLS). A second minimum in temperature (as observed by the balloon data) is located just below the ozone minimum, which is at about 23 km.

The temperature profile obtained from the TIMED/SABER satellite near the coastal region of Rio Grande do Sul on 23 October can be seen in Fig. 4d. This graph shows that the thermal structure in the upper troposphere and lower thermosphere was already returning to normal conditions. The typical condition in terms of temperature is noted by the high similarity between the TIMED/SABER and climatological profiles, where they match each other around the tropopause (15–17 km heights) and altitudes between 26 and 31 km (the highest altitude reached by the balloon). This confirms that the polar air mass, which is poor in ozone, was already leaving southern Brazil and is therefore in agreement with the ozone recovering process in Fig. 3d at about 20 km height.

The Brewer Spectrophotometer also performed measurements of the total ozone content between 18 and 24 October over São Martinho da Serra/RS-Brazil (29.53^∘ S, 53.85^∘ W) (Fig. 5). Ozone is in Dobson units (DU) and ordinary values were observed around 250 DU for 18 and 19 October. Still, ozone content dropped to an extreme value of ∼ 225 DU on 20 October. Afterwards, the ozone column began to recover and almost stabilized on 23 and 24 October.

The HYSPLIT/NOAA model (Fig. 1) also confirmed the event occurrence by using three distinct altitude levels (∼ 22–28 km). It showed the final position of the backward trajectories over southern Brazil where the air mass, which was poor in ozone, acted on 20 October (Fig. 1a) and 21 October (Fig. 1b). The backward trajectories correspond specifically to the heights of 22 (red line), 24 (blue line) and 28 km (green line) that are the same height levels used by Bittencourt et al. (2018) in this special issue for the analysis of the potential vorticity associated with this event.

The height levels in the HYSPLIT/NOAA model were chosen according to the height where the ozone maximum was identified. These trajectories show the arrival of the ozone filaments directly from the Antarctic region, which originated from the main ozone hole. Finally, the images of the OMI-ERS2 satellite (Fig. 6) also demonstrate the Antarctic ozone mass over a wide region in Argentina, Uruguay and southern Brazil. The air mass with low ozone concentration was more dominant from 20 to 22 October, especially in southern Brazil and Uruguay (the regions analysed here).

The influence of the Antarctic ozone hole may reach mid-latitudes. Results of this study corroborate previous ones such as performed by Kirchhoff et al. (1996), who reported an ozone content reduction in Santa Maria for a couple of days in October of 1993. Their results suggested that the ozone content presented two minimums: ∼ 250 DU on 19 and 28 October. In that study, the monthly average data were taken for two stations, one with data collected for 19 years and the other for 13 years, with typical total ozone content reported around 290 DU. In the present study, a larger decrease was observed when compared to Kirchhoff's, with a minimum of approximately 225 DU on 20 October, which can be considered a very intense influence of the ozone hole over southern Brazil. Similarly, validation of the Brewer Spectrophotometer data using daily averages from the TOMS satellite is also presented and, therefore, a positive correlation was observed. Additionally, the present study also confirmed the influence of the Antarctic ozone hole at mid-latitudes by using validated profiles of the ozone sounding with the TIMED/SABER and AURA/MLS satellites. The Brewer Spectrophotometer, backward trajectories by HYSPLIT/NOAA and images from the OMI-ERS2 satellite were also used and exhibited excellent agreement, thus corroborating data obtained from our sounding balloon.

Another study that showed ozone profiles created by ozonesonde for the region of Santa Maria was by Guarnieri et al. (2004), with data from November 1996 to April 1998, which were used to build the climatology used in this study. They also built the ozone climatology by seasons (autumn, winter, spring and summer), based on 35 ozonesonde profiles, showing that ozone content is higher in winter and spring (with a peak of ∼ 135 µhPa at ∼ 23 km height) than in summer and autumn. In the present study, a minimum value of 70 µhPa was observed at the expected ozone peak (∼ 24 km) in spring, which corresponds to a value of ∼ 52 % of typical ozone content observed by the climatology for spring. A maximum value of ∼ 110 µhPa (22 km height) was observed, while the climatology for the same altitude was about 120 µhPa. This intense ozone layer depletion was explained by the influence of the Antarctic ozone hole that reaches mid-latitudes in austral spring. This effect was remarkable from 20 to 21 October 2016. Furthermore, Guarnieri et al. (2004) reported a temperature profile climatology by using the same 35 records of ozone sounding data over Santa Maria. These data presented the tropopause at about 16–17 km height with a minimum temperature of −70^∘ C, whereas in our observations the tropopause was around 19 km with a corresponding temperature of ∼ −77^∘ C, which is much colder than the climatology at the same height. The characteristics of these temperatures show the impact of the ozone hole on the temperature.

In a future study, we intend to show the behaviour of UV-B radiation during an extreme event of influence of the ozone hole through the inclusion of the calibrated UV data from the Brewer Spectrophotometer as well as satellite data, as this would show the impact of the ozone hole on UV radiation.

All SABER satellite data used in this work have been downloaded using the SABER Custom Data Services tool available at http://saber.gats-inc.com/data.php. AURA/MLS satellite data have been downloaded by a user, available at https://aura.gsfc.nasa.gov (AURA, 2018). OMI-ERS2 satellite data are available at https://aura.gsfc.nasa.gov/omi.html (OMI, 2018). Radio occultation satellite data are available at http://www.cosmic.ucar.edu/ (COSMIC, 2018). The climatological data were obtained by Guarnieri et al. (2004). The ground-based data (Brewer Spectrophotometer) are not available online, but can be solicited directly by e-mail to José Valentin Bageston (bageston@gmail.com). The balloon data are not available online, but can be solicited directly by e-mail to Damaris Kirsch Pinheiro (damaris@ufsm.br).

The authors declare that they have no conflict of interest.

This article is part of the special issue “Space weather connections to near-Earth space and the atmosphere”. It is a result of the 6^∘ Simpósio Brasileiro de Geofísica Espacial e Aeronomia (SBGEA), Jataí, Brazil, 26–30 September 2016.