Corresponding author: S. Lal, Physical Research Laboratory, Navrangpura, Ahmedabad-380009, India. (email@example.com)
 The measurements of vertical distribution of ozone have been made over the Bay of Bengal and the Arabian Sea in spring 2006 during a cruise campaign, namely, the Integrated Campaign for Aerosol, Gases and Radiation Budget. The average tropospheric columnar ozone (TCO) values are found to be 36.1 ± 6.9 Dobson unit (DU) and 41.7 ± 5.0 DU over the Bay of Bengal and the Arabian Sea, respectively. In contrast to TCO, ozone mixing ratios are higher by about 10 ppbv in the lower 3 km over the Bay of Bengal due to Indo Gangetic outflow above the marine atmospheric boundary layer. Major contribution in the higher TCO value over the Arabian Sea is, possibly, by stratospheric intrusion when ozone in the upper troposphere is higher by about 20 ppbv. The lowest columnar ozone content of 22.4 DU was observed on 30 March in the central Bay of Bengal due to convective activity resulting into lower ozone throughout the troposphere, except between 6 and 10 km altitude. These near simultaneous ozone observations over the Bay of Bengal and the Arabian Sea in spring season have revealed the role of regional and long-range transport and local dynamics on the vertical distribution of ozone over these tropical marine regions.
If you can't find a tool you're looking for, please click the link at the top of the page to "Go to old article view". Alternatively, view our Knowledge Base articles for additional help. Your feedback is important to us, so please let us know if you have comments or ideas for improvement.
 Tropospheric ozone is a precursor of hydroxyl radicals (OH), which determine the oxidation capacity of the troposphere [Jacob, 1999]. It is a potential greenhouse gas [e.g. Gauss et al., 2003] and has deleterious effects on the human health and plants at the elevated mixing ratios in the atmospheric boundary layer [e.g. Adams et al., 1989]. Study of vertical distribution of ozone provides information on the chemistry as well as on the dynamical factors affecting its distribution. Tropospheric ozone has two main sources. In the upper troposphere (from above about 10 km to the tropopause), it is mostly transported from the ozone rich stratosphere. While in the middle (approximately 3–10 km) and lower troposphere (below about 3 km), it is formed by the photo-oxidation of pollutants like CO, CH4, and non-methane volatile organic compounds in the presence of NOx and sunlight [Crutzen, 1995 and references there in]. Ozone in the troposphere has highly variable lifetime from a few days in the moist tropical marine atmospheric boundary layer (MABL) (from the ocean surface to about 2 km height) to months in the free troposphere (from the top of the MABL up to the tropopause) [Fishman et al., 1991]. Hence, it is used as a tracer for transport in the free troposphere. The photochemical processes are rapid in the tropical Asian region due to higher levels of OH radicals, availability of intense solar radiation, and higher emissions of anthropogenic pollutants. Higher levels of OH radicals in the tropical region are due to higher levels of water vapor and actinic flux. The marine regions in the lower troposphere surrounding India get polluted from the transport of pollutants including ozone as well as its precursors from the Indian subcontinent [Lelieveld et al., 2001]. However, long-range transport can affect ozone distribution in the free troposphere [Cooper et al., 2010].
 The changes in ozone mixing ratios have important implications toward radiative forcing and climate change when they occur in the middle and upper troposphere [Gauss et al., 2003]. The free tropospheric ozone distribution is often affected by convective processes and stratosphere-troposphere exchange. The intrusion of ozone-rich midlatitude stratospheric air due to isentropic transport across the tropopause is an important phenomenon for the ozone enhancement in the tropical upper troposphere [Stohl et al., 2003]. Another important phenomenon that influences the tropospheric ozone distribution is deep convection [Folkins et al., 2002; Kley et al., 2007]. The tropical regions are characterized by strong upwelling during most of the year. This transports boundary layer pollutants into the free troposphere, from where horizontal winds can carry them to distant regions.
 Earlier, measurements of vertical distribution of ozone have been made during February–March 1998 and 1999 as part of the Indian Ocean Experiment (INDOEX) over the Indian Ocean and the Arabian Sea (AS) [Mandal et al., 1999; Zachariasse et al., 2000, 2001; Lal and Lawrence, 2001; Lelieveld et al., 2001; Chatfield et al., 2007]. These studies showed a layer of enhanced ozone and its precursors in the lower troposphere above the MABL over the study region due to transport of pollutants from the Indian subcontinent mostly during winter season. The north-easterly wind carried polluted air from the Indo-Gangetic Plain (IGP) to the receptor locations via the Bay of Bengal (BoB) [Lelieveld et al., 2001]. Since the transport from the north Indian region reaches first over the BoB, this region should get affected more. In view of this, several cruise campaigns have been conducted over the BoB during 2001–2009 to study transport of pollutants using surface-based measurements [Chand et al., 2003; Naja et al., 2004; Lal et al., 2006, 2007; Sahu and Lal, 2006; Sahu et al., 2006; Nair et al., 2011; Srivastava et al., 2012]. These observations showed higher levels of surface ozone and its precursors over the northern part of the BoB. Additionally, average levels of these trace gases are found to be higher over the BoB than over the AS [Srivastava et al., 2012]. Limited observations of vertical distributions of ozone were made only in the southern part of BoB during the INDOEX 1999 campaign [Lelieveld et al., 2001].
 An Integrated Campaign for Aerosols, Gases and Radiation Budget (ICARB) was conducted over large areas of the BoB and the AS during March–May 2006 [Moorthy et al., 2008]. This period of the campaign was chosen to cover both the regions synoptically in the premonsoon season. This is also the period of maximum photochemical activity due to lower solar inclination over these regions. Measurements of ozone profiles were made using balloon borne ozonesondes during this campaign. This is the first systematic study of the vertical distribution of tropospheric ozone covering both the marine regions. This valuable data set provided an excellent opportunity to investigate the effects of transport, convective processes, and other physical and chemical effects on the vertical distribution of ozone over these marine regions. The ozone distributions in the lower part of the troposphere up to 4 km height over both the marine regions have been analyzed in detail and published in Srivastava et al. . The present work is the extension of this work covering the entire troposphere including the 0–4 km height for the sake of completeness.
2 Campaign and Measurement Techniques
 The ozonesondes were launched from the ocean research vessel Sagar Kanya during its cruises SK 223A and SK 223B over the BoB and AS, respectively. The cruise track is shown in Figure 1. The cruise started on the east coast of India from Chennai port (13.1°N, 80.3°E) on 18 March 2006 and ended at Goa port (15.4°N, 73.8°E) on the west coast of India on 11 May. The total duration of the cruise including both the phases was 55 d. Further details of the cruise are given in Srivastava et al. .
 Vertical profiles of ozone, pressure, relative humidity, and temperature were obtained using balloon borne ozonesondes coupled to radiosondes. The ozonesondes are based on electrochemical concentration cell (ECC) [Komhyr and Harris, 1971]. Accuracy of ozone measurements using these sondes has been found to be ±5–10% up to 30 km height [Smit et al., 2007]. The meteorological parameters such as air temperature, relative humidity, and pressure were measured using radiosondes (Vaisala, RS-80). The temperature and pressure sensors have the accuracy of ±0.3°C, and ±0.5 hPa, respectively, below about 20 km. However, the relative humidity sensor has the accuracy of about ±2% near the ground and it decreases to ±15–30% in the 5–15 km height [Kley et al., 1997].
 These sondes were launched at around 1000 h Indian Standard Time (geomagnetic time +5.5 h). A total of 13 ozonesondes and radiosondes were launched from 18 March to 12 April 2006 from the ship during the first phase of the cruise covering an extensive area of the BoB and some parts of the Indian Ocean. In the second phase, 15 balloon soundings were made over the AS. These balloon sounding locations are marked on the cruise track (Figure 1).
3 General Wind Pattern
 The wind patterns averaged for the first leg of the cruise period (18 March to 12 April 2006), when the ship was over the BoB, are shown in the right panel of Figure 2 for 925 hPa, 500 hPa, and 200 hPa. These are based on the reanalysis data taken from NCEP (National Centers for Environmental Prediction, USA). The winds at 925 hPa were westerly over northern BoB. However, the winds were slower and changing direction in the central and southern BoB. At 500 hPa and 200 hPa, the winds were westerly over the land region as well as over the northern BoB. In the southern part, the winds were easterly at 500 hPa and south easterly at 200 hPa.
 The wind patterns for the second leg of the cruise, when the ship was over the AS, are shown in the left panel of Figure 2 at all the three pressure levels. These are the average winds during the cruise period of 18 April to 10 May 2006. The winds at 925 hPa over the northern AS were westerly and turning over the eastern region to become northerly. At 500 hPa, the winds were westerly over the northern AS, turning in the central part to become north-easterly over the southern part of the AS. The wind pattern further changed at 200 hPa to be westerly over the northern part and south-westerly over the central and southern AS.
4 Results and Discussion
4.1 Variation of Integrated Tropospheric Ozone
 The tropospheric columnar ozone (TCO) is obtained by integrating individual profiles observed up to the cold point tropopause height, which was obtained from the temperature profiles measured by radiosondes, which were flown together with the ozone sondes in each balloon flight. The TCO varied from 22.4 to 45 DU, with a minimum of 22.4 DU on 30 March in the central BoB and a maximum of about 45 DU on 20 March near the coast in the northern BoB (Figure 3). It was higher over the western and northern regions, which generally experience the continental outflow from the Indo-Gangetic Plain. The measurements near Chennai also show higher TCO (about 42 DU). TCO values over the southern region are about 30.1 DU and 28.6 DU on 8 and 10 April, respectively. These values are similar to those reported over a segment of the southern BoB during the INDOEX 1999 campaign [Chatfield et al., 2007]. The average TCO over the BoB during this period was 36.1 ± 6.9 DU.
 The TCO over the AS is different than that over the BoB. Except near Trivandrum, where tropospheric content was only 30.3 DU, all other measurements show columnar values higher than 36 DU. Highest values around 47.5 DU have been observed over two places, one off the coast of Gujarat and the other one over the south-eastern part of the AS. Higher TCO is generally observed in the northern AS. The average value of tropospheric content over the AS is 41.7 ± 5.0 DU, which is higher than over the BoB by about 15%. The observed column content during the INDOEX 1999 varied from about 30 to 50 DU for the latitude region of 5–20°N over the AS with a positive gradient from south to north [Chatfield et al., 2007]. The tropopause temperature varied mostly between −81 and −85°C while the height of this cold point tropopause varied in the range of 16.5–18.2 km (figure not shown). However, no strong relationship of the tropospheric ozone content either with tropopause temperature, tropopause height, or with average value of ozone in the first 1 km was found during this campaign.
 We have compared the computed TCO values based on balloon data with one derived from Ozone Monitoring Instrument (Figure 4). In the satellite-based algorithm, the profile of partial ozone columns in DU is retrieved at 24 layers from the surface to ~60 km (~2.5 km thick per layer) from backscattered UV (BUV) radiances in the spectral region 270–330 nm using the optimal estimation technique [Liu et al., 2005; Liu et al., 2010]. The NCEP reanalysis tropopause height is used as one retrieval level to separate the stratosphere and troposphere. Tropospheric and stratospheric ozone columns are integrated from the retrieved ozone profile. The spatial resolution of the retrievals equals to that of UV-1 measurements, 13 × 48 km2 at nadir. The estimated one sigma retrieval errors in TCO are typically within 2–4 DU. Figure 4 shows a comparison of the two data sets. The correlation coefficient (r2) is 0.78. Even the lowest ozone observed on 30 March over the BoB compares well with the satellite data. The differences are within the errors of the estimations.
4.2 Vertical Distribution of Ozone over the Bay of Bengal
 Contour plot of vertical distribution of ozone is shown in Figure 5 for the BoB. Higher levels of ozone up to 100 ppbv are observed up to an altitude of about 13 km in the northern part of the BoB as well as in the central part of the BoB. There are regions of low values (<40 ppbv) around 28–30 March and 8–10 April throughout the troposphere. Lower values of ozone are also observed above 13 km height. In the lowest part of the troposphere, the height of the low ozone region increases from almost zero to about 3 km as the ship moved from north to south.
 We have grouped these profiles in three categories, namely, costal, central, and southern regions. The coastal region includes ozone profiles measured on 20 and 22 March and 1 and 2 April. The central group consists of profiles measured on 24, 26, 28, and 30 March and 4 and 6 April. The southern group consists of profiles measured on 8 and 10 April. These average profiles are shown in Figure 6. The coastal region average profile shows very clearly higher ozone throughout the troposphere as compared to other two regions. Ozone values are mostly in the range of 60–80 ppbv in the 3–13 km height. The TCO values for the coastal region are higher than 40 DU. Srivastava et al.  have shown that higher ozone below 4 km was due to transport from the IGP. We find that higher ozone above 4 km is due to transport from the North African region via the North-Indian region. The winds are north-westerly at 500 hPa also (Figure 2). The trajectories over different regions (coastal, central, and southern) of the BoB are shown in Figure 7. These 7 d kinematic back trajectories are computed for different heights and collocated with balloon launch locations and time using Hybrid Single-Particle Lagrangian Integrated Trajectory (HYSPLIT) model [Draxler and Hess, 1997]. The trajectories for balloon launches over the coastal BoB show air masses mostly coming from Africa via central-western India in the 4–12 km region. Figure 6 also shows higher ozone in this altitude regime.
 The MOZAIC (Measurement of OZone and water vapour by AIrbus in-service airCraft) measurements of ozone profiles are available over Chennai (formally known as Madras) [Marenco et al., 1998, Sahu et al., 2011]. However, measurements in the recent time are very limited. High ozone values in the range of 60–75 ppbv in the 2–7 km height region have been observed on 26 April 2001 from MOZAIC measurements over this site [Sahu et al., 2011]. The average ozone profiles over the central and southern regions show lower ozone compared to coastal region throughout the troposphere. The ozone values over the central BoB are slightly higher compared to those over the southern BoB in the 5–10 km region. This could be due to back trajectories coming from IGP/Africa over the central BoB whereas almost all trajectories over the southern region are from the marine regions adjoining in Southeast Asia. The near surface level ozone values for the central and southern BoB are in the 20–30 ppbv range as compared to coastal region values of about 40 ppbv. Ozone values are mostly in the range of 40–60 ppbv in the 3–12 km height for these two regions. Large variability is observed in the central profiles with one sigma deviation in the range of 30–40%.
4.3 Vertical Distribution of Ozone over the Arabian Sea
 Ozone distribution over the AS (Figure 8) is very different from that over the BoB. Lower ozone values <30 ppbv are observed in the boundary layer throughout the AS. However, low ozone regions are not observed in the free troposphere as observed over the BoB. Also, ozone levels are higher (up to 100 ppbv) in the 13–15 km in most of the regions unlike over the BoB. It is also to be noted that over the northern AS, very high ozone values (up to 120 ppbv) are observed in the 9–12 km height range on 5 and 7 May.
 As for the profiles over the BoB, we have done similar grouping for the AS region also. The profiles observed on 27, 28, and 29 April and 4 and 9 May are grouped as coastal profiles. The profiles on 23 and 25 April and 1, 3, 5, and 7 May are grouped as central region profiles. The southern set consists of profiles measured on 12, 19, and 21 April. The three average profiles with one sigma standard deviation over the AS are shown in Figure 9. These profiles are very different than over the BoB. All the three profiles show low ozone around 20 ppbv near the surface. Ozone keeps on increasing from the lowest point till about 8–10 km. The highest ozone for the southern region profile is about 80 ppbv at about 8 km, whereas the central average profile has about 86 ppbv at 10 km. However, there is large variability in the profiles. Most of the back trajectories for the central AS are coming from higher latitudes compared to the coastal AS (Figure 7). This may result in higher ozone for profiles over central AS compared to coastal AS (Figure 9). Some of the profiles showing higher ozone in 10–12 km region are discussed further in section 4.5. Most of the trajectories over the southern AS, except in the 6–10 km altitude, are arriving from the cleaner marine regions of the Indian Ocean, explaining the lower ozone values. The higher ozone in the 6–10 km region over the southern AS (Figure 9) may be due to transport from higher latitudes (Figure 7). The southern profile shows lowest ozone as compared to other profiles in the 10–15 km height. As mentioned earlier, the TCO values in the southern region are in the range of 30–38 DU, whereas in the northern region, these are as high as 47 DU.
4.4 Effect of Convective Activity
 The average TCO over the BoB region was observed to be 36.7 DU, while on 30 March 2006 at a location of 12.39°N, 87.92°E, it was found to be only 22.4 DU, which is approximately 40% lower. Vertical profiles of ozone along with specific humidity and estimated equivalent potential temperature on this day are shown in Figure 10. The surface ozone on this day was found to be 14.6 ppbv, which was very low from the average value of about 20 ppbv in the southern BoB. The lower ozone continued up to about 6 km height. However, higher ozone in the 6–10 km region has been observed, which could be due to advection of ozone-rich air from continental regions.
Aspliden (1976) showed that strong convective activity results in almost constant values of equivalent potential temperature. During a convection event, ozone-deficient air is lifted from the MABL to the upper troposphere. Effects of convection on the vertical distribution of ozone have been discussed in detail by several workers [Kley et al., 1997, 2007; de Laat et al., 1999 and Folkins et al., 2002, 2006]. The equivalent potential temperature (θe) profile on 30 March was almost constant (~327 K) between 2–6 km (Figure 10) unlike other days where it was increasing with altitude. However, above 6 km, the θe profile was similar to other days. Further, it is observed that wind streamlines on 30 March at 300 hPa transported continental air to the site of measurement like any other day. Thus, O3 levels on 30 March are expectedly normal at higher altitudes.
 Further, on that day, the vertical velocity (omega >0.2 Pa/s) was directed downward over Eastern-Central India. However, at 600 hPa, air was almost stagnant at the measurement location and there was no net flow from any direction. This could be due to relatively higher pressure over this region (Figure 11). As a result, transport of ozone and/or precursors into this region was cut-off. Such meteorological features were not observed during other days. Further, the back trajectories below 6 km (Figure 12) span mostly oceanic regions between Southeast Asia and Australia. Thus, the air masses possibly affecting the measurement location were themselves not loaded with continental pollutants. The cloud cover over this region would have hindered photochemical production of ozone. All these processes seem to have resulted in conspicuously depleted ozone layer at lower altitudes (between surface and 6 km) on 30 March.
 Back trajectory analysis for 30 March reveals influence of marine air masses except at 1 km from India and for the height of 6–10 km (nonconvective region) from Africa. Measurements of vertical distribution of ozone were made during 2005 and 2006 over a West African site as part of African Monsoon Multidisciplinary Analysis program. Average ozone during March 2006 shows 70–75 ppbv in the 6–10 km height region [Thouret et al., 2009]. On the contrary, trajectories for the height range of 2–6 km and 10–18 km show transport from the Pacific Ocean side where a small cyclonic formation took place during that time. The height information from the back trajectories shows lateral transport for the height range from 1 to 6 km and downward transport in the height range of 6–10 km whereas from 10–18 km, it shows up lift of air mass from 4–5 km of altitude range (Figure 12 lower part).
4.5 Stratosphere-Troposphere Exchange over the Arabian Sea
 Average TCO over the AS was observed to be 41.7 ± 5.0 DU as compared to 36.1 ± 6.9 DU over the BoB. Also the maximum ozone mixing ratio was as high as 120 ppbv over the northern AS against about 100 ppbv over the northern BoB. Figures 8 and 9 clearly show that the northern and central AS has higher ozone in the 8–12 km height. Ozone profiles during 4–7 May, particularly on 5 (both morning and afternoon) and 7 May, show enhanced values in the 8–12 km height (Figure 13). High ozone values on these days could be attributed to the downward transport of ozone-rich air as indicated by potential vorticity (PV) maps (Figure 14). PV is a useful tracer to study the transport from the stratosphere to the troposphere as its values are very different in the two regions. PV value greater than 1.6 is an indicator of stratospheric intrusion [Poulida et al., 1996; Pan et al., 2004; Cristofanelli et al., 2006]. PV values are calculated using a Weather Research and Forecasting (WRF) model for flight days on 3, 4, 5, and 7 May. The model domain is centered at 25°N, 80°E covering nearly the entire South Asian region (4–40°N, 55–105°E) at a horizontal resolution of 45 km × 45 km. The initial and boundary conditions for meteorological fields are obtained from 6 hourly NCEP final analysis (FNL) fields (http://dss.ucar.edu/datasets/ds083.2/data/). More details of the model setup can be seen in Kumar et al. . Figure 14 clearly shows that stratospheric air mass begins to intrude the upper troposphere on 3 May. The event was most intense on 5 May (PV >2 unit) in the vicinity of the balloon launching location, thus explaining the very high ozone in the 8–12 km region. Conspicuous high ozone is also observed on 7 May, even though the event had retreated. The high ozone on 7 May could be attributed to the remnant ozone-rich air that had been brought down by the event over to this region. Both PV (Figure 14) and back trajectory analysis (Figure 7) suggest that the observed ozone peaks were associated with air of stratospheric origin from higher latitudes. Transport of stratospheric air containing higher ozone levels into the troposphere has also been observed over the Indian Ocean and Central Europe [Zachariasse et al., 2000].
5 Summary and Conclusions
 Measurements of the vertical distributions of ozone have been made over the BoB and the AS during the ICARB campaign (18 March to 9 May 2006) to study variability in tropospheric ozone distribution. A total of 28 balloon soundings carrying ozonesondes and radiosondes were conducted over these marine regions during this campaign.
 The tropospheric columnar ozone (TCO) is found to be 36.1 ± 6.9 DU over the BoB and 41.7 ± 5.0 DU over the AS. The lowest TCO (22.4 DU) was observed on 30 March 2006 over the BoB along with the lowest ozone mixing ratio of about 15 ppbv at the surface due to a combination of several meteorological features. The vertical profile observed on this day shows lower ozone values in the entire troposphere except in the height region of 6–11 km and around 14 km. The relatively higher ozone values in the 6–11 km are due to transport of ozone and possibly its precursors from the northern African region.
 Layered structures have been observed in the free tropospheric ozone distribution more frequently over the AS. This indicates that the effects of different source regions on ozone distribution are changing with height. Such layering had also been observed over the Indian Ocean and over the AS during the INDOEX 1998 and 1999 campaigns [Zachariasse et al., 2000] and over other marine regions [Newell et al., 1999].
 Vertical distributions of tropospheric ozone over the two tropical marine regions show distinct features (Figure 15). Higher ozone levels are observed in the lower troposphere (up to 3 km height) over the BoB, but above this height, ozone levels are higher over the AS. This is due to transport of ozone over the northern BoB from the Indo-Gangetic Plain in the lower height and to effects of convection over the central BoB. Intrusion of ozone-rich air from higher heights and latitudes is believed to be responsible for higher ozone mixing ratios over the AS. The upward motion over the BoB could be a part of the major Walker circulation. Back trajectory analysis and potential vorticity in the northern AS show downward motion on several days.
 We present valuable measurements of ozone vertical profiles from surface to tropopause over the data sparse tropical Asian marine region for the first time. The present work highlights the importance of photochemistry in the continental outflow containing anthropogenic air pollutants in the lower 3 km over the BoB and possible intrusion of stratospheric air enhancing ozone mixing ratio in the upper troposphere over the AS. The frequency of these events is still not known in the absence of adequate ozonesonde data. This requires regular measurements of ozone distributions for better understanding of atmospheric chemistry over the tropical marine regions.
 We are thankful to PRL and ISRO GBP ATCTM for the encouragement and support. We also thank the ICARB team and crew members of the Sagar Kanya ship for making these measurements possible. We are grateful to the NCEP for providing free access to the wind data. We are highly grateful to the anonymous reviewers for their valuable comments and suggestions, which have improved the quality of the MS significantly.