Evidence for anthropogenic influence over the central North Atlantic

Contributed by
K.S. Law, University of Cambridge, UK,¹ S.A. Penkett,² C.E. Reeves,² M.J. Evans,^1,6 J.A. Pyle,¹ S. Bauguitte,^2,7 T.J.Green,² B.J. Bandy,² G.P. Mills,² H. Barjat,³ D. Kley,⁴ S.Schmitgen,^4,8 P.S. Monks,⁵ G.D. Edwards,⁵ J.M. Kent,³ K. Dewey³ and A. Kaye³.

---

Introduction

The Azores summer detachment

The UK Meteorological Office C130 aircraft flew to the island of Santa Maria (35°N, 25°W) on 13 September 1997 and returned on 23 September 1997. From this base it completed 6 flights over the middle of the North Atlantic Ocean. The main chemical objectives of the flights included a study of ozone/CO relationships in changing water vapor regimes, in situ photochemical ozone production and loss rates, perturbation of the photochemical equilibrium of peroxides, a study of NO_Y/ozone relationships, and of the chemical form of NO_Y. Another major objective was to survey the composition of air over the North Atlantic Ocean between the North American and European/North African continents and the extent to which it is impacted by air pollution. The flights therefore contained many vertical profiles from the surface up to 8km to detect the presence of polluted layers seen previously on flights over the North Atlantic in 1993 during the NARE1 campaign (e.g., see Wild et al. [1996]).

Evidence for long-range pollution transport

During all the flights, to a greater or lesser extent, there was clear evidence for long-range transport of pollutants from continental regions, especially in the mid-troposphere, over the North Atlantic. One of the most dramatic examples was observed on 14 September 1997 when the C-130 flew into air circulating around Hurricane Erika, which had moved north-eastwards from the West Indies to the Azores. This produced some spectacular effects, including the transport of polluted air from the southeastern USA into the mid-Atlantic free troposphere south of the Azores. Trajectory analysis (not shown) suggests that frontal systems situated over the North American continent 5-6 days earlier had contributed to the uplift of pollution into the free troposphere over the North Atlantic and into the path of Hurricane Erika.

Figure 1. Vertical profile of trace gas concentrations measured on the United Kingdom C-130 Meteorological Research Flight aircraft on 14 September 1997 as part of the UK ACSOE NARE2 campaign (see Penkett et al. [2001] for details).

On 14 September 1997 the aircraft initially flew southwards from Santa Maria, then west into the cloud streamers emanating from the eye of the Hurricane, and then back to Santa Maria through air that had clearly been uplifted. The aircraft also cut vertically through the out-flowing air and a profile of the concentration of ozone, CO and total reactive nitrogen (NO_Y) are shown in Figure 1. High concentrations of these pollutants are clearly visible in layers at altitudes between 5 and 7km. The ozone concentration in these layers reaches 100ppbv, the NO_Y concentrations exceed 1ppbv and both are highly correlated with CO indicating a boundary layer rather than a stratospheric source.

The trajectory analysis showed that the polluted layers observed between 5 and 7km had been uplifted from the polluted boundary layer over the southeastern United States. The air was uplifted to about 9km over 2 days, then it slowly subsided for about 4 days before being intercepted by the aircraft. Lower down in the profile the air had been slowly subsiding over several days or had traveled in the marine boundary layer, where ozone is destroyed efficiently by sunlight and high water vapor concentrations. The ozone concentration close to the surface was only 40ppbv. This is typical of ozone concentrations measured during the ACSOE Mace Head (Ireland) experiment in the summer of 1996 in air from North America that had crossed the Atlantic close to the surface (e.g., see Evans et al. [2000]). Pollution therefore does appear to be mostly transported from the continents in the mid-troposphere over the North Atlantic.

A further point worth emphasizing is the marked anti-correlation between ozone and water vapor seen in the profiles. This is also shown in Figure 1 although the detail is difficult to discern because above 3km the water concentration is close to the detection limit of the sensor, that is, it is very dry air. This is not stratospheric air though, as shown by the high CO and CN (not shown) concentrations, by the trajectories, and by the low PV values calculated along their course that did not exceed 1 PV unit over the 6 days. The conclusion to be reached here is that an anticorrelation between ozone and water vapor does not necessarily indicate a stratospheric origin for the ozone.

The general picture of the atmosphere over the Atlantic Ocean during the sampling periods is therefore of long-range transport of ozone pollution from the continents at altitudes between 3 and 8km interspersed with uplifted marine boundary layer air. There is also evidence, at least in the spring, that ozone production can still take place in the free troposphere in plumes of polluted air. Only rarely was any trace of stratospheric influence detected. Below 3km there is evidence both for less polluted air being transported from the African continent, and for clean marine air in which photochemical destruction of ozone appears to dominate. The layering, seen in all the ACSOE flights, is a very general feature and is undoubtedly related to the overall vertical and horizontal transport processes that occur in the troposphere. See Penkett et al. [2001], Bauguitte et al. [2001], Reeves et al. [2001] and Edwards et al. [2001] for further discussion of these results.

The relationship between ozone and CO has been used previously to estimate the amount of ozone being exported from the continent into the background troposphere (e.g., see Parrish et al. [1993]; Chin et al. [1994]). This is because CO is a major precursor to ozone and is emitted in large quantities in industrial regions. Here, O₃:CO relationships were examined in air masses sampled during the 1997 summer C-130 campaign to investigate factors affecting the concentration of summertime ozone in the troposphere between ~1 and 8 km. In particular, we attempted to discern whether it is possible to separate out the effects of photochemical processes from mixing processes. Results are described more fully in Law et al. [2001].

Figure 2. Plot of ozone and CO concentrations measured from UK C-130 aircraft during flights from Santa Maria, Azores over the central North Atlantic in September 1997 as part of the UK ACSOE NARE2 campaign (see Law et al. [2001] for details).

Figure 2 shows a composite of all ozone and CO from flights out of the Azores. What is interesting is that different ozone and CO concentrations and ratios (varying from 0.3 to 0.8) exist at different levels of water vapor (not shown). In particular, higher ratios are evident in drier air masses. As well as higher ozone, these air masses also had higher levels of CO, CN, NO_Y and NO, indicating that they were more polluted and generally they were sampled in the mid rather than the lower troposphere (see earlier discussion). Air masses with lower ratios were moist and had lower ozone and CO concentrations and originated largely over the North Atlantic. What is leading to these differences? Is it photochemistry or mixing with 'background' air?

The Cambridge Lagrangian photochemical model, CiTTyCAT, described by Evans et al. [2000], was run along 5day, three-dimensional back trajectories calculated using meteorological analyses from the European Centre for Medium Range Weather Forecasts (ECMWF). Two typical cases were examined based on the analysis of data described earlier. Firstly, air originating from the polluted North American boundary layer that was uplifted, most probably by frontal systems [see Cooper et al., 2001] was considered and, secondly, air originating from the clean marine boundary layer over the North Atlantic. Results were compared to observed O₃:CO ratios and their dependence on precursor concentrations, water vapor and mixing with background air were examined.

The most important result is that a combination of both photochemistry and mixing are required to reproduce the observed O₃:CO ratios and concentrations. In polluted airmasses ozone can increase due to photochemical production but also due to lower water vapor, in air that has been uplifted and dried, leading to less photochemical destruction (also see Wild et al. [1996]). However, CO concentrations, which are not really affected by photochemistry over a 5-day period, were calculated to be much higher than observed over the North Atlantic, suggesting that mixing with background air must also occur. Mixing parcels with typical 'background' concentrations leads to increased O₃:CO ratios and decreased ozone and CO concentrations in line with observed values. In the case of clean marine boundary layer air, transported in the lower troposphere, mixing serves to maintain the observed ratio of O₃:CO around 0.2-0.4 by mixing in higher ozone descending from aloft. Therefore, the observed distribution of ozone and CO shows that both chemical and dynamical processes are important in determining the distribution and ratios of these tropospheric trace gases.

Conclusions

The quality of trajectory analysis is now such that many individual features of the chemical composition of air along the flight tracks of research aircraft such as the UK C130 can be accounted for with confidence. This includes uplift of continental boundary layer air, transport of marine boundary layer air, and long-range transport of pollution in the troposphere across the North Atlantic Ocean.

High concentrations of ozone are present in layers over the Atlantic between 3 and 8km in association with elevated concentrations of anthropogenic tracers such as CO, CN and NO_Y. The ozone concentration declines in the marine boundary layer, probably due to increased removal efficiency.

There is evidence that a significant part of the ozone observed over the North Atlantic in summer (and spring) has a pollution origin. The influence of ozone from the stratosphere appears to be small, at least below altitudes of 8 km.

Transport and mixing processes within the troposphere, as well as photochemistry, are clearly important in generating observed trace gas distributions over the central North Atlantic.

Acknowledgments

The ACSOE program was funded by the UK Natural Environment Research Council. Trajectory data was provided by ECMWF via the British Atmospheric Data Centre.

¹Centre for Atmospheric Science, University of Cambridge, Cambridge, England, United Kingdom.

²School of Environmental Science, University of East Anglia, Norwich, England, United Kingdom.

³Meteorological Office, DERA, Farnborough, Hampshire, England, United Kingdom.

⁴Institut fur Chemie, Kernforschungsanlage, Jülich, Germany

⁵Department of Chemistry, University of Leicester, Leicester, England, United Kingdom.

⁶now at Center for Earth & Planetary Physics, Harvard University, Cambridge, MA.

⁷Now at British Antarctic Survey, Cambridge, England, United Kingdom.

⁸LSCE, Unité mixte CEA-CNRS, Gif sur Yvette, France.

References

top ^