The Tropospheric Chemistry of Ozone in the Polar Regions

The Arctic troposphere (0 to ca. 8 km) plays an important role in environmental concerns for global change. It is a unique chemical reactor influenced by human activity and the Arctic ocean. It is surrounded by industrialized continents that in winter contribute gaseous and particulate pollution (Arctic haze). It is underlain by the flat Arctic ocean from which it is separated by a crack-ridden ice membrane 3 to 4 m thick. Ocean to atmosphere exchange of heat, water vapor and marine biogenic gases influence the composition of the reactor. From September 21 to December 21 to March 21, the region north of the Arctic circle goes from a completely sunlit situation to a completely dark one and then back to light. At the same time the lower troposphere is stably stratified. This hinders vertical mixing. During this light period, surface temperature reaches as low as −40 °C.

It is becoming increasingly apparent that the polar regions play an important role in the atmospheric cycle of ozone. What is perhaps less well recognized is that this is true not only in the stratosphere (above 8 to 10 km) but also in the troposphere (0 to 8 km). The aim of this paper is to “set the scene” for the following discussions by describing characteristics of geography, climate and chemical environment in the Arctic and Antarctic that help us to understand polar tropospheric ozone. The use of “polar” here refers to those areas poleward of 40° latitude. This represents approximately 36% of the total surface area of the globe. 40° is chosen to include areas of the northern hemisphere that are frequented by arctic air masses during winter as indicated by the mean position of the arctic front in January (Figure 1).

Ozone in the troposphere, though it represents only a relatively small fraction of the total column amount, is now recognized for its important role in the chemical and radiative balance of the atmosphere. Because of the lower tropopauses found in the polar regions, the contribution of the tropospheric component to the total column ranges from about 3–15%. In the north polar region the contribution falls in a narrower range of about 5–8%. Over Antarctica, during the spring when stratospheric ozone depletion is severe, the troposphere can comprise up to 15% of the total, otherwise it is about 3–5%. In the polar regions, there are several unique characteristics that play an important role in determining the distribution of ozone in the troposphere [Oltmans, 1992]. These include the extremely cold winter temperatures, long periods of darkness and sunlight, and transport regimes that may incorporate ozone precursors into polar latitudes or mix ozone from the stratosphere into the troposphere. Sources of anthropogenic ozone precursors are significant primarily in the northern hemisphere (NH).

The polar night is long and cold. Locations such as Alert at the northern edge of Ellesmere Island (82° N) do not receive any sunlight from late September until the beginning of March, and the average ambient temperature is normally below −30 °C during that time. This fact has interesting consequences for the chemistry that can occur in the ambient air at this part of the world. Most notably, the absence of sunlight effectively prevents the primary photochemical production of atomic and free radical compounds that initiate many of the important atmospheric chemical processes. Furthermore, the comparatively low temperatures generally lead to slower bimolecular reactions. This also leads to a drastically reduced water content of the air. All these factors combined imply that whenever airborne contaminants are somehow transported to the Arctic, their lifetime is considerably longer than at climatically more moderate mid-latitude regions. Combined with the absence of efficient transport routes out of the Arctic basin, it has the effect that the arctic atmosphere serves as a holding reservoir: a buildup of contaminants is expected — and observed. This picture changes at the time when the sun reappears. More active chemistry once again becomes possible through the photochemical production of reactive atomic and free radical compounds. Polar sunrise studies are specifically concerned with attempts to catch the chemical changes that take place during this transition from dark-phase to sunlight-irradiated chemistry.

Airborne dispersion of a gravitationally neutral tracer is determined by the macroscopic atmospheric motions. These atmospheric motions generally take shapes as wavelike meanders and closed circulations, in short generally referred to as eddies. The nature of the transport of air parcels is determined by the kinetic energy of these eddies as a function of their size and fluctuation period. According to Taylor’s statistical theory of plume-diffusion (particles continuously released over an infinitely long time in a fixed point [e.g., Pasquill and Smith, 1983], the energy-containing eddies with periods longer than the elapsed time (T) since the particle release determine the further plume shape. In reality, emissions take place over a finite time interval, or one is interested in the shape averaged over a finite time interval T_s. Taylor’s theory can be applied as long as T ≪ T_s, otherwise puff-diffusion theory, which takes into account the diffusion of particles relative to each other, is applicable. Puff diffusion was first addressed by Richardson in the 1920s and elaborated further by Batchelor in the 1950s by using similarity theory [Pasquill and Smith, 1983]. Eddies of the same size as a puff are the most efficient contributors to the increase of the puff-size by causing a complete shape distortion, whilst those that are significantly smaller only cause a gradual entrainment of clean air by mixing through the boundaries. Eddies much larger than a puff will cause a general displacement of the puff without dilution, viz. advection.

Nitrogen oxides (NO_x), through their control of tropospheric ozone production, play a major role in determining the global reactivity of the atmosphere. The concentration of these oxides varies by as much as a factor of 1000 between continental source regions and remote locations and fluctuates significantly with season at high latitudes. While NO_x levels appear to be rather low (<50 pptv) away from local sources, high levels of PAN, an important reservoir for NO_x, have been measured at the surface in the winter polar regions [Barrie and Bottenheim, 1991] and in the free troposphere north of 30°N [Singh, Salas and Viezee, 1986]. Furthermore, our recent global chemical transport model (GCTM) study finds PAN to be the major reactive nitrogen species in the northern latitudes [Kasibhatla et al., 1992].

In 1988 Barrie et al. [1988] drew attention to the startling drop in O₃ at polar sunrise at Alert and the concomitant increase in filterable Bromine (f-Br). As shown in the paper of Bottenheim in this workshop, the measurements indicated sharp periodic drops in the mixing ratio of O₃ with a time scale of several hours or less. On occasions the O₃ levels would be below the detection limit of the instruments used. Simultaneously, they confirmed dramatic increase in the levels of f-Br (more than 100 times the levels anticipated from sea salt aerosols) that had been measured earlier by Berg et al. [1983]. (Although one should point out that different filters had been used. In addition, one uses the term “simultaneously” loosely since the time of a single measurement is substantially longer for the f-Br measurements than for the O₃ measurements.) Many measurements since that time have confirmed the negative-correlation between the O₃ and the f-Br [e.g., Bottenheim et al., 1990; Oltman et al., 1989 etc.] and the effect has been observed at Barrow as well as Alert. Seasonal measurements of O₃ at Barrow indicate values well below the yearly average during polar sunrise. The sharp depletion seems to be correlated with stable conditions in the boundary layer with the depletion of O₃ occurring only within the first kilometer.

Low solar insolation and low precipitation rates dramatically slow down the oxidation and removal of many anthropogenic pollutants in the Arctic, including hydrocarbons and nitrogen oxides. These processes give rise to a seasonal pollutant accumulation in the Arctic often referred to as “Arctic Haze”, with a maximum in the winter-spring period. There is some evidence that the wintertime accumulation of these constituents occurs not only in the Arctic but also in other remote Northern Hemispheric locations as well. Penkett and Brice [1986] suggest that a hemispheric buildup of nitrogen oxides and hydrocarbons contributes to the observed springtime maximum in ozone which is usually attributed to stratospheric exchange. Based on modeling studies, Isaksen et al. [1985] have found that the arctic reservoir of nitrogen oxides can play a significant role in ozone production in the spring.

A wide range of halocarbon gases (those containing one or more of the halogens chlorine, fluorine, bromine and iodine) have been identified in the polar atmosphere. Their origins are various, from both anthropogenic and natural sources. Although much is known about the transformations of halocarbons in the stratosphere, and their apparent involvement in Antarctic stratospheric ozone depletion in particular, much less is know about their potential impact on tropospheric ozone chemistry. Nevertheless, the Arctic spring bromine “pulse” and negative correlation between particulate bromine and ozone is compelling evidence for halogen-ozone reactions in the polar troposphere. In the Antarctic, the progressive decline in free tropospheric ozone in austral summer has been attributed to greater UV penetration through the ozone-depleted stratosphere; a possible example of an indirect effect of halocarbons on tropospheric ozone.

This paper is a shortened version of one to be published in the refereed literature [Penkett, et al., 1993]. Its relevance to the NATO Workshop held in Nova Scotia in 1992 is associated with the particular condition of meteorology and chemistry that pertain in the Arctic and allows large concentrations of relatively reactive molecules to build up in wintertime at high latitudes of the northern hemisphere. Also, the location of the densely populated continents of North America, Europe and parts of Asia with respect to the polar areas ensures that significant fractions of the pollutants emitted in these land areas are able to pollute the polar areas to a high degree. This phenomenon was first identified as the arctic haze [S.A. Penkett, 1981], however, the co-presence of many gases at high concentrations could be of considerably more consequence to the pollution of the northern hemisphere as a whole.

Transport of pollutants from populated and industrialized continental areas affects the chemistry and radiation balance of the global troposphere [Duce et al., 1991]. A pollutant of particular interest is ozone (O₃), since its photolysis initiates the oxidizing processes in the atmosphere [Logan et al., 1981], and it is an important greenhouse gas, whose atmospheric trends are only poorly known [Watson et al., 1990]. Thus, the transport of ozone and its precursors from source regions affects the oxidizing capacity of the troposphere in the receptor areas and is important in climate change.

In June 1984 several French and German research groups joined in the STRATOZ III aircraft campaign to measure the global distribution of a number of trace gases. The first part of that campaign led across the Northern Atlantic with stops at Prestwick, Scotland; Keflavik, Iceland; Sondrestrom, Greenland, and Goose Bay, Canada, all falling in the zonal belt between roughly 55° N and 65° N latitude (see Figure 1). During the ascents from and descents to these airports vertical profiles were measured for NO by Drummond et al., [1988] and for O₃, CO, CH₄ by Marenco et al. [1989a,b]. Figure 2 presents the data coverage projected onto the altitude by longitude plane at 60° N latitude. Also measured was PAN but with a much lower sampling density [Rudolph et al., 1987]. The measurements took place between the early morning on June 4, 1984, beginning with the descent to Prestwick, and the late morning of June 6, 1984, ending with the ascent from Goose Bay, and resulted in altogether 8 vertical profiles. As indicated in Figure 2 the aircraft, while gradually climbing or descending, also covered large horizontal distances. Thus, the term “vertical profile” is somewhat of a figure of speech. The flight pattern, however, is useful in more evenly filling the altitude by longitude plane with data points.

Recently, episodic destruction of boundary-layer ozone has been observed in the arctic [Barrie et al., 1989, Barrie et al., 1988, Bottenheim et al., 1990, Mickle et al., 1889, Oltmans et al., 1983]. Those episodes, when ozone levels drop from the normal 30–40 ppb to unmeasurable (3< ppb) levels, appear to be associated with high concentrations of “filterable bromine” (bromine that can be collected on cellulose filters). While it is not clear what the exact nature of filterable bromine is, it was hypothesized by several authors [Barrie et al., 1988, Bottenheim et al 1990] that the active component of filterable bromine might be BrO radicals.

The relative remoteness of continental polar regions from natural and anthropogenic sources of various chemicals suggests that their atmosphere represents the best present-day example of the “background atmosphere” for their respective hemispheres and are, therefore, very sensitive to any natural and/or anthropogenic changes. Due to their distinct seasonal cycle with a long polar night and their very cold temperatures, polar regions can be considered as a kind of “giant natural laboratory”, in which it is probably much simpler than elsewhere to check the complex chemistry governing major biogenic cycles (S, N and C). Furthermore, a unique specificity of polar regions comes from the desposition of solid particles accumulated on polar ice sheets, which offers the possibility to reconstruct the paleoenvironnement of the Earth back to several hundred thousands years.

The principal oxidants of the earth’s atmosphere are O₃, OH radical and H₂O₂. The total atmospheric burden of these three species determines the oxidation capacity of the atmosphere [Thompson, 1992]. Each of these species reacts with different atmospheric trace constituents, and determines their atmospheric lifetimes. In particular, the OH radical reacts very fast with almost all naturally and anthropogenically emitted species; therefore, this radical has the nickname “tropospheric vacuum cleaner” [Graedel, 1978].

Atmospheric measurements of bromine compounds were stimulated by the recognition that bromine was a potential catalyst in the destruction of ozone. In the course of these studies, it became apparent that bromoform (CHBr₃) concentrations in the arctic atmosphere can be strongly seasonal; Cicerone et al. [1988] reported maxima at Point Barrow between December and February, and minima between June and August. Sturges and Barrie (1988) reported that atmospheric particulate Br peaked each year just after the Arctic dawn, with levels two orders of magnitude higher than could be explained by marine, automotive or crustal sources. Barrie et al. [1988] subsequently found that ozone depletion occurs in the arctic troposphere in springtime, and that ozone and filterable bromine concentrations were inversely correlated. These observations, together with the recent work of McConnell et al. [1992], have raised the question of the origin of arctic atmospheric bromine.

Non-methane hydrocarbons (NMHCs), particularly alkenes oxidized either by ozone, OH and NO₃ radicals can have a significant impact on the photochemistry of the remote background atmosphere [Donahue and Prinn, 1990], and are able, in turn, to influence the budget of these oxidants. Strong evidences for a residual tropospheric background hydrocarbons have been reported by several authors working in remote areas [Rudolph and Ehhalt, 1981; Bonsang and Lambert, 1985; Greenberg and Zimmerman, 1984]. Particularly, alkenes despite their high reactivity are found at significant levels from 10 pptv to 100 pptv in the southern hemisphere and even in the antarctic continent far from usual anthropogenic sources [Rudolph et al., 1989]. The existence of a marine production by outgassing of superficial seawater was first reported by Lamontagne et al. [1974] working in the Pacific Ocean, and the origin of light hydrocarbons was ascribed to local emissions from the ocean surface by photochemical degradation of dissolved organic carbon [Wilson et al., 1970]. Beside the chemical aspect linked to the photochemistry of the remote marine atmosphere, non-methane hydrocarbons are useful indicators of air/sea exchange kinetics due to their very different atmospheric lifetimes. Bonsang et al. [1991] have shown recently that it is possible to ascertain from the vertical distribution of NMHC in the atmosphere the existence of fast and non-steady state processes of exchanges between the ocean and the atmospheric boundary layer.

Phosgene (COCl₂) is produced in the earth’s atmosphere from the degradation of a variety of chlorinated compounds including tetrachloroethylene (C₂Cl₄), trichloroethylene (C₂HCl₃), chloroform (CHCl₃), methylchloroform (CH₃CCl₃), and carbon tetrachloride (CCl₄). These chlorinated compounds fall into two generic reactivity classes: 1. Tetrachloroethylene, trichloroethylene, chloroform., methylchloroform, the four reactive phosgene parent compounds (referred to here as the RPP compounds) that are primarily destroyed in the troposphere by reaction with OH; and 2. Carbon tetrachloride which is unreactive in the troposphere and is destroyed primarily by photolysis in the stratosphere. Thus the degradation of the RPP compounds lead to the production of tropospheric phosgene, while CCl₄ and to some extent also the RPP compounds leads to the production of stratospheric phosgene. Tropospheric phosgene is in turn believed to be removed from the atmosphere by rainout and ocean deposition (Singh, 1976), while stratospheric phosgene is thought to be destroyed by photolysis (Crutzen et al., 1978).

It has become clear in recent years that ice surfaces can efficiently promote chemical reactions between neutral gas phase species. The best known example involves the activation of chlorine in the presence of polar stratospheric clouds through the reaction of HCl with chlorine nitrate (ClONO₂) [e.g., Molina et al., 1987; Tolbert et al., 1987; Hanson and Ravishankara, 1991] or with HOCl [Hanson and Ravishankara, 1992; Abbatt and Molina, 1992]:

Halogen species play critical roles in the chemistry of ozone in the stratosphere. The roles played by bromine and chlorine in stratospheric ozone destruction via catalytic cycles have been studied extensively. These halogen species may also be important in the ozone chemistry of the Arctic troposphere [Barrie et al. 1988]. In particular, bromine chemistry has been implicated in the observed episodic destruction of tropospheric O₃ in the Arctic spring. Such an O₃ loss could be due to the liberation of bromine from inactive to active forms [Barrie et al., 1988; McConnell et al., 1992]. Conversion of inactive halogen species such as the hydrogen halides, chlorine nitrate, and bromine nitrate to forms that are capable of affecting the concentrations of tropospheric O₃ are likely to take place over the ice surfaces which are present during polar winter and early spring. Yet, the heterogeneous processing of bromine compounds on ice surfaces has not been studied. Therefore, we have carried out a series of laboratory measurements to investigate the reactive and non-reactive uptake of atmospherically important halogenated species such as ClONO₂, BrONO₂, HCl, and HBr onto ice layers located on the inner wall of a cylindrical flow tube at 200(±10) K.

Major tropospheric sources of HCl are the intrusion of acids [HNO₃, Martens et al., 1973, and H₂SO₄, Hitchcock et al., 1980] into sea spray aerosol [Cadle and Robbins, 1960], the direct release of HCl from coal power plants [Lightowlers and Cape, 1988], the release of Cl₂ [Hov, 1985], a photolytic source of Cl atoms which react with hydrocarbons to form again HCl, and a minor source is the photodegradation of chlorinated hydrocarbons [Behnke and Zetzsch, 1988, 1989a, and Becker and Zetzsch, 1989]. Subsequently, the slow gas-phase reaction of OH radicals with HCl is expected to form atomic Cl. These processes have been reviewed by Cicerone [1981], Warneck [1988], Friend [1990] and Keene et al. [1990] and are leading to global levels of atomic Cl around 10³cm⁻³ [Singh and Kasting, 1988], provided that the level of HCl is around 1 ppb [Vierkorn-Rudolph et al., 1984, Bächmann and Fuchs, 1987, Keene et al., 1991, although an upper limit of 0.25 ppb has been observed by Harris et al., 1990 in the marine boundary layer].

This paper reviews some of the aqueous photochemistry studied in our laboratories and elsewhere which could be extrapolated, albeit with due caution, to photochemical processes in a tropospheric situation. Attention has been limited to cases of chromophores that could possibly be found in an aerosol phase and which would have been in contact with halogens in the liquid-phase so as to have formed or be able to form halogen-chromophore combinations. Foremost amongst the chromophores having potential to be of significance in the liquid phase photochemistry of tropospheric components, we consider three: simple Iron(III) species; the iron (hydrous) oxides such as Fe₂O₃; and organic materials such as dissolved organic material (DOM), all of which are known to play a major role in the photochemical processes in natural waters.

An important role of photochemical reactions in controlling ozone concentrations in the background ‘unpolluted’ troposphere was first suggested nearly 2 decades ago [Crutzen, 1973; Chameides and Walker, 1973]. This followed the important paper by Levy [1972] where the existence of relatively large concentrations of hydrogen containing free radicals (OH and HO₂) and associated organic species derived from the breakdown of atmospheric methane, would be present in the sunlit atmosphere. The early suggestions have lead to the development of a comprehensive theory of tropospheric photochemistry which underlies our current understanding of the budget of tropospheric ozone and the oxidising properties of the atmosphere [Logan et al., 1981, Isaksen, 1988]. The central feature of this theory is the generation of a steady state concentration of OH radicals primarily through ultra violet photolysis of ozone and subsequent reaction with water of the excited O (¹D) oxygen atoms produced. The OH radicals react with many atmospheric trace gases, leading to their oxidation by a mechanism in which the OH radical can be regenerated. This oxidation process serves to control the concentration of many volatile organic compounds, including those containing sulphur, nitrogen and halogens, as well as the organic constituents such as NO, NO₂ and SO₂. The oxidation processes can lead either to ozone loss or additional ozone production, depending on the local availability of nitrogen oxides. The mechanism of these processes is now quite well known as a result of intensive studies over the last 20 years, stimulated by the problem of photochemical oxidant production in air polluted by combustion — derived NO _x and non-methane hydrocarbons. This has led to the development of sophisticated photochemical models of the production and loss of ozone in the boundary layer and in the free troposphere [Derwent and Jenkin, 1982; Crutzen and Gidel, 1983; Isaksen and Hov, 1987; Hough, 1991].

The phenomenon of ozone depletion in the springtime Arctic boundary layer is similar in several respects to the ozone hole in the Antarctic stratosphere. In both cases, heterogeneous chemistry on ice and acid-ice surfaces may be responsible for altering the partitioning of halogen-containing species between long-lived and photochemically labile forms. As a result of this repartitioning, the mixing ratios of ClO and BrO radicals become sufficiently high that self-reactions of these species can become the dominant rate-limiting steps in catalytic ozone destruction cycles. It is important, therefore, to understand the rates and mechanisms of these reactions under ambient atmospheric conditions.

PAN [CH₃C(O)O₂NO₂] has been measured ubiquitously in the troposphere, and in the stratosphere HO₂NO₂ has also been analysed. The thermal lives of both compounds as a function of pressure and temperature have been established reasonably well by laboratory studies, in particular for PAN. Other peroxynitrates including the halogenated derivatives are less important as NO_x carriers or temporary RO₂ reservoirs in the atmosphere at least at ambient temperature. Other loss processes of peroxynitrates, besides thermal decomposition, are the reaction with OH radicals and photolysis. A few experimental results also seem to support heterogeneous decomposition of RO₂NO₂ on surfaces. The relative importance of thermal decomposition rate, photolysis rate and rate of the OH reaction depends on temperature, pressure, spectral distribution of solar light intensity and the OH concentration. In principle, lower temperature and lower pressure at higher altitudes increase the importance of photolysis and OH reactions as sinks of RO₂NO₂, because thermal decomposition rates become much slower. The RO₂NO₂ are produced by the addition of NO₂ to RO₂ radicals. The RO₂-NO₂ bond energy is relatively weak and lies in the range of 85–120 kJ/mol according to recent measurements. Because these bond energies determine the thermal lifetimes, it can be estimated that at 1 bar total pressure and room temperature, the lifetime ranges from 10⁻¹ to 10⁴ s. In the present work, the thermal lifetimes of a variety of RO₂NO₂ species were studied as a function of temperature and pressure. The observed pressure dependencies of the thermal decomposition rates were fitted by calculations based on the Troe-treatment of unimolecular reactions. The data obtained within the present work allow predictions about the thermal decomposition rates of RO₂NO₂ species not yet studied.

It has recently been recognized that brominated methanes could play an important role in the chemical processes occurring in the atmosphere, since they may provide a source of Br atoms. Indeed, Br species are known to participate in the ozone catalytic destruction cycle initiated through the formation of BrO radicals: BrO radicals have been detected during Antarctic and Arctic springtime ozone depletion events in the stratosphere [Brune and Anderson, 1989; Carroll et al., 1989], and in the troposphere [Hausmann et al., 1992].

Dimethylsulfide (CH₃SCH₃, DMS) is one of the major natural organic sulfur compounds emitted to the atmosphere [Aneja and Cooper, 1989], and consequently much effort has been directed towards understanding its oxidation mechanisms particularly with the OH radical [Yin et al., 1990; Tyndall and Ravishankara, 1991]. All of the product studies on the reaction of OH with DMS have been carried out at room temperature and no information is currently available on the effects that temperatures such as those prevalent in the Arctic or Antarctic may have on the product distributions. For example, results from field measurements in the remote Southern Hemisphere show a higher yield of methanesulfonic acid (CH₃SO₃H, MSA) and a lower yield of SO₂ from DMS oxidation compared to other world areas [Berresheim et al., 1989, 1990], suggesting a change in the product distribution at low temperatures. The various mechanistic and product studies indicate that the products of the oxidation of DMS will be very much dependent on the atmospheric fate of the initially formed OH-DMS adduct and also the intermediate CH₃S radical whose loss will be controlled by reactions with either O₂ or trace species such as O₃ and NO_x. The kinetics and mechanisms of the reactions of CH₃S and O₂, O₃ and NO_x are still speculative.

It is widely hypothesized that catalytic cycles involving BrO_x species play an important role in the episodic destruction of ground-level ozone which is observed in the springtime Arctic boundary layer, although the exact mechanism for production of BrO_x radicals remains an open question [Barrie et al., 1988; Bottenheim et al., 1990; Finlayson-Pitts et al., 1990; McConnell et al., 1992]. The critical evidence linking ozone depletion with BrO_x chemistry is an observed negative correlation between ozone and filterable bromine [Bottenheim et al., 1990; Kieser et al., 1992]. In a recent field study of springtime Arctic boundary layer chemistry [Kieser et al., 1992], ozone concentrations and ethane concentrations were found to be correlated; this observation suggests that chlorine atoms (which react rapidly with ethane) may also be an important catalyst for ozone destruction under springtime Arctic conditions.

The ozone destruction observed at polar sunrise in the lower Arctic atmosphere has been suggested to be linked to catalytic reactions of BrO_x (Br, BrO) radicals [Barrie et al., 1988]. The main catalytical cycle was considered to be the following:

Over the past two decades, numerous field measurements have established that methyl iodide (CH₃I) is commonly present as a significant trace constituent of the planetary boundary layer, at concentrations in excess of 1 pptv [Lovelock et al., 1973; Rasmussen et al., 1982; Singh et al., 1983; Barnes et al., 1991; Oram and Penkett, 1991]. These field measurements have shown that the marine regions display systematically higher levels of CH₃I than the remote “clean” continental boundary layer, indicating that the oceans provide a source of CH₃I which is emitted directly into the troposphere. CH₃I has also been detected in polar air masses [Rasmussen and Khalil, 1983], in addition to the observation of springtime maxima in the level of particulate iodine in the Arctic [Sturges and Barrie, 1988; Barrie and Barrie, 1990]. The origin of these emissions is believed to be various types of macroalgae and phytoplankton, which may be found in both coastal waters and the open ocean [Lovelock, 1979; Chameides and Davis, 1980; Manley and Dastoor, 1987; Nightingale, 1991]. Although the precise mechanism is uncertain, it seems clear that these organisms produce CH₃I as part of their normal metabolic processes by “methylating” the iodide present in sea-water and, as a result, the water becomes locally super-saturated with CH₃I causing a flux from the aqueous to the gaseous phase [Singh et al., 1982; Nightingale, 1991]. Consequently, in marine locations of high biomass activity, elevated levels of CH₃I (10–40 pptv) have been observed [Rasmussen et al., 1982; Barnes et al., 1991; Oram and Penkett, 1991]. The formation of CH₃I represents, therefore, an important link in the transportation of the life-essential element iodine in the biosphere, providing a means of sea-to-land transport to balance the flux of dissolved I⁻ entering the sea from the land [Lovelock, 1979].