New
Library
About
Online Catalog
Ask Dr. Global Change
Site Map
Links


HomeTOCPreviousNext

Changes in Tropospheric Chemical Composition

Changes in O3

Tropospheric ozone plays an important role in the atmosphere because its photolysis in the presence of water vapor is the primary source for hydroxyl radical (OH), which is responsible for the removal of many important trace gases. Moreover, it is an effective greenhouse gas (especially in the upper troposphere), is toxic to organisms, and contributes to global biogeochemistry because of its UV absorption (see Chapter 5). Ozone also causes damage to many organic materials, particularly if carbon-carbon unsaturation is present, such as in rubber products.

    Tropospheric ozone stems from two processes, downward transport from the stratosphere, and in-situ photochemical production from the oxidation of hydrocarbons or carbon monoxide (CO) in the presence of NOx. Ozone is removed from the troposphere by in-situ chemistry and by deposition at the Earth's surface. The effect of decreased stratospheric ozone on the influx of stratospheric ozone into the troposphere is unknown and is not discussed here. The general effect of increased UV-B radiation on the chemical budget of ozone is discussed below.

Fig. 6.2. Development of observed global total ozone densities and calculated averages of JO3 given as annual averages normalized to 1979 levels (from Fuglestvedt et al., 1995).

The photolysis of ozone followed by reaction of O(1D) atoms with water vapor is a major photochemical sink for O3, and the main source of OH radicals:

O3 + hn (l £330nm) ® O(1D) + O2 (R1)

O(1D) + H2O ® OH + OH (R2)

Only a small fraction of O(1D) atoms react with water vapor; most O(1D) atoms undergo relaxation to ground state O(3P) via collisions with molecular oxygen or nitrogen:

O(1D) + M(N2, O2) ® O(3P) + M (R3)

Ground state O(3P) atoms react mainly with molecular oxygen forming ozone again. The reaction of O(1D) atoms with water vapor, R2, is the rate-determining step in the photochemical loss of tropospheric ozone through the photolysis of O3 itself. Additional losses of O3 occur through reaction with HO2 radicals via R4 and (to a lesser extent) through reaction of OH radicals with O3 via R5:

HO2 + O3 ® OH + 2O2 (R4)

OH + O3® HO2+ O2 (R5)

The photochemical loss rate of ozone is therefore approximately:

LO3 = [O3]{ F JO3 + k4[HO2] + k5[OH]}

where F is the fraction of O(1D) that reacts with water vapor (typically a few percent).

The production of ozone in the troposphere occurs through the conversion of NO to NO2 by peroxy radicals (HO2and RO2), followed by photolysis of NO2:

RO2 + NO ® RO + NO2 (R6)

HO2 + NO ® OH + NO2 (R7)

NO2 + hn (l £ 420 nm) ® NO + O(3P) (R8)

O(3P) + O2 + M ® O3+ M (R9)

Hydrocarbons and carbon monoxide (CO) provide the peroxy radicals through their oxidation and are consumed in the process, while NOx is conserved and thus acts as catalyst in ozone formation. The conversion of NO to NO2 by peroxy radicals is the rate-limiting step, since the photolysis of NO2 (R8) and the subsequent reaction (R9) are rapid. The ozone formation rate is given approximately by:

PO3 = [NO]{ k6[RO2] + k7 [HO2]}

    The chemical budget of ozone in a given region is governed by the rate of net ozone production, PO3 - LO3. The effect of increased UV-B radiation on the chemical budget of ozone, i.e. the change of PO3 - LO3, is complex. An increase in UV-B radiation will increase JO3 and hence increase the rate of ozone destruction LO3. Considering the production of ozone, JNO2 is quite insensitive to increasing UV-B radiation (see Table 6.1), but the increase of UV-B radiation will lead to increased concentrations of hydroxyl radicals and thus to increased concentrations of HO2 and RO2 radicals, which will enhance the production of ozone if NOx is available. An increase in UV-B radiation will enhance both destruction and production process for tropospheric ozone. Since PO3 is dependent on the NOx abundance and LO3 is, to first approximation, insensitive to NOx, the sign and magnitude of trends in tropospheric O3 (from UV-B radiation increases) are dependent on the NOx concentration.

    Several model studies have addressed the question of how tropospheric ozone may respond to UV-B increases. Liu and Trainer (1988) used a zero-dimensional model (chemistry only, no transport) to show that the sign (increase or decrease) of tropospheric ozone changes depends sensitively on the local NOx concentrations. At the high NOx levels typical of urban conditions, increases in UV-B are expected to result in higher levels of surface ozone, so that exceedances of air quality standards may be more frequent (De Leeuw and Van Rheineck Leyssius, 1991), or alternatively more stringent regulations may be required to meet current standards (Gery et al., 1987). Qualitatively similar results were found using one-dimensional models (chemistry plus vertical transport) by Thompson et al. (1990) and by Ma (1996), but with quantitative results shown to depend also on other chemical constituents (e.g. hydrocarbons, CO). The study by Ma (1996) further suggests that in the upper troposphere the changes in ozone are probably rather small, due to the usually very low concentrations of water vapor which therefore reduce the importance of reactions R1-R3 (and thus of JO3). Increases in JO3 become more important in the middle and lower troposphere, where higher levels of water vapor are found, and especially in the continental boundary layer if significant NOx concentrations are present. The results of these studies are sensitive to whether the calculations are carried out over polluted areas (e.g., the troposphere above industrialized regions) or more pristine regions (e.g., over remote ocean regions). Fuglestvedt et al. (1994; 1995) used a two-dimensional model of the global troposphere (i.e., allowing for variations with both altitude and latitude). Calculations suggest that at high latitudes of the Southern Hemisphere there is a large reduction in tropospheric O3 in response to stratospheric ozone depletion in the spring, but the maximum effect, a 16 % reduction in tropospheric O3 at 2.5 km altitude over 60-80° S, is delayed relative to the October minimum in stratospheric ozone by about 2 months, to early summer (December) when solar UV-B actinic fluxes are highest. At middle and high latitudes of the Northern Hemisphere, increased UV-B radiation is predicted to increase tropospheric O3 in mid-spring (April) because of high levels of NOx and other ozone precursors (CO and hydrocarbons), but some reductions were also noted for high northern latitudes in late spring (May and June).

    Very little observational evidence exists in support of these model calculations, due mostly to the sparcity of tropospheric ozone measurements (especially vertical profiles) and to the difficulty of attributing any observed trends to UV-B increases rather than any other of the numerous factors that also contribute to the tropospheric ozone budget, e.g., trends in emissions of the precursors NOx and hydrocarbons. The most relevant observations are those obtained in polar regions, because the changes in stratospheric ozone (and therefore tropospheric UV-B levels) are larger than those found for mid-latitudes and the tropics. Long-term observations at the South Pole indicate that some enhanced net destruction of surface O3 may already have occurred in association with the large stratospheric ozone losses in that region (Schnell et al., 1991). More recently, Taalas et al. (1997) have studied vertical profiles obtained from balloon-borne ozone sondes over 1988-1994 at Marambio, Antarctica (64° S), and at Sodankyla, Finland (67° N). During months with substantial stratospheric ozone loss, ozone levels were found to be reduced also in the upper troposphere (6-8 km), averaging 12.8% reductions in the Antarctic and 10.0% in the Arctic. It was proposed that both reductions in the ozone flux from the stratosphere to the troposphere and the changes in the photochemistry of the upper troposphere are possible reasons for this correlation between upper tropospheric and lower stratospheric ozone concentrations. In the middle (3-5 km) and lower (0-2 km) troposphere, tropospheric ozone deviations were found to be negative over Antarctica but positive in the Arctic, in qualitative agreement with the model-derived expectations given the more polluted conditions present in the high latitudes of the Northern Hemisphere (relative to Antarctica).

    Tropospheric ozone concentrations are sensitive not only to UV radiation, but also to many other factors including concentrations of nitrogen oxides, water vapor, carbon monoxide, methane and non-methane hydrocarbons. Long-term changes in the emission sources of these compounds are known to be occurring, and their effects on tropospheric ozone, relative to the effects of increased UV radiation levels, are at present highly uncertain.

Changes in HOx

The odd-hydrogen radicals (HOx = OH + HO2) play a key role in tropospheric chemistry, as already illustrated in reactions R1-R9. Important sources of these radicals include the photolysis of ozone, hydrogen peroxide (H2O2), formaldehyde (HCHO), and several other inorganic and organic species. Table 6.1 shows that many of the corresponding photolysis rate coefficients are sensitive to the amount of stratospheric ozone.

    Model calculations consistently show that increased UV-B actinic fluxes (associated with stratospheric ozone reductions) yield higher tropospheric concentrations of the HOx radicals (Liu and Trainer, 1988; Thompson et al., 1990; Madronich and Granier, 1992; Fuglestvedt et al., 1994; 1995; Ma, 1996). For example, Fuglestvedt et al. (1995) calculated that the increases in UV-B radiation over 1979-1993 have led to an increase in OH on the global scale of about 8%, with the largest fractional increases (about 40 %) found for high southern latitudes in October. The modeling study by Ma (1996) investigated the potential changes in the concentrations of OH, HO2, and CH3O2 in the troposphere for a 10 percent reduction of stratospheric ozone. The concentration of OH is predicted to increase by 8 to 9 percent in the lower troposphere over global scales, due mainly to the increase in JO3. The concentrations of HO2 and CH3O2 are predicted to increase by about 4 to 6 percent respectively, due to the increased oxidation rate of CO and hydrocarbons by OH.

    There is little direct observational evidence in support of the model-predicted increases in HOx. Several studies have attempted to estimate recent global trends in OH concentrations based on measurements of methyl chloroform concentrations and the requirement that its sources (which are fairly well known) be balanced by its atmospheric sink, which is believed to be primarily reaction with OH. The study by Prinn et al. (1995), which updates the earlier study by Prinn et al. (1992), found no statistically significant trend on OH over 1978-1990. More recently, a re-analysis of the methyl chloroform data (Krol et al., 1998) suggests a slight positive OH trend, ca. 0.4-0.5 % yr-1 over 1978-1993 at the five monitoring stations considered. However, such OH trends can result from many different atmospheric changes, including from increases in O3 and H2O2 (precursors to OH) especially in the northern hemisphere as a result of changing emissions of NOx, CO, and hydrocarbons; from increases in water vapor concentrations above tropical oceans (possibly associated with climate warming); and from changes in temporal and geographical patterns of other human activities (e.g., biomass burning). It is at present difficult to assess how much of any observed trends in global OH concentrations can be attributed to UV-B increases stemming from stratospheric ozone depletion.

Changes in CH4 and CO

The atmospheric concentrations of methane (CH4) and carbon monoxide (CO) are determined by the balance between emission sources (many of which are related to human activities) and atmospheric removal, mostly by OH radicals. For CO an important additional source is the atmospheric oxidation of CH4 and other hydrocarbons. Short-term and long-term variations in atmospheric CH4 and CO concentrations stem from the superposition of these sources and sinks, and, when local concentration measurements are considered, from transport processes that re-distribute the gases to regions far from the original emission sources. The effect of possible tropospheric OH increases (from stratospheric ozone depletion) on the average concentrations of CO and CH4 is estimated in many cases to be smaller than, or comparable to, the effects of changes in emission sources, and is therefore difficult to detect by inspection of the available long-term CO and CH4 monitoring data records. The estimated effects of the tropospheric UV-B increases are not negligible (at least based on model estimates), and must therefore be given some consideration as contributing to the overall trends (Liu and Trainer, 1988; Thompson et al., 1990; Madronich and Granier, 1992, 1994; Bekki et al., 1994; Fuglesvedt et al., 1994, 1995).

    Analyses of polar ice cores (Etheridge et al., 1998) indicate that pre-industrial atmospheric CH4 concentrations were about 650 ± 40 ppb. Atmospheric methane concentrations started to increase about a hundred years ago, probably because of human activities linked to rising population. The rate of increase has not been constant. In the late 1970s the rate of increase was approximately 16 ppb/yr, while in the later part of the 1980s the increase was about 9 ppb/yr. To the extent that increased UV-B radiation is expected to lead to higher levels of OH, a correspondingly faster loss of CH4 is expected. Figure 6.3 shows the change of CH4 concentrations calculated with a two-dimensional model (Fuglesvedt et al., 1995), that are expected to result from increased UV-B radiation alone. The changes in the global annual growth rate are also shown. The figure shows that the increased UV-B radiation may have reduced the global methane concentration by as much as 30 ppb from 1980 to 1993, and thus may have contributed partly to the overall trend declines observed since 1980.

Fig. 6.3. Predicted change in the global level and growth rate of CH4 due to UV increases (from Fuglestvedt et al., 1995).

CO concentration measurements suggest an increase in the CO abundance of 1 percent per year over the past 40 years in the Northern Hemisphere (Zander et al., 1989), while no significant trend was found in the Southern Hemisphere (Brunke et al., 1990). A large decrease in the CO concentrations was observed between 71°N and 41°S during June 1990 - June 1993, with a decrease of 15 to 18 ppbv at most stations located north of 25°N, and of 8 to 12 ppbv between 41°S and 20°N (Novelli et al., 1994). To the extent that OH increases are expected in association with stratospheric ozone depletion, global decreases in CO concentrations are expected. Granier et al. (1996) used a global three-dimensional chemical transport model to investigate possible causes for the observed CO decrease, and concluded that changes in total ozone abundance may be responsible for a global decrease in CO of about 3.5 ppbv and 1.7 ppbv in the Northern and Southern Hemispheres, respectively, accounting for about 20% of the observed CO decreases. The large remaining decreases were attributed to decreased industrial and transportation CO sources, to decreased biomass burning, and to a global decrease in surface air temperatures for several years following the eruption of Mt. Pinatubo.

    An additional study, though not directly related to stratospheric ozone depletion, illustrates the sensitivity of tropospheric CH4 and CO to stratospheric transmission of solar UV-B wavelengths. Observations show sharp increases in the growth rates of CH4 and CO in the tropics and middle southern latitudes for several months following the eruption of Mt. Pinatubo on June 15, 1991. This volcanic eruption emitted approximately 20 Mtons of SO2 and 3-5 km3 of ash into the upper troposphere and lower stratosphere. Calculations using a radiative transfer model showed that the tropospheric UV-B actinic flux in the tropics was attenuated by about 12% immediately after the eruption due to direct absorption by SO2, and was perturbed for up to 1 year after the eruption due to scattering by sulfate aerosols (Dlugokencky et al., 1996). This study suggested that the decreased UV-B flux caused a decrease of OH concentrations and therefore lead to the observed anomalously large growth rates for CH4 and CO during late 1991 and early 1992. Methyl chloroform, whose atmospheric removal is also initiated by OH, showed a small positive anomaly in Cape Grim, Australia, consistent with observations for CH4 at comparable latitudes. The increased growth rates were short-lived, as CH4 and CO growth rates showed strong decreases during late 1992 and 1993 (Dlugokencky et al., 1994; Novelli et al., 1994). Bekki et al. (1994) suggested that for several years after the Mt. Pinatubo eruption, faster rates of removal of CO and CH4 may have resulted lower stratospheric ozone (and therefore higher tropospheric UV and OH); however measurements of the isotopic composition of CH4 are more consistent with decreased emissions, possibly from biomass burning (Lowe et al., 1997).

    In summary, the potential perturbations to CO and CH4 concentrations from increased tropospheric UV-B levels are only one of many factors contributing to the observed trends and variations. Changes in emission sources are well-recognized, although not fully quantified. Changes in OH concentrations have also likely occurred due to many factors, in addition to UV-B changes, including trends in precursor species such as hydrocarbons and NOx (and therefore tropospheric ozone), changes in emissions of biogenic (natural) hydrocarbons from changes in land use, climatic changes in temperature and water vapor, and possibly even other changes in actinic fluxes due to trends and variabilities of clouds and aerosols. The net response of atmospheric CO and CH4 concentration is, to first order, a result of the superposition of multiple driving factors, but it remains exceedingly difficult to separate and quantify the importance of each of these contributions

Changes in H2O2

Hydrogen peroxide (H2O2) is one the principal oxidants in the troposphere, and plays an important role in the aqueous phase oxidation of SO2 to SO4=. Model calculations (e.g., Fuglesvedt et al., 1994, 1995) suggest that tropospheric H2O2 concentrations should increase in response to enhanced tropospheric UV-B actinic fluxes, in parallel to the expected increases in HOx (especially HO2). The calculated H2O2 increases are greatest in low NOx regions, where the formation of peroxides is the predominant fate of HOx radicals. Measurements at Eurocore (central Greenland, 72°N) showed a 50% increase of the H2O2 concentration in the firn/ice during last 200 years, with most of the increase having occurred between 1960 and 1988 (Sigg and Neftel, 1991). Recent studies at Summit, Greenland (72°N, in 1995) confirmed the H2O2 increases found in ice cores, and showed a further increase of the H2O2 concentration since 1988, leading to an overall increase of 60±12% during the last 150 years (Anklin and Bales, 1997). Photochemical model calculations for Summit (Neftel et al., 1995) indicate that ozone depletion over 1980-1990 (from 395 to 360 Dobson Units) gives an atmospheric H2O2 increases of about 7% for summer, which could account for about one third of the increase observed over that time period. Anklin and Bales proposed that the recent increase could be partly due to increasing UV-B radiation caused by the stratospheric ozone depletion and a combination of changes in tropospheric chemistry from other causes such as increased emissions of hydrocarbons and CO.

HomeTOCPreviousNext
U.S. Global Change Research Information Office, Suite 250, 1717 Pennsylvania Ave, NW, Washington, DC 20006. Tel: +1 202 223 6262. Fax: +1 202 223 3065. Email: . Web: www.gcrio.org. Webmaster: .
U.S. Climate Change Technology Program Intranet Logo and link to Home