From Section 11.3 it is apparent that the atmospheric concentrations of CH4 can be affected by changes in the oxidizing capacity of the atmosphere, particularly changes in global or regional OH concentrations. Earlier theories on increases of CH4 were based on the hypothesis that as more carbon monoxide is put into the atmosphere it will deplete OH and hence cause CH4 concentrations to rise, even if CH4 emission rates do not increase (Chameides et al., 1977; Sze, 1977; Thompson and Cicerone, 1986). More detailed studies, including the atmospheric changes that also increase the production of OH, have suggested that long-term changes of OH are not as large as may be expected by considering the increases of carbon monoxide alone. Hence, the concentrations of CH4 currently observed, which are more than double pre-industrial levels, are accepted to be driven primarily by increased emissions of CH4 itself (Khalil and Rasmussen, 1985). If the pre-industrial OH concentration is taken as a standard, the present values are ~5-10% below this level, and during the ice ages OH would have been ~15-30% above this level (Lu and Khalil, 1991; Pinto and Khalil, 1991). Yet the question remains whether there is an important effect on the concentrations of CH4 over time and accumulation in the atmosphere caused by changes of OH, especially in recent decades. For instance, one proposed theory of the recent slowdown of the CH4 trend is that OH is increasing as a feedback from stratospheric ozone depletion, which stimulates its production by increasing ultraviolet (UV) radiation in the troposphere (Madronich and Granier, 1992).
It is not yet possible to measure OH on a global scale; hence, it is not possible to know with certainty the global average of OH. Models are used to calculate global OH concentrations. In constructing such models it becomes apparent that the atmospheric chemistry and physics are very complicated and include many feedbacks. As the completeness of the models is difficult to establish, predicted trends are not reliable, as already discussed. An empirical method has been used to determine global or hemispherical mean OH concentrations and long-term trends. In this method a tracer with known atmospheric concentrations, sources and OH reactivity is used to determine average OH levels. The tracer of choice is the solvent 1-1-1 trichloroethane, commonly known as methyl chloroform (CH3CCl3), because it has a number of characteristics that make it a reliable indicator of OH such as: (i) accurate global measurements over nearly 30 years; (ii) well-known emissions based on industrial production; (iii) removal mostly by reacting with OH; and (iv) a relatively short lifetime of ~5 years that reduces the uncertainties in the OH estimates (Lovelock, 1977).
A global mass balance for any atmospheric constituent can be written by equating the rate of change in concentration with time (dC/dt), with the emissions (S) less than the losses due to reactions. For gases such as CH3CCl3 the losses due to reactions with OH are proportional to the concentrations in the atmosphere or k[OH][CH3CCl3] where k is the reaction rate coefficient. Then, on a global scale we can calculate OH as equal to (S ~d[CH3CCl3]/dt)/(k[CH3CCl3]) where S is determined from industrial production records; the usages of CH3CCl3, d[CH3CCl3]/ dt and [CH3CCl3] are determined from direct atmospheric measurements; and k is determined from laboratory studies.
Actual calculations use more complex spatially and temporally resolved models, a more detailed mass balance taking into account other sinks, potential effects of transport processes and more sophisticated techniques to obtain a de-convolution by which OH is calculated. The results of two such calculations are shown in Fig. 11.2. The mean levels are about 9.4 x 105 and 1.2 x 106 molecules/cm3 in Prinn et al. (2001)
-Q— Prinnefa/. (2001) -•-- Butenhoff and Khalil (2005)
Fig. 11.2. Calculated concentrations and trends of OH in the atmosphere using methyl chloroform as a tracer. (From Prinn et al., 2001 and Butenhoff and Khalil, 2005, personal communication.)
and Butenhoff and Khalil (2005, personal communication) respectively, which are consistent with atmospheric chemistry models mentioned earlier. In Fig. 11.2 the results of Prinn et al. (2001) are adjusted by a constant multiplicative factor of 1.25. The additional benefit of these calculations is the ability to detect trends in OH concentration over time. According to these calculations there is a suggestion of small increases during 1979-1990 of ~0.7% ± 0.8% per year and 0.7% ± 1% per year that are not statistically significant, and decreases in 1990-2000 of 1.8% ± 1.1% per year and 2.4% ± 1.2% per year respectively for the studies reported by Prinn et al. (2001) and Butenhoff and Khalil (in preparation). Note that these two studies use different data-sets and models.
In general the year-to-year changes are variable, but this cannot be said to represent the changes in the real atmosphere as the variabilities introduced by the model and measurements are significant. The calculations are also uncertain for the recent years since the inventory of CHgCClg has become increasingly unreliable.
Was this article helpful?