2.3.2. Discrete Sample Measurements of Methane

During 1996-1997 the determination of the global distribution of atmospheric methane continued from 49 sampling sites of the Carbon Cycle Group's cooperative air sampling network. Sampling was started at four new sites during this 2-year period. Provisional annual mean values for 1996-1997 are given in Table 2.7.

In mid-1997 routine analysis of discrete air samples began on a new analysis system (MAGICC). For the most part, the analytical techniques remained the same. Methane is analyzed by gas chromatography with flame ionization detection using a HP 6890 GC, nearly identical to the ones used at MLO and BRW for in situ measurements. A two-column separation scheme is used to isolate CH₄ from other species in each air sample. Methane mole fractions are calculated by ratioing the peak response (area) for each sample to the average peak response of bracketing aliquots of standard gas, then multiplying this ratio by the CH₄ mole fraction assigned to the standard gas tank. Sample air is extracted from the flask and flushed through the sample loop using a metal bellows pump. When using the previous Carle 7 system, CH₄ and CO sample loops were flushed in series; now the analytical systems are set up in parallel and flushed sequentially starting with N₂O/SF₆, then CO/H₂, then CH₄, and finally CO₂.

TABLE 2.7. Provisional 1996 and 1997 Annual Mean CH₄ Mole Fractions From the Air Sampling Network

Square brackets indicate insufficient data to calculate the annual mean.

The trend of methane is an important constraint on methane's global budget, and it impacts environmental policy. Methane has been increasing in the Earth's atmosphere since the industrial revolution, and it is clear that the increase is due to anthropogenic activities. The globally-averaged rate of increase has varied since the early-1980s, but, over the past decade, it has slowed by more than a factor of two [Dlugokencky et al, 1994]. The overall decline in growth rate has not been fully explained. Possibilities that could partially explain the decreased growth rate are increased OH radical concentrations due to long-term decreasing trends in stratospheric O₃, and stabilization of emissions from anthropogenic sources such as rice agriculture and cattle. A recent analysis of the CMDL global CH₄ averages shown in Figure 2.13 [Dlugokencky et al., 1998] suggests that the decreasing growth rate of atmospheric methane can be explained as a chemical system approaching steady state where emissions and [OH] have remained about constant from 1983-1996.

Fig. 2.13. (a) Globally averaged CH₄ mole fractions (symbols; where ppb is used to abbreviate nmol mol^-1). The solid line is a deseasonalized trend fitted to the global averages. The dashed line is a fit of Equation (3) to the global averages calculated using CH₄(ss) = 1779 nmol mol^-1, CH₄(t = t₀) = 1615 nmol mol^-1, and t = 10 yr. (b) Instantaneous, smoothed growth rate for globally averaged atmospheric CH₄. The curve is calculated as the derivative of the solid curve in (a). Symbols are annual increases determined from the trend line above.

The change in the global burden of CH₄ is given by:

d[CH₄]/dt = Q - [CH₄]/t (1)

where [CH₄] is the global CH₄ burden (calculated directly from the CMDL globally-averaged mole fractions), Q is the sum of all emissions, and t is the total atmospheric CH₄ lifetime. The term [CH₄]/t is equal to the photo-chemical sink. Equation (1) can be rearranged to give:

Q = d[CH₄]/dt + [CH₄]/t (2)

The total global source, Q, and the atmospheric sink determined from our measurements are plotted in Figure 2.14a. The trend was calculated for each year from the deseasonalized curve fitted to the CH₄ global means, and the sink was calculated from the globally averaged burden for each year and constant lifetime, t = 8.9 years [Prinn et al., 1995]. Contributions to the sink from soil processes are small and have been ignored. Since [OH] is assumed constant, interannual variability in OH is accounted for in the source term. No significant long-term trend in [OH] has been detected globally [Prinn et al., 1995]. Therefore, Q is really a pseudo-source since it includes deviations from the mean atmospheric lifetime. We converted from ppb in the marine boundary layer to Tg (where 1 Tg = 10¹² g) with the global conversion factor, 1 ppb = 2.767 Tg [Fung et al., 1991]. There are significant interannual variations in Q, but a straight line fitted to Q gives a slope of 0.3 ± 0.6 Tg yr^-1 for the source suggesting that over the period of our measurements a significant trend in the source was not present. The increase in the CH₄ sink is due to the increasing CH₄ burden. Therefore, it appears the long-term decrease in CH₄ growth rate is an approach to steady state where total source strength and [OH] are constant.

Fig. 2.14. (a) Global methane source (circles) and sink (triangles) calculated using the mass-balance equation from the global burden and annual increase determined by CMDL and CH₄ lifetime determined by Prinn et al. [1995]. (b) Differences between annually averaged CH₄ mole fractions in the latitude zones 30-90ºN and 30-90ºS. The average difference is 123.6 nmol mol^-1, and the trend in the difference is -0.6 ± 0.2 nmol mol^-1 yr^-1.

Alternately, consider CH₄ at steady state, where d[CH₄/dt = 0, and

Q = [CH₄]_ss/t , (the subscript "ss" implies steady state). The approach to steady state for this system, with a zeroth-order source and pseudo first-order loss of CH₄ is given by:

[CH₄](t) = [CH₄]_ss - ([CH₄]_ss - [CH₄]₀) exp(-t/t) (3)

where the non-zero burden at the time our measurements started is accounted for. The terms [CH₄]_ss, ([CH₄]_ss-[CH₄]₀), and t are determined from a nonlinear curve-fitting routine. Equation (3) was fitted to our deseasonalized global averages. A steady state CH₄ mole fraction of 1779 ppb and t = 10.0 yr were determined, and a curve calculated using these coefficients is plotted in Figure 2.13a. Agreement between this simple kinetic model and the global averages is good. While sources have appeared to remain nearly constant, a small, yet unexplained, trend may exist in the difference in CH₄ burden between the high-northern and high-southern latitudes (Figure 2.14b).

Two data analysis methods (mass balance and kinetic expression) suggest that the global burden of CH₄ has been approaching equilibrium since the early 1980s. This has important implications for policy. If CH₄ sources and [OH] remain at 1996 levels, the global burden will increase to about 1780 ppb with a time constant of about 10 years and then stabilize. This result impacts scenarios of future climate and stratospheric O₃ (since CH₄ + Cl --> HCl + CH₃ is a temporary reservoir of active chlorine). We caution that quantifying small, long-term trends in specific CH₄ source strengths remains impossible, therefore, the future burden of atmospheric CH₄ cannot be predicted with certainty.

[BACK] [CONTENTS] [NEXT]