Atmospheric Observations

Atmospheric concentrations of methane have been measured systematically for nearly 20 years (Rasmussen and Khalil, 1981; Khalil and Rasmussen, 1983, 1990a; Blake and Rowland, 1988; Steele et al., 1992; Khalil et al., 1993; Dlugokencky et al, 1994). There is an additional record from many independent measurements spanning another 15 years or so back to the early 1960s (Khalil et al., 1989). For earlier times, the ice core record is the only source of information. It extends back over 150,000 years, but for our interest here only the last 1000 years or so are important (Rasmussen and Khalil, 1984; Chappellaz et al., 1990; Etheridge et al., 1992). This record is summarized in Figures 1 and 2. The first panel of these figures shows the time history of methane over the last 1000 years, 100 years, and the most recent decades, and the second panel shows the trends of methane over the same periods.

These data establish two important results: First, that methane concentrations have increased by a factor of about 2.5 over the last 100 to 200 years. And second, that the rates of increase reached peak values during the 1980s, but have been declining since. The rapid increases observed in the 1980s suggested that methane could contribute significantly to global warming in the future. These observations were the compelling reason that drove much of the research on a systematic study of the cycle of atmospheric methane. Later we will return to why these trends are changing.

The most recent decades of data shown in Figure 1 contain many features that reflect the production and destruction processes of methane. There are two salient patterns: The seasonal cycles and the latitudinal concentration gradient. These features are shown more clearly in Figures 3 and 4, respectively.

The data shown are taken at Earth's surface at locations that are far from local sources. As such these concentrations and their patterns represent the large-scale distribution of methane in the atmosphere. In the vertical, up to the tropopause, methane mixing ratios remain nearly the same, implying that the actual concentration of methane in molecules/cm3 falls off at the same rate as the density of air or approximately 12.5%/km. At higher altitudes, in the stratosphere, the concentrations (molecules/cm3) fall at an approximate rate of 5%/km as shown in Figure 5.

These observations provide qualitative evidence for several important conclusions regarding the global cycle of methane. Clearly, the methane cycle is out of balance when considered over decadal time scales as evidenced by the generally increasing trends. Moreover, this imbalance arose over the last 100 to 200 years since, before that time, the concentration was unchanging, at least over the previous 1000 years. This would mean that in recent times, more methane is put into the atmosphere than is being removed annually. Second, the latitudinal gradient suggests that the production of methane is considerably higher in the Northern Hemisphere compared with the Southern Hemisphere, if we assume that the destruction processes are similar in

these factors into consideration for each point in the atmosphere and for each unit of time. For our purposes here we will deal with a simplified concept whereby the entire atmosphere is regarded as a single reservoir into which we put methane from its sources and from which methane is removed by a series of chemical and physical processes. Moreover, we take each loss process to be proportional to the amount of methane present at any given time. Considerations of transport of methane are no longer explicitly needed since all methane stays within the atmosphere, and what is moved from one part by the winds, goes to another location, still within the global box, thus not affecting the amount of methane in the global atmosphere. This model can be stated as:

dt t

Here C is the amount of methane in Tg (1 Tg = 1012 g), S is the emissions from all sources in Tg/yr, and t is the effective atmospheric lifetime in years, i is a composite lifetime due to all the processes that remove methane from the atmosphere, or 1/t = l/ii + l/i2 + • • • + 1 /t;V where xx, i2, ..., iN represent the lifetimes due to each of N processes. For direct comparisons with measurements C can be converted to ppbv and hence S is expressed in ppbv/yr. The conversion factor is 1 Tg % 2.8 ppbv in the global atmosphere.

In the mass-balance equation, we know the solution (C) based on the atmospheric measurements, so our task is to find the two remaining parts of the budget, namely the emissions (S) or the lifetime (i). For the case of methane, the lifetime is calculated independently, so that the mass-balance equation is essentially a tool for finding the sources that have combined emissions that are consistent with the measured concentrations and calculated loss rates [Eq. (1)]. How Eq. (1) is used will be discussed next; then we will show the recent budgets consistent with the known constraints and the mass balance expressed by Eq. (1).

Since there are many sources, the mass-balance equation by itself is insufficient to constrain how much methane comes from each source. Nonetheless, we can use it to estimate the total annual emissions of methane. Based on a calculated lifetime of 10 years (i), to be discussed later, and a global burden of 4800 Tg (C) obtained from global measurements (Figs. 1 to 5), and a current rate of change of about 20 Tg/yr (idC/dt; Fig. 1), we see from Eq. (1) that the total worldwide emissions from all sources should be about 500 Tg/yr. This is a useful benchmark.

There are a number of ways to improve on Eq. (1) that will allow us to reduce the uncertainties in the estimate of emissions from individual sources or combinations of sources. We will discuss three approaches here that have been useful in developing better global budgets. One method is to consider the long time series, such as the ice core data over several centuries, and apply Eq. (1) to two different time periods. This method uses the observed changes over long time periods to determine the ratio of anthropogenic to natural emissions. If we assume that several hundred years ago the concentration of methane in the atmosphere was determined entirely by natural processes, we can then estimate the emissions that would be required to satisfy


Eq. (1). This is particularly simple since at that time there are no significant trends, so S = C/x. Based on the ice core data for C, we can estimate the emissions to be 1700Tg/yr/10yr= 170Tg/yr. We have already done the calculation for recent times suggesting present emissions of 500Tg/yr. This would imply that there are new sources, presumably due to human activities, amounting to some 330Tg/yr (Khalil and Rasmussen, 1990b). We have assumed that the lifetime of methane is the same now as it was a century or more ago. There is reason to believe that this is a good approximation, but the matter is open to question.

Another approach is to consider the latitudinal distribution of the various sources, determined by independent data or measurements. Then, a more detailed version of the mass-balance model, which takes into account the budget of methane over small regions of Earth's surface, can be used to determine whether the estimated rate of emissions from the sources is compatible with the measured atmospheric concentrations within each location. This method uses the latitudinal distribution to constrain the strength of the sources (Fung et al., 1991; Brown, 1993; Hein et al., 1997). For instance, we can rule out the oceans as the major source because that would require a more even distribution of methane across the hemispheres than is seen in Figure 4.

A recent approach at constraining the estimates of emissions uses carbon isotopes in methane. Normal measurements of methane cannot distinguish between the molecules of methane that come from one source or another, so only the total amount or concentration C is measured. It has become possible to measure the methane with different isotopes of carbon—specifically 12CH4, 13CH4, and 14CH4 (Tyler, 1986; Stevens and Engelkemeir, 1988; Wahlen et al., 1989; Quay et al., 1991, Lassey et al., 1993). In this case, we can get more information on the global sources (and sinks, or loss processes) of methane since we can now have three equations similar to Eq. (1), one for each isotope. We now have to independently balance three types of methane (12CH4, 13CH4, and 14CH4) instead of just the sum of all types, using the same sources. There are fewer combinations of sources and emission rates that would balance all types than there are for just the total methane. Recently, stable isotopes of hydrogen in methane (12CH3D) have also been measured (Bergamaschi and Harris, 1995), which would add a fourth type of methane to constrain the sources. This work requires a knowledge of not only the isotopic combination of methane in the atmosphere, but also of the amounts of each type emitted by the sources, and the atmospheric lifetimes of each type of methane. Isotopic measurements hold considerable promise for reducing the uncertainties in the budget of methane.

0 0

Post a comment