Several elements of biological and geological importance occur naturally with two or more stable isotopes. For example, there are two stable isotopes of carbon with atomic mass numbers 12 and 13, and a radioactive isotope with atomic mass number 14. The relative amounts of any two isotopes of the same element vary because natural differences in isotope composition of biogeochemical compounds result from processes of formation and destruction which fractionate the isotopic pools representing substrate and product [Hoefs, 1987]. Because of isotopic fractionation, biogeochemical researchers can exploit measurements of isotope ratios in compounds found in various earth, ocean, and atmospheric systems to study a wide variety of processes. The measurements can also be used as tracers to study flux rates among compartments of the biotic and abiotic environment including cycling of carbon, nitrogen, sulfur and other elements in various forms including trace gas species.

The differences in carbon isotopic composition among various compounds are usually very small (often only on the order of a few parts per thousand or less) and must be measured both accurately and precisely to be useful in biogeochemical studies. Only about 1 in 10¹² carbon atoms has mass 14. Because of this, the use of ¹⁴C is usually restricted to radioactive counting techniques which provide gross information about biogeochemical cycling (in addition to its main purpose of age dating material containing carbon). Conversely, because about 1 in 89 carbon atoms has mass 13, differences in the stable isotope ratio of ¹³C/¹²C among various compounds can be measured quite precisely, allowing detection of very small differences. The measurement results are nearly always expressed in d notation, using the following equation where ¹³C/¹²C_sample is the carbon isotope ratio in a sample and ¹³C/¹²C_standard is the carbon isotope ratio of Pee Dee Belemnite carbon, a conventional standard:

Isotopic information provides a major constraint on estimates of the global methane (CH₄) budget and its sources and sink processes. The latter include CH₄ loss from reaction with OH, Cl and O(¹D), as well as consumption by soil bacteria. Ehhalt [1974] first determined a budget relating sources and sinks of CH₄ to its atmospheric burden. In doing so, he divided sources into fossil fuel (radiocarbon dead) and biogenic (radiocarbon modern) categories and estimated that dead CH₄ sources accounted for roughly 10­20% of all atmospheric CH₄. Subsequent studies of ¹⁴CH₄ by Wahlen et al. [1989a], Manning et al. [1990], and Quay et al. [1991] have resulted in similar fossil fuel estimates.

Stevens and Rust [1982] proposed that a mass-weight-ed stable carbon isotopic balance between CH₄ sources and sink processes and the d¹³C-CH₄ value in the atmosphere would help constrain the CH₄ budget. They illustrated their contention using atmospheric data from rural Illinois and information on CH₄ sources and sinks available at the time. They used a flux-weighted average value for the d¹³C of CH₄ entering the atmosphere and allowed for a slight ¹³C enrichment in the resulting atmospheric CH₄ isotopic value from its partial chemical loss by reaction with OH, the principal CH₄ sink. (CH₄ remaining in the atmosphere is slightly enriched in ¹³C because of the OH preference for reacting with ¹²CH₄ over ¹³CH₄. This phenomenom is known as a "kinetic isotope effect," or KIE for short.) Resulting calculations by Stevens and Rust helped apportion CH₄ sources according to whether they were relatively ¹³C-heavy or ¹³C-light with respect to atmospheric CH₄. Since that beginning, additional measurements of atmospheric d¹³C-CH₄ have further constrained the CH₄ budget, provided information on seasonal cycles in CH₄ sources and sink processes, and helped elucidate recent trends in CH₄ mixing ratio [e.g., Wahlen et al., 1989a; Quay et al., 1991; Lowe et al., 1991, 1994, and 1997; Thom et al., 1993; and Tyler et al., 1993 and 1999].

One effective measurement protocol is to conduct multi-year regularly scheduled CH₄ isotope monitoring at fixed surface sites representing well-mixed background air to compare to other widely spaced sites [i.e., Quay et al., 1991; Lowe et al., 1991, 1997; and Tyler et al., 1993, 1999]. From observations, nearly all surface sites monitored have an appreciable but variable seasonal cycle in d¹³C-CH₄ with less negative values (¹³C-enriched) in summer and more negative values (¹³C-depleted) in winter. The cycle appears to be source driven, particularly in the northern hemisphere, although some of the seasonality can be explained by a seasonal increase in OH in warmer months which enriches the CH₄ isotope ratio.

Isotopic measurements of atmospheric CH₄ can be used along with model calculations to provide explanations for observed CH₄ behavior. For example, using a relatively simple one-dimensional (1D) box model with coarse resolution, the d¹³C-CH₄ seasonal cycle in the Southern Hemisphere, represented by data taken at Baring Head, New Zealand (41°S), can be explained partly by the KIE of CH₄ reaction with OH, but the amplitude is large enough to indicate that the cycle is partially source driven, possibly by biomass burning during the austral spring [Lassey et al., 1993; Lowe etal., 1997]. Similarly, the global mean d¹³C-CH₄ value of ~ -47 ppm as determined from stations regularly monitored in the Northern Hemisphere and from less frequent measurements in the Southern Hemisphere, when coupled with ¹⁴CH₄ measurements, implies that about 11% of the total CH₄ release rate is derived from biomass burning [Quay et al., 1991].

I can illustrate further how measurements and modeling of isotopic CH₄ are used to answer questions about atmospheric CH₄ and its sources and sinks processes from some of the research being done at UC Irvine. In so doing, I will also address areas where more questions have arisen, how these are being resolved, and show how the results of many researchers globally, all working toward a common understanding of the CH₄ budget, must be used to succeed.

At UC Irvine we have established time series of CH₄ measurements at Niwot Ridge, Colorado (1989­present) and Montaña de Oro, California (1995­present). The measurements at both locations include CH₄ concentration and d¹³C (approximately bi-weekly) and ¹⁴C (less frequently). Our overall precision of measurement on CH₄ mixing ratio and d¹³C from air samples is ±5 ppb and ±0.05 ppm, respectively. In Tyler et al. [1993] we reported on measurements of d¹³C-CH₄ at Niwot Ridge (40°N) for the period 1989 to 1992 (Figure 1). Those data showed a distinct difference between winter and summer values, with the d¹³C of CH₄ being approximately 0.5 ppm lighter in winter than in summer (a curve fit of the form y = a + b · t + c · sin (w · t) + d · cos (w·t) had an r² = 0.63 with seasonal amplitude 0.3 for the 4 year period) and also showed a trend in d¹³C-CH₄ of -0.10 ppm/year over the four years.

Figure 1. 13C/12C ratio in atmospheric CH4 at Niwot Ridge, Colorado from January, 1, 1989 to December 31, 1992, measured by isotope ratio mass spectrometry and expressed relative to PDB carbonate in per mil. Curve fit to data is unweighted least squares fit (1 harmonic) of form y = ­47.26 ­2.760E­4 x T + 0.0219 sin (wT) ­ 0.1269 cos (wT) with T in days relative to January 1, 1989 and w = 2p/365. Slope is equal to ­0.10 ppm/year over the four years. Error bars are one standard deviation from the average of multiple samples. [Adapted from Tyler et al., 1993].

We also maintain intercalibration programs of isotopic measurements of CH₄ samples with the National Institute of Water and Atmospheric Research, New Zealand (NIWA) and of CH₄ mixing ratios in air with the National Oceanic and Atmospheric Administration, Boulder, Colorado (NOAA). Through a sample exchange program with NIWA, a subset of the air samples collected at Montaña de Oro and Baring Head is measured by both laboratories while still in its original canisters. The obvious advantage to intercalibrating between the two isotope data sets is that each research group can incorporate the others' published data for other sites and dates into its own global calculations. Measurements reported by Lowe et al. [1991, 1994, and 1997] have consistently shown a seasonality in CH₄ mixing ratio and d¹³C of CH₄ from Baring Head for the period 1988-1996. The seasonal amplitude variation in d¹³C of CH₄ has a range of 0.6 ppm peak to peak while the period from mid-1991 to end of 1992 saw d¹³C drop by ~ 0.2 ppm.

We have used a 2D tropospheric photochemical model developed at UC Irvine to make model calculations of atmospheric CH₄ and its sources and sinks. The model has been described in detail in Gupta [1996] and Gupta et al. [1998]. In brief, it has a horizontal resolution of 10 ° and extends from the surface to 24.5 km with a vertical resolution of 0.5 km. Transport coefficients (advective and diffusive) are derived from the Geophysical Fluid Dynamics Laboratory general circulation model [Plumb and Mahlman, 1987]. A reasonably complete gas phase photochemical scheme containing only CH₄ as a parent hydrocarbon with parameterized tropospheric heterogeneous losses is included. All species including CH₄ and its isotopic species are calculated with the time dependent photochemical scheme up to 16.5 km. Above 16.5 km, hydroxyl and chlorine radical concentrations are prescribed on a monthly basis and their values are adopted from the Oslo 2D tropospheric-stratospheric global model.

In Gupta et al. [1996] we used annual surface sources of CH₄ based on published research by ourselves and others to initiate the model. Methane source magnitudes and distributions were slightly modified from those appearing in Fung et al. [1991]. The assigned average d¹³C-CH₄ for individual sources was adapted from source budgets appearing in Cicerone and Oremland [1988] and newer information from additional CH₄ source studies. The adopted source magnitudes and assigned d¹³C-CH₄ values data appear in Table 1.

Lake
Swamps
Tundra
Rice paddies
Animals
Landfills
Gas venting and leakage
Oceans
Coal minings
Biomass burning
Termites

64
85
5
90
85
42
53
5
27
42
32

-59*
-56*
-55*
-59
-62
-50
-38
-40
-37
-26
-57

Page 2