Karen B. Bartlett

Review and assessment of methane emissions from wetlands

The number of emission measurements of methane (CH4) to the atmosphere has increased greatly in recent years, as recognition of its atmospheric chemical and radiative importance becomes widespread. In this report, we review progress on estimating and understanding both the magnitude of, and controls on, emissions of CH4 from natural wetlands. We also calculate global wetland CH4 emissions using this extensive flux data base and the wetland areas compiled and published by Matthews and Fung (1987). Tropical regions (20° N-30° S) were calculated to release 66 TgCH4/yr, 60% of the total wetland emission of 109 Tg/yr. Flux data from tropical wetlands, reported only within the last four years, are currently restricted in geographic coverage. Additional data from other regions will be required to confirm these calculated large emissions. Although emissions from subtropical and temperate wetlands (45° N-20° N and 30° S-50° S) were relatively low at 5 Tg/yr, the process-oriented focus of most of the research in this region suggests that work at these latitudes may serve as models to examine controls and possible uncertainties in estimating fluxes. These types of efforts are frequently not possible in more remote, globally significant wetlands. Northern wetlands (north of 45° N) were calculated to release a total of 38 TgCH4/yr (34% of total flux); 34 Tg/yr from wet soils and 4 Tg/yr from relatively dry tundra. These latitudes have been the focus of recent intensive research. Significant differences between the relatively large flux data bases accumulated in the two primary measurement areas, northern Minnesota and the Hudson Bay Lowlands of Canada, indicate that extrapolation from one wetland region to another may be subject to considerable error. Global emissions were also compared to fluxes calculated using the wetland areas published by Aselmann and Crutzen (1989) in an effort to assess uncertainties due to wetland area estimates. Further refinement of wetland CH4 emissions awaits flux measurements from large areas currently lacking data, particularly in the tropics and the Siberian Lowlands, more realistic assessments of seasonal active periods, and accurate, up-to-date habitat classification and measurement.

Methane emissions from tundra environments in the Yukon‐Kuskokwim delta, Alaska

Over a 6-week period from July 3 to August 10, 1988, we made measurements of the flux of methane by enclosure techniques from major tundra environments in the Yukon-Kuskokwim Delta of Alaska (60°45′N). Emissions from wet meadow tundra averaged 144 ± 31 mg CH4/m2/d and ranged from 15.6 to 426 mg CH4/m2/d, varying with soil moisture and temperature. Flux from the drier upland tundra was approximately 2 orders of magnitude lower and averaged 2.3 ± 1.1 mg CH4/m2/d. Consumption of ambient levels of methane was sporadically measured at these drier sites, and emissions ranged between −2.1 and 18.1 mg CH4/m2/d. Tundra lakes emit methane from the open water surface as well as from fringing aquatic vegetation. The presence of vegetation significantly enhanced flux over open water rates. Average fluxes from a variety of sites with vegetation ranged between 62.7 and 153.5 mg CH4/m2/d. Calculated diffusive fluxes from open water varied with lake size, the large lakes emitting 3.8 mg CH4/m2/d and small lakes emitting an average of 77 mg CH4/m2/d. An updated estimate of global emissions from tundra indicates an annual flux of approximately 11 ± 3 Tg CH4.

Methane emissions along a salt marsh salinity gradient

The seasonal flux of methane to the atmosphere was measured at three salt marsh sites along a tidal creek. Average soil salinities at the sites ranged from 5 to 17 ppt and fluxes ranged from below detection limits (less than 0.3 mgCH4 m-2 d-1) to 259 mgCH4 m-2 d-1. Annual flux to the atmosphere was 5.6 gCH4 m-2 from the most saline site, 22.4 gCH4 m-2 from the intermediate site, and 18.2 gCH4 m-2 from the freshest of the three sites. Regression of the amount of methane in the soil with flux indicates that changes in this soil methane can account for 64% of the observed variation in flux. Data on pore water distributions of sulfate suggests that the activity of sulfate reducing bacteria is a primary control on methane flux in these transitional environments. Results indicate that relatively high emissions of methane from salt marshes can occur at soil salinities up to approximately 13 ppt. When these data are combined with other tidal marsh studies, annual CH4 flux to the atmosphere shows a strong negative correlation with the long term average soil salinity over a range from essentially fresh water to 26 ppt.

Seasonal variation of methane emissions from a temperate swamp

Methane flux measurements were made at four sites in a freshwater temperate swamp over the 13 month period of April 1985 through May 1986. Emissions were highly variable both between sites and over time at any one site. Ebullition from sediments was an important component of methane release. Although release of methane through bubbling occurred in only 19% of the measurements made between April and June 1985, when instrumentation allowed us to separate diffusive and bubble fluxes, ebullition accounted for 34% of the total flux during this period. Methane release rates showed a strong seasonal variation, with highest emission rates observed in early spring and again in late summer, which was associated with changes in plant growth and physiology. Emission rates were partially correlated with sediment temperature, but the relationship was not straightforward, and resembled a step function. Emissions responded strongly to temperature change through the range of 10–16°C. At winter sediment temperatures between 4–9°C, CH4 flux continued at low rates (0–28 mg CH4 m−2d−1; average = 7.9 mg CH4m−2d−1) and appeared insensitive to changes in sediment temperature. Annual methane emission from three constantly flooded sites (mean water depth = 35 cm) was 43.7 +/- 7.8 gm−2 (standard error); annual flux from a bank site was 41.4 +/- 20.5 gm−2. A comparison of flux measurements from fresh and saline wetlands in the immediate area of Newport News Swamp emphasizes the importance of edaphic factors in controlling flux.

Carbon monoxide and methane over Canada: July–August 1990

Carbon monoxide (CO) and methane (CH4) were measured in the 0.15- to 6-km portion of the troposphere over subarctic and boreal landscapes of midcontinent and eastern Canada during July–August 1990. In the mid-continent region, Arctic air entering the region was characterized by relatively uniform CO concentrations (86–108 parts per billion by volume (ppbv)) and CH4 concentrations (1729–1764 ppbv). Local biomass burning and long-range transport of CO into the area from industrial/urban sources and distant fires did frequently produce enhanced and variable concentrations. Emissions of CH4 from the Hudson Bay lowlands was the primary source for enhanced and variable concentrations, especially at altitudes of 0.15–1 km. In eastern Canada, most of the observed variability in CO and CH4 was similar in origin to the phenomena described for the midcontinent region. However, unexpectedly low concentrations of CO (51 ppbv) and CH4 (1688 ppbv) were measured in the midtroposphere on several flights. Combined meteorological and chemical data indicated that the low CO-CH4 events were the result of long-range transport of tropical Pacific marine air to subarctic latitudes.

Carbon monoxide and methane in the North American Arctic and Subarctic troposphere: July–August 1988

Measurements of carbon monoxide (CO) and methane (CH4) were made in the North American Arctic during July–August 1988. The distribution of CH4 was variable in the atmospheric mixed layer (0–2 km), with concentrations determined primarily by interactions of biogenic emissions from wet tundra and turbulent mixing processes. Carbon monoxide exhibited little variation in unpolluted mixed layer environments indicating a minor role for biogenic sources and/or sinks in determining its distribution. In the free troposphere (2–6 km) both CO and CH4 were variable. Concentration gradients were most frequently associated with intrusions of upper tropospheric or stratospheric air into the midtroposphere, emissions from forest and tundra fires, and long-range transport of enhanced concentrations of these gases from unidentified sources. Summertime haze layers exhibited midtropospheric enhancements of CH4 similar to those measured in winter Arctic haze events. However, these summer pollution episodes did not exhibit positive correlations with particulate sulfate. The summer Arctic and subarctic haze events observed during the Arctic Boundary Layer Expedition (ABLE 3) were primarily a result of forest and tundra fires of natural origin. The tendency for relatively high variability of CO and CH4 at altitudes of 3–6 km indicates that ground-based monitoring will not provide an adequate assessment of the chemical composition of the Arctic troposphere to support future global change studies.

Methane Emissions from Northern High-Latitude Wetlands

Methane emissions from northern high-latitude wetlands are an important consideration for understanding past, present, and future atmospheric concentrations of this important greenhouse gas. In this chapter we review progress on measuring methane emissions from northern wetlands and, through a model, estimate emission variability in relation to one component of climate variability. Our conclusions are as follows: (1) Methane emissions from northern wetlands are dependent on both soil moisture and temperature. The relative influence of these soil climate parameters is quite variable from one region to another, as is the magnitude of the net emission rate to the atmosphere. Some important wetland regions have not been surveyed for methane emissions (e.g., the Siberian Lowlands); further progress on defining global emissions from northern wetlands awaits field data from these areas. (2) Our preliminary modeling of the sensitivity of methane flux from northern wetlands to variability in temperature indicates that feedbacks from this source are unlikely to significantly influence rates of climate change during the initial stages of a global warming.

Natural and anthropogenic methane sources in New England

We have recently completed a methane emissions inventory for the New England region. Methane emissions were calculated to be 0.91 Tg yr-1, with wetlands and landfills dominating all other sources. Wetlands are estimated to produce 0.33 Tg CH4 yr-1, of which 74% come from Maine. Active landfills emit an estimated 0.28 Tg CH4 yr-1, 60% of which are generated from twelve landfills. Although uncertainty in the estimate is greater, emissions from closed landfills are on the same order of magnitude as active landfills and wetlands; 0.25 Tg CH4 yr-1. Sources of moderate magnitude include ruminant animals (0.05 Tg CH4 yr-1) and residential wood combustion (0.03 Tg CH4 yr-1). Motor vehicles, natural gas, and wastewater treatment make only minor contributions. New England is heavily forested and the soil uptake of atmospheric methane in upland forests, 0.06 Tg CH4 yr-1, decreases emissions from soils by about 18%. Although uncertainties remain, our estimates indicate that even in a highly urbanized region such as New England, natural sources of methane make the single greatest contribution to total emissions, with state totals varying between 8% (Massachusetts) and 92% (Maine). Because emissions from only a few large landfills dominate anthropogenic sources, mitigation strategies focused on these discrete point sources should result in significant improvements in regional air quality. Current federal regulations mandate landfill gas collection at only the largest sites. Expanding recovery efforts to moderately sized landfills through either voluntary compliance or further regulations offers the best opportunity to substantially reduce atmospheric methane in New England. In the short term, however, the large contribution from closed, poorly regulated landfills may make the attribution of air quality improvements difficult. Mitigation efforts toward these landfills should also be a priority.

Airborne nitrous oxide observations over the western Pacific Ocean: September–October 1991

The Langley tunable diode laser instrument package incorporated an additional channel to measure nitrous oxide (N2O) during the Pacific Exploratory Mission (PEM) West A. These measurements represent the first airborne, fast response (5-s) N2O data set obtained within the troposphere. Most data were recorded over the western Pacific between 0°N and 45°N latitude, 110°E and 180°E longitude, and 0.3 to 12 km altitude. Important observations include a decreasing N2O latitude gradient of approximately 0.4 parts per billion volume (ppbv) from northern midlatitudes to the equator, a decreasing N2O longitude gradient of 0.4 ppbv from the western Pacific to the central Pacific at northern midlatitudes, and an enhancement of 0.2 ppbv in the boundary layer (altitudes below 0.5 km) relative to the rest of the tropospheric vertical profile. Other observations include increased N2O mixing ratios within both urban and biogenic affected air masses and reduced N2O mixing ratios in stratospheric intrusions. These relationships with air mass source characteristics are exhibited in the large-scale correlations between N2O and CO, CH4, and CO2 in the free troposphere. Atmospheric inputs of N2O are examined and the relative strengths of continental biogenic and anthropogenic/industrial sources are estimated. The data set is also examined for evidence of an oceanic source of N2O.

Methane in the tropical South Atlantic: Sources and distribution during the late dry season

Methane (CH 4 ) mixing ratios in the South Atlantic basin were sampled from a DC-8 aircraft during the TRACE A expedition over the months of September and October 1992. This high-precision (± 0.1%), high-resolution data set (independent measurements every 5 s) includes a total of 67,335 observations and ranges from 1586.6 to 2152.8 parts per billion by volume (ppbv). The observed values were influenced by emissions from biomass burning and local urban/industrial areas, stratosphere-troposphere exchange, active convective mixing, and possible long-range transport of pollution. Average mixing ratios increased from the atmospheric mixed layer (0-2 km) to the free troposphere (2-6 km) and were greatest at altitudes above 6 km. A longitudinal trend observed between 5°N and 40°S latitude suggests that CH 4 inputs were greater on the South American side of the basin. A latitudinal trend in values between the near-surface and 11 km exhibits the expected gradient from north to south, but differs from the latitudinal gradient observed in surface level clean air at this time. Higher concentrations at altitude, at least at this time of year, indicate that ground-based atmospheric monitoring sites may underestimate pollution inputs to the region.