Transcontinental methane measurements: Part 2. Mobile surface investigation of fossil fuel industrial fugitive emissions

Atmospheric methane, CH₄, is a potent greenhouse gas whose integrated radiative impact is larger than that of CO₂ on a twenty year time-scale (IPCC, 2007), while indirect effects over CH₄'s entire chemical lifetime suggests an even larger contribution to the atmospheric radiative balance (Shindell et al., 2005). Growth of the atmospheric CH₄ mixing ratio slowed in the 1990s, almost stabilizing in the following decade, a trend proposed to result from reduced anthropogenic emissions, primarily fossil fuel industrial, FFI, activity (Aydin et al., 2011), competing with wetland emission increases (Bousquet et al., 2006; Simpson et al., 2012), although, decreased microbial emissions also has been proposed to underlie the trend (Kai et al., 2011). CH₄ growth has resumed since 2008 (Dlugokencky et al., 2011; Heimann, 2011; Rigby et al., 2008).

Natural (145–260 Tg yr⁻¹) and anthropogenic (264–428 Tg yr⁻¹) CH₄ sources release a combined ∼582 ± 87 Tg yr⁻¹ (IPCC, 2007). FFI activity is one of the largest CH₄ sources, with release during production, refining, and distribution (NRC, 2010). FFI CH₄ is ancient, allowing isotopic discrimination from modern CH₄. Also contributing is the natural CH₄ seepage from geologic reservoirs. Lassey et al. (2007) estimated all fossil sources account for 30% of the global budget.

CH₄ flaring occurs during oil production and largely has been stable at 1.4–1.7 ×ばつ 10¹¹ m³ yr⁻¹ CH₄, although significant reductions occurred since 2005 in Russia, the country with the highest flaring emission rates. Flaring efficiency is estimated at 98–99.5%, but can decrease to 90% in strong winds (Johnson and Kostiuk, 2002). This implies a lower limit global budget of 2–5 Tg yr⁻¹ from petroleum production regions. The flaring inventory may miss contributions from smaller, but more frequent, flares due to difficulty in satellite observations (Gallegos et al., 2007). Refining FFI CH₄ emissions are important and generally located near the customer, not production area. North American refining is the largest in the world (total US refining capacity is 18.2 million bbl. day⁻¹ (7.2 ×ばつ 10⁸ L day⁻¹)) followed by Asia (EIA, 2011b). US refineries primarily are located in the Northern Gulf of Mexico region and California to a lesser extent.

Natural gas distribution system emissions arise from buried pipelines, gas processing plants, and storage fields (Chambers et al., 2006). A survey of industry and EPA estimates for US, Canadian, and European systems suggested CH₄ emissions are <2% of gas throughput (Delucchi, 2003). Global coal CH₄ emissions are estimated at 35 Tg yr⁻¹ (NRC, 2010).

Urban areas are strong CH₄ sources (Shorter et al., 1996; Thomas and Zachariah, 2010), which often contain numerous strong, spatially co-located CH₄ sources (Leifer et al., 2012), such as industrial, landfills, and sewage treatment plants. Emissions also arise from dispersed sources including natural gas distribution, residential wood and natural gas heating, and urban park wetlands (Piccot et al., 1996). Vehicles emit fossil CH₄, mostly from urban areas with US emissions of ∼2 Tg yr⁻¹ (Piccot et al., 1996).

Anthropogenic sources exhibit annual and cyclical trends, for example, natural gas heating (and leakage) is winter dominated (Sasakawa et al., 2010), while application of market pricing to Soviet production lowered CH₄ emissions by ∼10 Tg yr⁻¹ (Dlugokencky et al., 2011). Cyclical anthropogenic trends also occur from FFI activity like regularly scheduled refinery maintenance shutdowns.

Source strengths for global CH₄ budgets are derived in two manners, from top-down estimates based on atmospheric measurements and inversion modeling and from bottom-up inventory estimates of individual sources (NRC, 2010). SCIAMACHY (Scanning Imaging Absorption Spectrometer for Atmospheric Chartography (Bovensmann et al., 1999; Frankenberg et al., 2011; Buchwitz et al., 2005)), launched in 2002, enabled the first global dry-column CH₄ measurements from space. More recently, GOSAT (Greenhouse gases Observing SATellite) (Kuze et al., 2009) also provides tropospheric global CH₄ data. However, a comparison between top-down and bottom-up estimates can indicate under-inventoried or un-inventoried CH₄ sources, e.g., Hsu et al. (2010) for the Los Angeles Basin. Such comparisons are challenging at mesoscales (Schneising et al., 2009). Ground- and satellite-based measurements are complementary. Ground-based measurements have very high accuracies but are intrinsically sparse. Satellite data products are not sparse, have global coverage, and enable top-down estimates of budgets for CO₂ and CH₄ to be determined (see Schneising et al., 2011 and references therein). However as current satellite footprints are large, 30 ×ばつ 60 km and 10.5 ×ばつ 10.5 km for SCIAMACHY and GOSAT, respectively, direct source interpretation is limited to large-scale (order 100 km) emission features (Frankenberg et al., 2005). Sadly, Envisat contact was lost on 9 April 2012 ending a decade of SCIAMACHY measurements.

Global satellite data of the dry columns of greenhouse gases are needed to infer global budgets (Bergamaschi et al., 2009); however, current and recent satellite instruments do not resolve most small-scale sources. Often, satellite data retrievals are validated by airborne (Wecht et al., 2012) or ground station measurements (Schneising et al., 2012). Fixed station and mobile ground measurements near emission sources are valuable for the investigation and identification of individual sources (Herndon et al., 2005; Pétron et al., 2012; Shorter et al., 1996) but generally span few satellite pixels, which requires mesoscale to continental-scale data.

High spatial-resolution, transcontinental (Florida to California), mobile surface CH₄ data were collected 6–12 Oct. 2010 near important CH₄ sources (Fig. 1A), supplemented with S. California survey data collected Feb. 2012. Observations for important natural sources, wetlands, fire, and natural terrestrial seepage are presented in Farrell et al. (2013). The goal of the present study is to report on FFI-related CH₄ data to understand better their contribution. Although CH₄ surface anomalies were better compared with FFI activity than wetlands, the lack of meteorological transport modeling and information on boundary layer height prevented flux calculations for most sources.

The survey data provides a single-route snapshot; thus, satellite data are required for larger spatial and temporal context. Surface data and SCIAMACHY CH₄ anomaly trends along the survey path showed similar patterns that were not correlated with retrieval factors like altitude (Fig. 1C) or humidity. Seasonal SCIAMACHY CH₄ data then were compared qualitatively with inventories for anthropogenic and natural sources and winds to identify large and inventory errors even for qualitative comparison.

The methodology and approach and data quality are described in detail in Farrell et al. (2013) and briefly summarized herein. In situ CH₄ measurements were made (Fig. 1A) by flame ion detection on a gas chromatograph, GC, aboard a 10-m recreational vehicle (RV), which enabled near-continuous mobile data collection over extended areas at speeds up to highway, and analysis a few months later. Additional data were collected in 2012 with a cavity ring-down spectrometer (Greenhouse Gas Monitor,

This study found strong and extensive positive surface methane anomalies associated with a range of FFI activities including refining, production, natural gas distribution, and coal loading. Surface and satellite-derived (SCIAMACHY) methane anomaly trends were similar. For some areas, such as Corpus Christi, Texas and Cantarelle, Mexico, and some seasons (e.g., winter), satellite anomaly patterns were consistent with dominant FFI emissions. In other areas and seasons, numerical transport

This work was supported by the Gulf of Mexico Hydrates Research Consortium administered by the Center for Marine Resources and Environmental Technology at the University of Mississippi through the Department of Energy's Cooperative Agreement Award No. DE-FC26-06NT42877 and the National Science Foundation, ATM Rapid Response program, Award No. 1042894 and NASA award No NNX12AQ16G. I.U.P. received ESA (GHG-CCI project), DLR (SADOS project) and the State and the University of Bremen support.