Observations of Atmospheric Carbon Dioxide and Methane

Atmospheric CH₄ and CO₂ levels continue to increase. In the case of CO₂, this increase is unambiguously a result of anthropogenic emissions, primarily from fossil fuel combustion, with North America accounting for about 20% of global emissions. The recent rise in global CH₄ concentrations (see Figure 8.1), on the other hand, has been attributed primarily to biological, not fossil, processes on the basis of a concomitant decrease in the global mean ¹³C:¹²C ratio and the tropical origin of the increase (Nisbet et al., 2016; Schaefer et al., 2016; Schwietzke et al., 2016). Two recent analyses render the causes of recent CH₄ growth rate changes less clear. First, studies have pointed out that the tropospheric CH₄ sink may not have been constant over recent years as had been assumed (Rigby et al., 2017; Turner et al., 2017). Secondly, Worden et al. (2017) suggest that atmospheric δ¹³C of CH₄ may have decreased because of less biomass burning, thus allowing for an increase in isotopically heavier fossil fuel CH₄ sources. Nonetheless, these results mostly pertain to the global mean and do not directly bear on potential trends in North American emissions. Despite the recent increase in oil and gas production due to new extraction technologies, both inventories and atmospheric inversions do not reveal an increase in North American CH₄ emissions (Bruhwiler et al., 2014; Miller et al., 2013; U.S. EPA 2016; see Figure 8.3). Normalizing CH₄ and CO₂ emissions using a 100-year global warming potential (GWP) indicates that U.S. radiative forcing from CH₄ emissions from 2000 to 2013 equates to just 13% of that from CO₂. Changes in U.S., Canadian, and Mexican energy systems will affect the atmospheric trends of anthropogenic CO₂ and CH₄, but U.S. GHG emissions currently are dominated by CO₂ and are likely to remain so for the foreseeable future.

Much less certain than anthropogenic CO₂ sources is the balance of biogenic sources (respiration and fire) and sinks (photosynthesis). There is general agreement that the terrestrial biosphere of the United States, and North America as a whole, acts as a CO₂ sink (see Figure 8.3 and Table 8.1; Hayes et al., 2012; King et al., 2015), but there is substantial uncertainty about the location of and reasons for the sinks. There is evidence that their interannual variability is driven largely by climatic factors. For example, Peters et al. (2007) presented evidence for a direct effect of drought on the North American sink. Understanding the spatial and temporal variability of sinks is critical, because positive feedbacks between net ecosystem CO₂ exchange and climate represent a first-order uncertainty in climate projections (Bodman et al., 2013; Booth et al., 2012; Friedlingstein et al., 2006, 2014; Huntingford et al., 2009; Wenzel et al., 2014; Wieder et al., 2015). At hemispheric and global scales, atmospheric CO₂ data have proved to be a powerful constraint on the representation of the carbon cycle (including, to some measure, feedbacks) in climate models (e.g., Cox et al., 2013; Graven et al., 2013; Keppel-Aleks et al., 2013; Randerson et al., 2009). The present generation of global atmospheric inverse models is limited by the accuracy and resolution (generally about 1° × 1°) of meteorological transport, availability and accuracy of prior flux emissions, uncertainty about the spatial coherence of prior flux errors, and the limited set of observation sites shown in Figure 8.2. Together, these limitations mean that, at present, global atmospheric inverse models cannot unambiguously resolve source-sink patterns below the scale of 5 to 10 million km². A new generation of regional and local models using much higher resolution meteorology (e.g., approaching the approximately 1- to 4-km resolution used by Lauvaux et al. [2016] and McKain et al. [2015]) will be more capable of assimilating data from the sites in Figure 8.2. Without quantitative knowledge of the spatial structure of flux uncertainties (Cooley et al., 2012; Ogle et al., 2015) and atmospheric transport errors (Díaz Isaac et al., 2014; Lauvaux and Davis 2014), these high-resolution inverse systems will have limited ability to determine the spatial structure of fluxes (Lauvaux et al., 2012a, 2016). Nonetheless, these improved inversion systems should enable better understanding of the climate-carbon relationship in North America.