The North American Carbon Budget

Since the Industrial Revolution, human activity has released into the atmosphere unprecedented amounts of carbon-containing greenhouse gases (GHGs), such as carbon dioxide (CO₂) and methane (CH₄), that have influenced the global carbon cycle. For the past three centuries, North America has been recognized as a net source of CO₂ emissions to the atmosphere (Houghton 1999, 2003; Houghton and Hackler 2000; Hurtt et al., 2002). Now there is greater interest in including in this picture emissions of CH₄ because it has 28 times the global warming potential of CO₂ over a 100-year time horizon (Myhre et al., 2013; NAS 2018).

The major continental sources of CO₂ and CH₄ are 1) fossil fuel emissions, 2) wildfire and other disturbances, and 3) land-use change. Globally, continental carbon sources are partially offset by sinks from natural and managed ecosystems via plant photosynthesis that converts CO₂ into biomass. The terrestrial carbon sink in North America is known to offset a substantial proportion of the continent’s cumulative carbon sources. Although uncertain, quantitative estimates of this offset over the last two decades range from as low as 16% to as high as 52% (King et al., 2015). Highlighted in this chapter are persistent challenges in unravelling CH₄ dynamics across North America that arise from the need to fully quantify multiple sources and sinks, both natural (Warner et al., 2017) and anthropogenic (Hendrick et al., 2016; Turner et al., 2016a; NAS 2018). Adding to the challenge is disagreement on whether the reported magnitudes of CH₄ sources and sinks in the United States are underestimated (Bruhwiler et al., 2017; Miller et al., 2013; Turner et al., 2016a).

At the global scale, about 50% of annual anthropogenic carbon emissions are sequestered in marine and terrestrial ecosystems (Le Quéré et al., 2016). Temporal patterns indicate that fossil carbon emissions have increased from 3.3 petagrams of carbon (Pg C) per year to almost 10 Pg C over the past 50 years (Le Quéré et al., 2015). However, considerable uncertainty remains in the spatial patterns of emissions at finer scales over which carbon management decisions are made. Most importantly, the sensitivity of terrestrial sources and sinks to variability and trends in the biophysical factors driving the carbon cycle is not understood well enough to provide good confidence in projections of the future performance of the North American carbon balance (Friedlingstein et al., 2006; McGuire et al., 2016; Tian et al., 2016).

2.1.1 Approaches for Estimating Carbon Budgets

Historically, the existence (if not the magnitude) of the land sink has been confirmed by inventory-based approaches involving the extrapolation of ground-based measurements to regional, national, and continental scales (Caspersen et al., 2000; Goodale et al., 2002; Pan et al., 2011). Regional- to ­continental-scale estimates of the magnitude and variability of the terrestrial carbon sink differ substantially among assessments, depending on the measurement or scaling approach used and the budget components considered (Hayes and Turner 2012; King et al., 2015). Estimations of land-based carbon budgets over large domains, typically involving a combination of measurements and modeling, generally can be categorized as either “top-down” (atmosphere-based) approaches or “bottom-up” (biosphere-based) approaches (e.g., field measurements and ecosystem process models).

Top-down approaches provide a reliable constraint on overall land-atmosphere carbon exchange based on direct measurement of spatial and temporal patterns in CO₂ concentrations. Regional-scale estimates of net ecosystem exchange (NEE; i.e., the net exchange of CO₂ between land and atmosphere) are derived from these observations using different techniques ranging from simple boundary-layer budget approaches (Wofsy et al., 1988) to upscaling eddy covariance data (Jung et al., 2009; Xiao et al., 2014) to more complex inverse modeling of atmospheric transport (Gurney et al., 2002). Atmosphere-based estimates are broadly inclusive and treat all surface-atmosphere CO₂ exchange as one integrated flux. However, such estimates have limited attribution information on 1) stock changes within individual components, 2) internal processes, 3) lateral transfers, or 4) the exact location of carbon sinks and sources, which is derived from ­biosphere-based approaches.

Plot-based measurements serve as the basis for bottom-up approaches—either directly, as input to inventory-based methods (e.g., Birdsey and Heath 1995; Stinson et al., 2011), or indirectly through their use in calibrating ecosystem process models (e.g., McGuire et al., 2001). Although researchers can apply bottom-up approaches at broad scales to estimate flux components individually, evidence suggests there are important carbon pools and fluxes that are undersampled, have large or unknown uncertainties, and are not inventoried or modeled (Hayes et al., 2012; Warner et al., 2017). Despite these limitations, bottom-up methods (e.g., inventories) typically are cited in broader-scale carbon cycle assessments (e.g., Goodale et al., 2002; Pacala et al., 2007; Pan et al., 2011) that favor these approaches for their use of large amounts of measurements, ability to track the total change in ecosystem carbon pools, and comparability among estimates.

2.1.2 Carbon Cycling Synthesis Efforts

Terrestrial carbon budget estimates at global, national, and continental scales have proliferated in recent years. Prominent examples are the Forest Inventory and Analysis (FIA) Program of the U.S. Forest Service (fia.fs.fed.us) within the U.S. Department of Agriculture (USDA), the National Aeronautics and Space Administration’s (NASA) Carbon Monitoring System (carbon.nasa.gov), and the National Oceanic and Atmospheric Administration’s (NOAA) CarbonTracker (esrl.noaa.gov/gmd/ccgg/carbontracker; see also Appendix C: Selected Carbon Cycle Research Observations and Measurement Programs). The U.S. Forest Service is adopting a new approach to carbon accounting that moves FIA data through time by attributing changes in the complete set of pools to disturbance and land use (Woodall et al., 2015). The goal of this new approach is to provide improved estimates of the magnitude and uncertainty of carbon fluxes, along with more detailed information on the drivers and fate of carbon change. In the last decade, the understanding of the North American carbon budget has moved beyond terrestrial emissions and sinks to incorporate anthropogenic, aquatic, and coastal margin CO₂ and CH₄ dynamics. Since the First State of the Carbon Cycle Report (SOCCR1; CCSP 2007), multiple research efforts have aimed to synthesize and reconcile estimates across the key components of the continental-scale carbon cycle. A series of studies borne from the REgional Carbon Cycle Assessment and Processes (RECCAP) initiative has provided diagnosis and attribution of carbon cycle dynamics for global regions, including North America (King et al., 2015). Designed to advance research from SOCCR1 toward the Second State of the Carbon Cycle Report (SOCCR2), several “interim synthesis” studies organized by the North American Carbon Program (NACP; nacarbon.org) compared observational, inventory-based, and modeled estimates of carbon stocks and fluxes across sites (Schwalm et al., 2010), within subregions (Schuh et al., 2013), and over the continent (Huntzinger et al., 2012). Currently, the Global Carbon Project (globalcarbonproject.org) develops global- and regional-scale estimates of CO₂ (Le Quéré et al., 2018) and CH₄ (Saunois et al., 2016) budgets. Collectively, these efforts comparing and synthesizing information across various sources of data and methods have improved the understanding of the North American carbon cycle.

2.1.3 Chapter Objectives

This chapter synthesizes the latest scientific information on the North American carbon budget, incorporating terrestrial, anthropogenic, aquatic, and coastal margin CO₂ and CH₄ dynamics. The estimates used to develop the continental-scale budget presented here are summarized from previous results based on different methodological approaches encompassing three countries (i.e., Canada, the United States, and Mexico), the U.S. National Climate Assessment regions, and the major carbon sectors (see Figure 2.1).

Specifically, this chapter follows the estimates of North American carbon stocks and fluxes synthesized and reported in Chapter 3 of SOCCR1 (Pacala et al., 2007). That analysis defined the reported estimates as “ca. 2003” to represent the approximate time period of SOCCR1. Here, these estimates are updated for the 2004 to 2013 time frame, or the decade since SOCCR1. However, SOCCR2 does not always rigidly follow these exact dates when combining and reconciling various reported estimates of the different components that make up the carbon budget. As explained where appropriate within this chapter, some datasets have a temporal resolution allowing precise time periods to be summarized, but others do not. As such, this chapter attempts to synthesize the various budget components using reported estimates and datasets generally representative of the 2004 to 2013 time period. Also summarized in this chapter are the historical and current context of continental carbon fluxes and stocks; recent findings of indicators, trends, and feedbacks; and a discussion about social drivers and implications for carbon management decisions.