pogenic.4 Any changes in emissions and removals from these lands are thus considered anthropogenic, regardless of whether natural factors contributed to those changes.
In addition to the definitional ambiguities, monitoring of anthropogenic CO2 emissions is greatly complicated by the natural cycling of CO2 through the terrestrial biosphere and oceans (Figure 1.3). The terrestrial biosphere takes up approximately 120 Gt C yr–1 through photosynthesis and releases almost all of it back to the atmosphere through respiration by plants, animals, and microbes (IPCC, 2007a). Photosynthesis occurs only during daylight hours in the growing season, whereas respiration occurs at all times, albeit at a reduced rate in some seasons (i.e., winter outside the tropics). This diurnal and seasonal imbalance can be quite large; the CO2 sources and sinks that they create are often larger than fossil-fuel fluxes in the same location, except in cities or close to power plants where fossil-fuel emissions are concentrated. Moreover, if we ignore tropical deforestation, terrestrial ecosystems represent a net sink that averaged 2.7± 1.0 Gt C yr–1 over 2000-2008 (Le Quéré et al., 2009). The cause of this sink is not completely understood, although a substantial fraction is due to forest regrowth and other land-use changes in the temperate zone (CCSP, 2007, Chapters 2 and 3) and the remaining fraction may be caused by CO2 fertilization (Friedlingstein et al., 2006). The size of the net terrestrial flux can change from year to year by as much as 5 Gt C (Baker et al., 2006a), in part from anthropogenic fires in tropical forests associated with El Niño events (Randerson et al., 2005; van der Werf et al., 2009b), but is usually within a range of ±1 Gt C yr–1.
The oceans are also a sink for carbon averaging 2.3 ± 0.5 Gt C yr–1 from 2000 to 2008 (Le Quéré et al., 2009). By measuring the changing chemical properties (e.g., pH, pCO2) of the surface ocean from research vessels and commercial ships of opportunity, the annual sink assignable to an ocean basin can be estimated to a precision of about ±10 percent (e.g., Watson et al., 2009). These measurements show that variations in the oceanic sink are too small to explain the multi-gigaton fluctuations in the atmospheric increase of CO2 (IPCC 2007a; Le Quéré et al., 2009). Moreover, because of their comparatively high accuracy, estimates of the oceanic sink provide a valuable constraint on estimates of the magnitude of land sinks at regional and global scales (because the land sink equals the fossil-fuel source minus ocean uptake minus the atmospheric increase).
Fluctuations of natural CO2 sources and sinks create a difficult signal-to-noise problem for efforts to estimate anthropogenic emissions with atmospheric measurements. Seasonally fluctuating background sources and sinks that contribute to the CO2 signal may be of the same order as the emission reductions that might be required under a treaty. The signal-to-noise problem is further exacerbated by the fact that annual fossil-fuel and deforestation emissions represent only about 1 percent of the CO2 in the atmosphere (IPCC, 2007a). This means that anthropogenic emissions will change the average CO2 abundance by only a small amount as air moves across a country over a period of hours to a few days. Thus, an effective way to uniquely identify many large emissions sources is to measure the perturbation in air close to the source, before mixing dilutes the added CO2. The plume of increased concentration above a major point source can be of order 1-10 percent above the background concentration (see Chapter 4).
Because of the signal-to-noise problem, the natural carbon cycle would have to be monitored as part of any effort to monitor anthropogenic emissions. Monitoring the carbon cycle would also constrain estimates of “leakage,” in which reduced emissions in one region or sector lead to increased emissions in another (i.e., soil carbon releases from land newly cultivated for biofuels; see Searchinger et al., 2008; Tilman et al., 2009). Further, the effectiveness of any climate treaty is based on the stabilization of greenhouse gas abundances, whether from anthropogenic or natural sources. Current models indicate that climate change feeds back on natural ecosystems and the ocean to produce new sources or reduce sinks of greenhouse gases, with most of the feedbacks amplifying climate change. For example, warming might cause arctic tundra to emit large quantities of CO2 and CH4, causing further climate change, even more releases of CO2 and CH4, and so on in a positive feedback loop (Walter et al., 2006; Zimov et al., 2006; IPCC, 2007a; Schuur et al., 2009). These effects need to be detected early to ensure that