Chapter 8: Observations of Atmospheric CO₂ and CH₄

Overview/Main Issues

The chapter has a clear organizing structure and clear findings that summarize the current state of research on monitoring global CO₂ and CH₄ and using inverse analysis to resolve carbon emission and uptake estimates. Discussed below are some suggestions for how the chapter could be augmented to provide a more complete picture of the state of research in this realm—in particular regarding discussion of current understanding of North American trends in CH₄ emissions.

Statement of Task Questions

• Are the goals, objectives and intended audience of the product clearly described in the document? Does the report meet its stated goals?

If the goal of this chapter is to define the stock and trends of CO₂ and CH₄ in the atmosphere and to present best estimates of carbon emissions and sinks at global, continental, and national scales, then yes, it does an excellent job of that. The first paragraph of introduction alludes to that goal “Atmospheric concentration measurements of these two species provide fundamental constraints on sources and sinks, quantities that need to be monitored and understood in order to guide societal responses to climate change. These atmospheric observations also have provided critical insights into the global carbon cycle and carbon stocks and flows among major reservoirs on land and in the ocean.” However, the paragraph could be phrased to more directly state these as goals. The audience isn’t specified, but implicitly it is the same audience as the overall SOCCR2 report.

• Does the report accurately reflect the scientific literature? Are there any critical content areas missing from the report?

Yes, the report accurately reflects the scientific literature. However, more discussion about controversy surrounding inverse analysis of CH₄ emissions in North America is needed.

Mention of the ongoing debate could be merged with discussion about what are the limits on trend detection from inverse analysis using the current array of measurements.

• Are the findings documented in a consistent, transparent and credible way?

Yes, the main findings are well documented.

• Are the report’s key messages and graphics clear and appropriate? Specifically, do they reflect supporting evidence, include an assessment of likelihood, and communicate effectively?

Figure 8.1 needs some work. Putting both CO₂ and CH₄ together in the figure does not work well. It is confusing why only the non-fossil fuel emissions (a negative value, thus a sink) are presented. It would be more informative to show both the fossil fuel source and the non-fossil, net sink, separately for CO₂ and CH₄. Because this is a North American carbon cycle report, it would helpful to separate the North American contribution from the total global contribution. This figure could refer to elsewhere in the chapter for partitioning of the fossil fuel source by energy type if it is presented.

Figure 8.2 shows a good illustration of the expanded CO₂ observation network (not GHG monitoring network as stated in the figure caption). Can a corresponding network of CH₄ observations be presented? Part of reason for not being able to quantify trends in CH₄ could be that the observing network is still too sparse.

Figure 8.3. The CO₂ “emission” panel is not clear. The values are negative, a CO₂ sink. Does that mean this is only the non-fossil fuel contribution as was shown in Figure 8.1. The text does clarify a little that the CO₂ is a sink, nevertheless the figure ought to be able to stand alone to be the main thing some readers will see by looking at the on-line version of the report. This figure would be more informative if it included the fossil fuel source. Although that’s not a result from inverse analysis, it puts the net, non-fossil fuel uptake in context.

Figures are referred to in the text out of order.

• Are the research needs identified in the report appropriate?

The chapter gives a good overview of research needs and next-generation observations that are coming available now.

• Are the data and analyses handled in a competent manner? Are statistical methods applied appropriately?

Statistics from multi-model comparisons are used where appropriate. Uncertainty analysis for the individual models isn’t summarized.

• Are the document’s presentation, level of technicality, and organization effective? Are the questions outlined in the prospectus addressed and communicated in a manner that is appropriate and accessible for the intended audience?

Yes, this is reasonably well done.

• What other significant improvements, if any, might be made in the document?

One place for improvement would be to recognize that the status of North American trends in CH₄ emissions remains somewhat controversial. The chapter reports an absence of evidence for CH₄ emission trends in temperate or boreal regions. Recent papers suggesting otherwise are not mentioned (e.g. Hendrick et al., 2016; Jacob et al., 2016; Kort et al., 2014; Turner et al., 2016). This finding is controversial, and some other recent literature contradicts it, e.g., see Bruhwiler et al., 2017; Miller et al., 2013; Turner et al., 2015; Wecht et al., 2014. Nevertheless, the chapter should point out the disagreement. The chapter could more strongly highlight the point that there is no apparent trend in North American CH₄ emissions despite clear indication that production and use of CH₄ has been increasing in recent decades. Table 2.1 of the 2017 EPA report (EPA 430-P-17-001) shows 2015 has lower total emissions than 2005, with the 2015 upticks in “natural gas systems” and “manure management” nearly cancelled by downturns in emission from landfills. Recent studies generally report that EPA national inventories and EDGAR global inventories are too low for parts of the U.S. (e.g., Miller et al., 2013). If there is no trend in total CH₄ emission despite increased activity that historically has been an important emission source, that is a very important finding— suggesting an

improvement in technology that is reducing leakage and by-product losses. To put this point in perspective, it would help to report what the minimum detection limit would be for quantifying a trend in CH₄ emissions. Is the increase in CH₄ use high enough to exceed the uncertainty in inverse analysis of the sources? Figure 8.3 shows an uncertainty in inverse model estimates of CO₂ and CH₄ source/sink based on standard deviation between models. That is useful, but there are uncertainty analyses presented in the individual inverse analysis papers that provide better quantification of the uncertainty and explore its causes.

The text mentions in several places that the bottom-up inventory estimates of U.S. biospheric CO₂ emissions show less interannual variability than the results from inverse analysis of atmospheric data. The text does note that inventory estimates are based on Forest Inventory and Analysis (FIA) sampling that is only repeated at 5-year or longer intervals. The current approach to estimating U.S. biospheric emissions is simply not designed to provide annual estimates. Comparing the inverse analysis and inventory estimates at annual time steps is not appropriate. A better comparison would be to examine whether decadal averages agree. One can make the point that biospheric CO₂ exchange should be viewed as a multi-year average, but there is no need to belabor the issue. If annual estimates of biospheric CO₂ exchange inventories are desirable, the chapter could point this out as a critical future research need.

Throughout the chapter, the way CO₂ sinks are presented should be checked to ensure consistency with the rest of the SOCCR2 report. Presenting a CO₂ sink as a negative emission value is mathematically correct, but requires the reader to be paying very close attention. Just a note to be sure this is consistent throughout.

There needs to be some additional discussion about the CH₄ sinks. Its lifetime is mentioned, but text doesn’t mention that destruction by OH is the main sink.

The SOCCR2 report should somewhere mention the contribution from ¹³CO₂ isotopes to our understanding of the carbon cycle. If not in overview chapter, then it could be noted here.

• Are the key findings well stated and supported by the detail provided in the chapter?

The three key findings are well stated, sufficiently quantitative, and give a good summary of the supporting evidence and its uncertainties.

Key Finding 1 presents the incontrovertible result that the global atmospheric burdens of CO₂ and CH₄ are increasing. The point made in this finding could be sharpened by giving the pre-industrial values of CO₂ and CH₄ for reference as a final sentence in the finding. Thus the finding would read: “stand at ……, compared to xx ppm and yy ppb for CO₂ and CH₄, respectively in the pre-industrial atmosphere”.

Key Findings 2 and 3 present estimates of emissions and sinks for CO₂ and CH₄ in North America estimated by inverse analysis. Key Finding 2 reports a fairly constant CO₂ emission with small variance and a sink that is about 1/3 of the continental source but has nearly 50% variability and suggestion of increasing trend. Inverse analysis for the land sink disagrees with the inventory estimate.

Key Finding 3 is that CH₄ emissions over North America are fairly constant and do not show clear evidence of trends, unlike global emissions which have been growing over the period. In the key finding statement it isn’t clear how the inverse analysis results compare to reported emissions because they are given in different units (Tg CH₄ vs CO₂ equivalent-100yr). Please use common units.

Conversion to CO₂ equivalent can be added to the text elsewhere if there is a section comparing the budgets of CO₂ and CH₄.

• Are there other key findings or critical literature that are missing?

Turner et al. (2016) and subsequent responses should be mentioned to better characterize the extent of debate in the community about CH₄ emission trends in North America.

P317, Line 28 onwards: Cite Kort et al. (2014) for CH₄ emissions from the Four Corners region of the southwestern U.S.

See comments and references regarding methane budget in Chapter 2.

Line-Specific Comments

P312, Line 24-25

It would be helpful to report the global emission trend and total global sink for the same period in order to address the obvious question of whether the proportion is holding constant or not. Emissions doubled from 5000 Tg in 1980 to 10000 in 2015. Sink increased by 2.5x from 2000 in 1960 to 5000 in 2015

P314, Line 13

There are no CO₂ data yet from the National Ecological Observatory Network (NEON).

P324, Line 8-10

As currently being deployed, the NEON network will NOT be reporting any CH₄ concentrations. They are measured but not going to be computed and reported because the CH₄ calibration was cut. This report would be a good forum to point out this penny-wise pound-foolish decision.

P319, Line 23-25

Does “emission of less than –500 Tg C” mean a greater sink (e.g. of –600 TgC)?

P341, Table 8.1.

Fossil fuel column: boreal and temperate North America do not add up to the North America total.