|Re-Review of revised manuscript "Model simulations of atmospheric methane 1997-2016 and their evaluation using NOAA and AGAGE surface- and IAGOS-CARIBIC aircraft observations" by Zimmermann et al.|
The revised manuscript of Zimmermann et al. presents an updated analysis of the global CH4 budget (and trends), using now 17 global monitoring stations instead of only 6 stations in the initial discussion paper, while leaving out coastal stations (used in the initial discussion paper), which cannot be well reproduced by the EMAC model. Therefore the authors addressed the general comments (1) and (2) of my previous review.
However, the authors did not address at all my previous comment (3) and still analyze only one hypothesis for the global CH4 increase, namely the combined increase of CH4 emissions from shale gas in North America and from tropical wetlands. Given the large number of hypotheses discussed in the literature, I find it very unsatisfactory that the analysis of the paper is still limited just to this single scenario. In addition, the authors do not even make any attempt to motivate the choice of this particular scenario. While there is some evidence from several studies that agricultural CH4 emissions from ruminants have increased (a scenario which unfortunately is not further analyzed in the paper), a persistent increase of tropical wetland CH4 emissions over the whole 2007-2016 period remains very speculative. The authors only briefly state in section 4.2 (lines 399-400): "Enhanced precipitation in the regional summer season (Nisbet et al., 2016; Bergamaschi et al., 2013) may be a possible cause of growing tropical wetland emissions". However, none of the 2 cited papers analyzed in any detail the tropical precipitation patterns, nor do Zimmermann et al. present any own meteorological analyses. While some studies in the literature found some correlations between tropical wetland CH4 emissions and ENSO induced anomalies in precipitation [e.g. Pandey at al., 2017], the reported anomalies appear directly related to the ENSO patterns (with a typical duration in the order of 1 year) - but to my knowledge do not support any persistent increase over the entire 2007-2016 period.
Given the restriction of the paper to a single scenario (which - in my view - is not the most likely one) the added value of the paper to the analysis of the global CH4 trends remains very limited.
Furthermore, the presentation of the paper is not satisfactory - in the following just some examples:
(a) Introduction, page 3 (lines 84-86): "In other words, it was found by both authors that the combined fossil CH4 sources (1985-2002) must have been much stronger (factor of 2), at the expense of microbial sources." This conclusion was drawn only by Schwietzke et al. (2016), but not by Schaefer et al (2016). Then the authors continue: "Further, it was concluded that fossil fuel related sources had decreased." This was indeed concluded in both studies. However then the text continues: "Although the findings of the two articles are not necessarily in conflict...." As the studies have first been presented to give the same conclusions (which I think is not correct regarding the first statement): why should they then be in conflict?
(b) Presentation of results: The first 1.5 pages of the results section ("4 Simulation results" and "4.1 The period 1997 through 2006" largely discuss again the model setup, especially the use of the tagged tracers (repeating 3 times the finding that the sum of the tagged methane tracers is equal to the "reference tracer" with all emission sources - a finding which is basically trivial, since the system is setup as linear system).
(c) apart from the restriction of the analysis to a single scenario (see my general comment above) the discussion of this single scenario is only very limited. At the end of section 4.2.1 the authors state:
"Kirschke et al. (2013) and Turner et al. (2016), however, found that an increase by 17-22 Tg/y could explain the renewed methane growth and 30-60% of this could be attributed to increasing U.S. anthropogenic methane emissions, which supports our results with 20.70 Tg/y emission increase including 8.38 Tg/y", but do not mention that the results of Turner et al. (2016) have been questioned by Bruhwiler et al , highlighting in particular several methodical issues in the Turner et al. (2016) paper. In general the discussion of the results should be put much more in context with the existing literature.
(d) the presentation of the conclusions and outlook should better summarize the real conclusions (and limitations) of the study. E.g. the discussion of the "2nd order polynomial extrapolation predicts steady state after 13 years" seems rather hypothetical and the statement " NOAA/AGAGE station methane data are updated annually so further updates are expected" rather trivial.
A further issue of the paper is that the applied optimization technique is rather simple. In principle such simple techniques (with a very limited number of parameters) can be useful for quick analyses and for illustrative purposes. However, this should be put into context with more sophisticated inverse modelling techniques and should include a discussion of the limitations of these simple techniques. The optimization technique used in this study is rather similar to simple synthesis inversion techniques, typically used ~20 year ago (e.g. [Hein et al., 1997]) - using fixed global spatial and temporal emission distribution patterns. A very critical issue of these synthesis inversions, however, is that they are prone to the so-called aggregation error [Kaminkski et al., 2001].
Given the discussed limitations of the revised paper, I cannot recommend the paper for publication in ACP.
Bruhwiler, L. M., et al., U.S. CH4 emissions from oil and gas production: Have recent large increases been detected?, J. Geophys. Res.-Atmos., 122(7), 4070-4083, 2017.
Hein, R., P. J. Crutzen, and M. Heimann, An inverse modeling approach to investigate the global atmospheric methane cycle, Global Biogeochem. Cycles, 11, 43-76, 1997.
Kaminski, T., P. J. Rayner, M. Heimann, and I. G. Enting, On aggregation errors in atmospheric transport inversions, J. Geophys. Res., 106(D5), 4703-4715, 2001.
Pandey et al., Enhanced methane emissions from tropical wetlands during the 2011 La Niña, Scientific Reports volume 7, Article number: 45759, 2017.