Interactive comment on “ Barriers to predicting changes in global terrestrial methane fluxes : analyses using CLM 4 Me , a methane biogeochemistry model integrated in CESM ”

I have some concerns about the model being over parameterized with many components either difficult or even impossible to verify with real world measurements. I also feel that some of the complexity of the model was poorly proportioned (e.g. large effort expended on various transport mechanisms of methane from the site of formation to the atmosphere, each mechanism with a high uncertainty about how well these are performing; but no prognostic wetland finding scheme or wetland PFTs). However, with those stated concerns, I find the authors did a commendable job of stating the model weaknesses and exploring the model with sensitivity analyses. The authors are also very clear about the problems with such a high complexity model, i.e. over parameterization. While I am not fully convinced that their model is able to truly represent wetland methane dynamics (given the level of parameterization), it does present a worthwhile step forward and also advances our understanding of the strengths and weaknesses of high-complexity wetland CH4 modelling. Given the authors careful dissection of the model and openness to explain its weaknesses, I can find no major faults with the paper. I recommend publication with only technical revisions.

based upon several previous models (including [Walter et al., 2001;Wania et al., 2010] [ Zhuang et al., 2004]) and some interesting innovations. The model is evaluated over the modern period and is used for projections up to year 2100.
I have some concerns about the model being over parameterized with many components either difficult or even impossible to verify with real world measurements. I also feel that some of the complexity of the model was poorly proportioned (e.g. large effort expended on various transport mechanisms of methane from the site of formation to the atmosphere, each mechanism with a high uncertainty about how well these are performing; but no prognostic wetland finding scheme or wetland PFTs). However, with those stated concerns, I find the authors did a commendable job of stating the model weaknesses and exploring the model with sensitivity analyses. The authors are also very clear about the problems with such a high complexity model, i.e. over parameterization. While I am not fully convinced that their model is able to truly represent wetland methane dynamics (given the level of parameterization), it does present a worthwhile step forward and also advances our understanding of the strengths and weaknesses of high-complexity wetland CH4 modelling. Given the authors careful dissection of the model and openness to explain its weaknesses, I can find no major faults with the paper. I recommend publication with only technical revisions.
In general the writing and layout of the paper is very good. The paper is very long and ideally would be shorter. However, the paper does present a lot of material and is not repetitive.
Specific comments: abstract L18: I think you don't intend the 'm-2'. p1735 L6: Put the sources in the same order as they are of magnitude of flux.
L18: Anoxia? Anoxia is usually important for methane production... Christensen [Christensen et al., 1996] using a simple wetland finding approach p1737 L4: If this high latitude band was not defined the same as your band in the paper, please state it so the reader can compare.
L14: This source is not confined to trees. Also there has been numerous papers [Ferretti et al., 2006;Nisbet et al., 2009] of late that suggest this source is much less than the original paper [Keppler et al., 2006] suggested. Those should be referenced for an estimate instead. P1763 : Given the major problems with using site scale measurements to evaluate a large grid model, why did you not try air mass back trajectory [Worthy et al., 2000] or a more regional satellite and site measurement approach [Melack et al., 2004]? Basically, I think you should have tried for an in-between of the satellite/inverse-models approach and the site-level measurements.
L19: These sites are all high-latitude, were these chosen to minimize the problems CLM4 has with low latitude NPP? P1765 L17: Given how uncertain your model results are, and the high level of parameterization, I am not sure this would be an improvement for the inversions.
P1779 L16: We seem to be missing part of this line? The sentence does not make sense.
P1780 L12: I do not understand this annual average seasonal inundation factor. What does a value of 0.95 mean? Please provide a better description of this factor.