the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 10 Feb 2026
Interactive Simulation of Methane and Hydrogen Soil Deposition in ECHAM5/MESSy Atmospheric Chemistry Model (EMAC) v2.55 with the new Submodel BIODEP (v1.0)
Abstract. Methane (CH4) and hydrogen (H2) play critical roles in atmospheric chemistry and climate processes. CH4 is a powerful greenhouse gas, whereas H2, although not a greenhouse gas itself, indirectly affects radiative forcing by modifying the atmosphere's oxidative capacity and therefore the concentrations of CH4, ozone (O3) and stratospheric water vapor. Hydrogen is predominantly removed through microbial uptake in soils, while approximately 6 % of CH4 is taken up by soils, a factor that contributes significantly to its overall atmospheric budget. Soil uptake depends on various soil characteristics, including type, temperature, moisture, and for CH4, nitrogen deposition. Accurately representing these influences requires a detailed understanding of both atmospheric conditions and land surface and hydrological properties. However, many Earth system models currently use fixed soil deposition rates for H2 and CH4, without accounting for variations in soil properties. We present BIODEP, a new biogenic deposition submodel that has been integrated into the ECHAM/MESSy Atmospheric Chemistry model (EMAC). BIODEP dynamically simulates the uptake of CH4 and H2 by soil, based on local meteorological and soil conditions. With BIODEP, the soil sinks of CH4 and H2 are updated online based on the meteorological conditions, atmospheric composition, and land surface properties provided by the EMAC model. The EMAC model is coupled to the JSBACH land surface and vegetation model. This allows for a consistent and interactive treatment of soil sinks within the atmospheric chemistry model. Modeled global mean soil uptakes of 62.7 ± 11.7 Tg yr−1 for H2 and 30.2 ± 4.8 Tg yr−1 for CH4 are consistent with previous studies, and the resulting atmospheric mixing ratios show good agreement with observations from the NOAA GML Carbon Cycle Cooperative Global Air Sampling Network, evaluated over the period 2009–2019. In addition, comparison with column-averaged CH4 (XCH4) observations from the Greenhouse Gases Observing Satellite (GOSAT) demonstrates that EMAC reproduces the global and zonal-scale methane distribution with small mean biases, providing independent support for the accuracy of the simulated soil methane sink. This development makes EMAC a state-of-the-art model to interactively simulate atmospheric chemistry, including both the soil sinks of CH4 and H2. This enables more consistent simulation of trace gas budgets and an improved assessment of the feedbacks between land surface processes, atmospheric composition, and future climate and emission scenarios.
Competing interests: At least one of the (co-)authors is a member of the editorial board of Geoscientific Model Development.
Publisher's note: Copernicus Publications remains neutral with regard to jurisdictional claims made in the text, published maps, institutional affiliations, or any other geographical representation in this paper. While Copernicus Publications makes every effort to include appropriate place names, the final responsibility lies with the authors. Views expressed in the text are those of the authors and do not necessarily reflect the views of the publisher.- Preprint
(13031 KB) - Metadata XML
- BibTeX
- EndNote
Status: final response (author comments only)
- RC1: 'Comment on egusphere-2026-371', Anonymous Referee #1, 23 Mar 2026
-
RC2: 'Comment on egusphere-2026-371', Anonymous Referee #2, 30 Mar 2026
This manuscript presents results from a new configuration of EMAC in which previously developed deposition parameterizations for H2 and CH4 are driven by temperature, moisture, and N from a comprehensive land model (rather than reanalysis). The approach is interesting. Unfortunately, the resulting simulation of H2 exhibits significant biases in both seasonality and concentration in NH, i.e., where the soil sink is dominant.
Given that this is the primary focus of this study, additional analysis (and very likely additional development) are clearly needed before this paper can be accepted for publication.
Major comments1. It would be helpful for the authors to describe similarities/differences in the biological processes that control the CH4 and H2 soil sink. I understand the strong link between CH4 oxidation by OH and H2 source, but the link between their soil sinks is not clear
2. Given the importance of soil moisture, has there been assessments of the soil model performance for this metrics?
3. The general shape for r_sm and r_t for H2 and CH4 seem quite similar (H2 should be added to Fig. 1) but equations are quite different. Does this reflect differences in the underlying biological processes? If not, why aren't the same functional forms used for both?
5. Since the model is driven by climatological emissions. It's unclear why the authors focus on the correlation between observed and model time series. It seems that it would be more appropriate to focus on the seasonality.
6. Can the authors describe how the new representation of CH4 soil sink improves the simulation of CH4 relative to previous versions of the EMAC model (I assume that the soil sink was not represented). It's not clear in the current manuscript.
7. The model exhibits very significant biases wrt the seasonality and meridional gradient of H2.
Biases wrt seasonality are found in many models (see analysis by Paulot (2024); Tardito Chaudhri (2025)) but it seems worse here and deserves further analysis.
Models of varying complexity have also been able to capture the meridional gradient in H2 well for 10+ years (see Price et al.) including EMAC with ERA5 soil moisture. This model does not capture the H2 gradient well as shown in Fig. 7b. Again more analysis are clearly needed to understand and hopefully address this large bias. I am concerned that the core issue (as alluded by the authors) may be the simulated soil moisture. However, if it's the case, this begs the questions whether this configuration of the model is well-suited to represent H2 (including analysis 4.3).
The authors should also comment on why biases in soil moisture are less important for CH4?
8. There are observations of H2 deposition velocity. The authors should show how the simulated vd(H2) compares against these observational estimates (see Brown et al. (2026) for instance)Technical comments
1. Line 14: Which time-period does this correspond to?
2. Line 40 and elsewhere. The number of significant figures does not reflect the large underlying uncertainty
3. Line 44 Is the H2 sink also sensitive to the N deposition. If not, I would suggest rephrasing for clarity.
4. Line 63. Please state the years that are considered explicitly
5. Line 73. COVID in 2019?
6. Line 83. Subsubsection-> Subsection
7. Line 96. Specify which variables you are nudging (wind, temperature, ... )
8. Line 125. The RETRO dataset seems quite dated considering that the authors focus on 2009-2019 period. Why not use anthropogenic CO emissions to estimate H2 emissions (see Paulot et al. (2021,2024), Brown et al. 2025). How does it compare with these more recent inventories. It's not clear why the authors use a climatology rather than time-varying emissions?
9. Use consistent notation. For instance, on line 155, you use $\Psi$ to designate the volumetric water content but on line 175 you use SM.
10. There is no reference to Fig. 1. Please use it to support your discussion of r_sm, ...
11. Line 216. The authors need to provide the equations used to represent H2 uptake. I understand that they are introduced elsewhere but this is also true for CH4.
12. Line 220. "The water content and thickness of the top layer are calculated from porosity, total column water content and
empirical thresholds depending on the eolian and sand and loess loam fractions (Surawski et al., 2025) while these fractions
are taken from the LDAS/GLDAS data sets (Rodell et al., 2004), porosity and total water column content are derived from the
optimized soil hydrology in EMAC, provided by the JSBACH submodel. "I don't understand this sentence (I suspect it's missing a period). I also don't understand why you are not using the eolian, sand, and loess loam fractions (line 235) from JSBACH. This treatment seems to be introducting an uncessary inconsistency. Further, are the GLDAS and Shangguan datasets for f_clay, ... the same?
12. Table 3 needs to be revised.
a. Please use "surface mixing ratio" rather than atmospheric burden at surface. I don't think it is reported in Petron (2024) for instance.
b. I don't understand why the only comparison for vd(H2) is against Surawsk et al. (2025). This is not meaningful since the authors tuned their model to match this estimate.
c. Detailed budgets for H2 and CH4 have been published in recent years and should be included in this table (see Brown et al. (2026) and references therein for instance)13. Fig. 3. It seems that all observations have correlation >0. Can the colorbase be adjusted accordingly. Please provide correlation and information about the model bias for the top panel (scatter plot in SI would be helpful)
13. Fig. 6. Give reference for fitting strategy.
14. Line 324 "... annual variability (see Figure B2), which suggests that the EMAC model with BIODEP accurately represents soil sink dynamics in at remote stations"
H2 variability at these stations is not driven by the colocated vd(H2) and it's not clear how you are drawing this conclusion.
15. Line 346. 235ppv? Seems like a typo.Citation: https://doi.org/10.5194/egusphere-2026-371-RC2
Viewed
| HTML | XML | Total | BibTeX | EndNote | |
|---|---|---|---|---|---|
| 200 | 120 | 16 | 336 | 41 | 51 |
- HTML: 200
- PDF: 120
- XML: 16
- Total: 336
- BibTeX: 41
- EndNote: 51
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1
Review of “Interactive simulation of methane and hydrogen soil deposition in ECHAM5/MESSy Atmospheric Chemistry Model (EMAC) v2.55 with the new submodel BIODEP (v1.0), by Anna Martin et al.
General comments
Results from a new global model which simulates the atmospheric chemistry and land surface fluxes of hydrogen and methane is presented. Some of the equations used in the model are unclear (in particular the units used – see specific comments below). The model does a reasonable job at simulating the global distribution of H2 and CH4, but its seasonality, especially in the Northern Hemisphere, appears less well simulated (although the authors claim otherwise). Although the soil deposition schemes for the two gases are described in some detail, the other main determinant of seasonality (reaction with OH in the atmosphere) is not discussed, and it seems that the relative strengths of these two sinks is perhaps incorrectly described by the model. I think this needs to be investigated before the paper can be accepted.
If the equations describing the model can be clarified, and some extra discussion about the atmospheric sinks, and their interplay with the soil sinks, can be included then this paper should be suitable for publication.
Specific comments
l40 H2 sources (not emissions). Both % values are over-precise, given the uncertainties.
l58 How does the interactive soil moisture compare to ERA5 (and reality)?
P3 I don’t think Table 1 is referred to anywhere in the text.
I’m a bit confused how the COVID-19 pandemic resulted in data gaps before 2020 – but may it is possible if it disrupted data analysis of data collected earlier? Please clarify.
l120 I don’t understand the scaling applied to CH4 emissions – please clarify.
l124 of the Appendix
P6 Equations 1-3 and surrounding text: Units are only partially given (e.g., what are the units of A, B and J?). The units of eq.2 mixes ppbv and mg m-3. Please clarify and make these equations dimensionally balance.
l147 Clarify that GT is (presumably) also dimensionless.
P6 Equation 6: The volumetric soil water must be a fraction, not a %? Please clarify.
l160 Presumably these scaling factors are also dimensionless.
L166 values
L170 Suggest clarify that you mean soil oxidation rate (as opposed to atmospheric oxidation rate, which is more commonly discussed for methane).
L174 Suggest state that the scheme yields an optimum soil moisture content of 0.2.
P8 Equation 12 isn’t dimensionally balanced. Please clarify.
P9 Figure 1b: I was expecting the soil response factor rT to vary between 0-1 (like rSM and rN; and based on equation 7), but it varies from 0-3.5. Please explain.
P9 Figure 1 caption: I am unsure what the “initial bulk density” of 14.7 kg m-3 is, please clarify.
L221 Aeolian.
L228 rate -> velocity
L228 The value in Figure 2 is 0.032 cm, not 0.033 cm
P11 Table 3. I don’t understand what is meant by atmospheric burden “at surface” and “total column”. Surely the atmospheric burden is just the total mass in the atmosphere, and can’t be qualified like that?
L266 4a should be either 3a or 4b?
L269 4b should be 4a?
L270 I disagree that the CH4 seasonal cycle “shows generally good agreement”. In Figure 4a, observations show a summer minimum in the Northern Hemisphere, whilst the model has a late summer maximum. The Southern Hemisphere seasonal cycle is not very clear from Figure 4a (partly due to the poor choice of scale), but is also not great.
I found it surprising that the discussion around Figure 4 and the seasonal cycle didn’t include mention of the atmospheric oxidation of CH4 by OH as a significant driver. Whether atmospheric oxidation or soil uptake dominates the seasonality is a key question, and this probably varies with sites. I’d expect there to be some difference between hemispheres, as there is much less land in the SH, so it is more likely to be more dominated by the atmospheric OH sink. It looks to me that your model simulations do not have quite the right balance between these two drivers of seasonality in CH4.
L273 ‘likely’ – you could check whether outlier stations are having an influence – then you would know one way or the other.
P14 Figure 4a: You have not chosen a good scale to clearly illustrate the seasonal cycle.
P14 Figure 4: Are all the plots averages over 2009-2019?
P14 Figure 4b: Would be clearer if you used different shaped symbols for the model and observations (e.g., cross and circle).
P18 Figure 7: Similarly to CH4, the modelled seasonal cycle of H2 in the NH is nowhere near in phase with the measurements, suggesting something is not right. Again, there is a balance between soil deposition and reaction with OH, but for H2, the soil sink is now dominant. Tardito Chaudhri and Stevenson (2025: https://acp.copernicus.org/articles/25/7369/2025) have also looked at this.
L381 “the model captures the general shape of the observed seasonal cycles”. Maybe it gets the shape, but the phase is shifted several months.