Modeling biochar effects on soil organic carbon on croplands in the MIMICS (MIcrobial-MIneral Carbon Stabilization) model

Han, Mengjie; Zhao, Qing; Wang, Xili; Wang, Ying-Ping; Ciais, Philippe; Zhang, Haicheng; Goll, Daniel S.; Zhu, Lei; Zhao, Zhe; Guo, Zhixuan; Wang, Chen; Zhuang, Wei; Wu, Fengchang; Li, Wei

doi:https://doi.org/10.5194/gmd-2023-114

Preprints

https://doi.org/10.5194/gmd-2023-114

Preprints

Submitted as: model description paper

10 Jul 2023

Submitted as: model description paper |

| 10 Jul 2023

Status: this preprint is currently under review for the journal GMD.

Modeling biochar effects on soil organic carbon on croplands in the MIMICS (MIcrobial-MIneral Carbon Stabilization) model

Mengjie Han, Qing Zhao, Xili Wang, Ying-Ping Wang, Philippe Ciais, Haicheng Zhang, Daniel S. Goll, Lei Zhu, Zhe Zhao, Zhixuan Guo, Chen Wang, Wei Zhuang, Fengchang Wu, and Wei Li

Abstract. Biochar application in croplands aims to sequester carbon and improve soil quality, but its impact on soil organic carbon (SOC) dynamics is not represented in most land models used for assessing land-based climate mitigation, therefore we are unable to quantify the effect of biochar applications under different climate conditions or land management. To fill this gap, here we implemented a submodel to represent biochar into a microbial decomposition model named MIMICS (MIcrobial-MIneral Carbon Stabilization). We first calibrate MIMICS with new representations of density-dependent microbial turnover rate, adsorption of available organic carbon on mineral soil particles, and soil moisture effects on decomposition using global field measured cropland SOC at 58 sites. The calibration of MIMICS leads to an increase in explained spatial variation of SOC from 38 % in the default version to 47 %–52 % in the updated model with new representations. We further integrate biochar in MIMICS resolving its effect on microbial decomposition and SOC sorption/desorption and optimize two biochar-related parameters in these processes using 134 paired SOC measurements with and without biochar addition. The MIMICS-biochar version can generally reproduce the short-term (≤ 6 yr) and long-term (8 yr) SOC changes after adding biochar (mean addition rate: 25.6 t ha^-1) (R² = 0.65 and 0.84) with a low root mean square error (RMSE = 3.61 and 3.31 g kg^-1). Our study incorporates sorption and soil moisture processes into MIMICS and extends its capacity to simulate biochar decomposition, providing a useful tool to couple with dynamic land models to evaluate the effectiveness of biochar applications on removing CO₂ from the atmosphere.

Received: 07 Jun 2023 – Discussion started: 10 Jul 2023

Download & links

Preprint (PDF, 1657 KB)

Supplement (2899 KB)

Download & links

Mengjie Han et al.

Status: final response (author comments only)

CEC1:
'Comment on gmd-2023-114', Juan Antonio Añel, 03 Aug 2023

Dear authors,
Unfortunately, after checking your manuscript, it has come to our attention that it does not comply with our manuscript guidelines.

As the handling topical editor pointed out before, your manuscript must include the version numbers for the models used in the title (and please, along the manuscript the first time they are mentioned).

Moreover, your submission does not comply with our "Code and Data Policy".
https://www.geoscientific-model-development.net/policies/code_and_data_policy.html

Your submission does not provide the repositories for the data that you use in your work. This includes all the relevant input and output data or datasets and observations. Even the "Data Availability" section is missing in the manuscript.

I note that your manuscript should not have been accepted in Discussions, given this lack of compliance with our policy. Therefore, the current situation with your manuscript is irregular. In this way, if you do not fix these issues, we will have to reject your manuscript for publication in our journal.

Therefore, please, publish your data in one of the appropriate repositories, and reply to this comment with the relevant information (link and DOI) as soon as possible, as it should be available before the Discussions stage. Also, in your reply, include the new fixed title for the manuscript and provide details about the different versions of the models and code used to perform your work.
Then, remember to address it and include the new information in any potentially reviewed version of your manuscript.

Juan A. Añel

Geosci. Model Dev. Executive Editor

Citation: https://doi.org/10.5194/gmd-2023-114-CEC1
- AC1:
  'Reply on CEC1', Wei Li, 09 Aug 2023
  
  We thank the editor for the suggestions. We added the model version number in the title as “Modeling biochar effects on soil organic carbon on croplands in a microbial decomposition model (MIMICS-BC_v1.0)”.
  We have archived the model code and data on Zenodo using a new link (https://doi.org/10.5281/zenodo.8223088). The “Code availability” section in our manuscript was updated to “Code and data availability. The source code of MIMICS-BC_v1.0 and the data used in this study are available at https://zenodo.org/record/8223088 (last access: 8 August 2023).”
  The MIMICS-BC_v1.0 is a MIMICS-based model version for biochar (BC). In includes processes in MIMICS-TSM_b (The optimal version of MIMICS used for biochar addition, excluding biochar effects on soil organic carbon (SOC)), MIMICS-BC_D (including biochar effects on SOC by modifying deprotection rate of SOC in the MIMICS-TSM_b) and MIMICS-BC_DV (including further biochar effects on SOC by modifying the microbial maximum reaction velocity in MIMICS-TSM_b) in our manuscript (see section 2.2 and 2.3 for details). The source codes for all versions are available at https://zenodo.org/record/8223088 (last access: 8 August 2023).
  
  Citation: https://doi.org/10.5194/gmd-2023-114-AC1
  - EC1: 'Reply on AC1', Sam Rabin, 11 Aug 2023
    
    Thanks for the revisions! I'm sure that getting the above comment was stressful, so I'm sorry for missing the issues on my initial review. I wanted to post here to say for the record that the reviews should proceed as normal.
    
    Citation: https://doi.org/10.5194/gmd-2023-114-EC1
RC1: 'Comment on gmd-2023-114', Anonymous Referee #1, 07 Sep 2023

The authors adapted and modified MIMICS model to simulate cropland SOC and the effects of biochar application on cropland SOC. Although some of the processes implemented remain empirical and largely relies on parameter tuning (e.g., the nutrient enriching effect of BC application is simulated through linking BC application rate to the turnover factor rather than changing the C:N ratio of SOC), the development of MIMICS seemed reaching reasonable improvement in simulation results and could fit for the authors’ further purpose for model application.
I have just one general comment except for the following minor comments: some of the justifications on why the three aspects (MIC density-dependent process, adsorption and soil moisture) were selected for model improvement in the introduction would be better.
Line 57: “density-dependent microbial processes”: this needs some explanation. I have some basic biogeochemical background but still, I cannot guess its meaning the first time I see it.
line 165: Equation 4, the sign of square root of LIT_{tot} seems to cover 1.2. Is this intended or an error? Equation (6) might have a similar issue.
Line 175: a few words on the theoretical basis of why turnover rate is density-dependent might be helpful. Readers can then understand why, if this process is critical, it is missing from the original MIMICS version.
Line 180: how the adsorption process in original MIMICS is represented given that it already has the deprotection process being simulated? The SOC might first be adsorbed in the original MIMICS before they can be de-protected?
Line 279: why a spatial resolution of 0.5º was selected? The MIMICS model should be resolution-independent and in principle we can run for it for each site?
Line 291: Is the climate from CRU-JRA or WorldClim (line 131) being used?
Lines around 290: I understand MIMICS-TSM_b was optimized using SOC observations (considered being at equilibrium) on 58 sites. But then there are 134 paired measurements with BC application. My questions are: (1) Is MIMICS-TSM_b further optimized using the control SOC measurements from these 134 paired observations or you just directly optimize MIMICS-BC targeting only ΔSOC? (2) when optimizing for ΔSOC, how are the training and test samples being handled? (3) in evaluating for ΔSOC, did you pick up the simulated SOC corresponding to the specific years of SOC measurement in field after BC application?

Citation: https://doi.org/10.5194/gmd-2023-114-RC1
RC2:
'Comment on gmd-2023-114', William Wieder, 29 Sep 2023
Han and co-authors calibrate the MIMICS soil biogeochemical model to with the addition of density dependent processes, sorption, and soil moisture scalars and validate results with independent observations. They subsequently add a biochar parameterization to the model that is calibrated against field experiments. In the process they conducted additional experiments and sensitivity analysis.
Despite all of this work, I kind of wanted more- more than just results from a calibration and validation exercise. Why does it matter? This is a chance to show off and illustrate why calibrating a model like this matters and how the calibrated parameters alters projections of non-steady state soil C dynamics. For example, does the transient behavior of some version of the calibrated model respond to an idealized warming experiment, compared to the uncalibrated (default) parameterization? How do microbial biomass or soil respiration fluxes change with biochar additions in different versions of the MIMICS-BC calibration? These are just ideas for the authors to consider. I also appreciate you’ve already done a ton of work, and present additional suggestions to help clarify the study being presented.
I found the inconsistent organization of the paper made the work difficult to follow. Greater attention to the organization of the text will help readers follow all that’s going on with this work. For example:
The introduction claims to calibrate density dependent processes, sorption, and soil moisture scalars with data of SOC densities from 58 sites. Subsequently, it seems like different data were used to calibrate the new biochar submodule (line 98-104). This is fine, and it lays out a clear expectation for readers. What immediately follows in section 2.1, however, is a bunch of details on the biochar measurement that were collected for what I assumed was the second half of this calibration activity.

Similarly, the caption describing new features of the model in Fig 1 is reverse of how the introduction is set up.

Finally, the conclusion seemingly conflates: new processes added to the model (e.g., adsorption and soil moisture), agricultural management practices (not addressed in this paper), and long-term field experiments for biochar addition (although I'm still not really clear why this is needed based on the results presented?).

Line 261 Wow, the matrix works for non-linear models like MIMICS? This seems like it needs to be described, as it would be an important contribution, the details of which seem appropriate for a journal like GMD.

Section 3.1. I wonder if the data are sufficient for calibrating all the features for part one of the calibration (density dependent processes, sorption, and soil moisture scalars)? Declines in model performance with the validation data for the ‘optimzed model’ (MIMICS-TSMb) shown in Table S5 suggests that the model isn’t that well calibrated? I also wonder why the model with (nearly) the highest AIC values was chosen as the best?

Finally, the authors have 5 figures in the main text and 20 in the SI. This isn’t a nature paper. Would including more of the results in the main part of the manuscript clarify the experimental design (e.g. Fig S1) and results (e.g, Fig S8)?
Clarifications + Minor and technical concerns:
Line 43: This clause is incomplete, “therefore add further constraint to stabilize future warming under 2 ℃.: I’d remove it or clarify and reference what’s being stated said.

Lines 41-45. A two-sentence paragraph is pretty short. Maybe combine with other text?

Line 57. These statements seem inaccurate or misleading. Since 2015 MIMCIS included an implicit density dependent turnover term (see Wieder et al. 2015; Georgou et al. 2017, also Eq. 4) and one CN version of MIMICS does too (Kyker-Snowman et al. 2020) as well as Zhang et al., 2020. Moreover, different versions of MIMICS also include a ‘water scalar’ that reduces microbial kinetics (Wieder et al. 2019)

Section 2.1. How was depth handled for all of these different measurements? How do you handle this depth to which you’re calibrating the model?

Line 122, I might make this independent validation activity more obvious. It seems like there’s a really nice model calibration (58 sites; Fig 2) and validation (224 sites; Fig S8) for the non-biochar part of the study, but this is kind of obscure in all sections of the paper.

Line 137-141. This text seems redundant with text in the following paragraph. Can it be removed?

Line 144. Why is Fig S3 included, it seems identical to Fig 1 but with a different caption (that seems redundant with text already in the manuscript. Consider removing?
Line 169, 240, and elsewhere. Is deprotection a word? I think the MIMICS papers call in desorption.

What is the ecological or theoretical implications of modifying Km by the soil moisture term (eq 14)? I’m trying to figure out if (or why) he half saturation constant of substrates would be moisture limited or if concentrations of microbes or substrates themselves would be modified by soil water content? Some discussion of the assumptions here would be helpful.

Line 223, Table S3. This seems like a low %N for wood, but what crop inputs are included in the site-level calibration that are even providing wood litter inputs?

How are you handling wood litter inputs into MIMICS (Fig 1)?

Section 2.2.5. Does this mean that you’re making steady state assumptions about agricultural soils that they reflect present day crop productivity (less harvest removal) and litter quality estimates? This should maybe be stated clearly in the methods (e.g. line 260)?

Line 231 At first pass, this assumption of adding char directly to physically protected, available and chemical protected pools seem surprising. This is mainly because in my mind biochar has a turnover time that’s longer (~600 years, but I don’t know where this number came from) than the residence time of the default MIMICS pools, and a distinct chemical signature. Maybe provide a brief justification for the simplifying assumptions being made here? I see this is discussed at the end of section 4.2, but maybe a brief explanation is still warranted in the methods.

Ex 15 & 16 + Table S1. are the units for Fv and Fd correct here?

Section 2.3 I also am struggling to understand the assumptions of ‘negative priming’ vs. ‘positive priming’ that are going into the biochar calibrated parameterizations here.
First, I think of priming as a microbial explicit process and you’re working with a microbial explicit model- but the mechanisms for negative priming seem more to do with substrate affinity of biochar for available SOM, which is an abiotic process. Is this assumption consistent with discussion of negative priming elsewhere in the literature (e.g. Zimmerman et al 2011 paper cited earlier?). Sorry, this is literature I’m not very familiar with.

Second, if biochar is supposed to have a “strong adsorption affinity for organic matter” shouldn’t you modify some part of eq 7-9, not eq 5 as is presently done? I understand why you did it this way, eq 5 is simple to modify, whereas I’m not sure what the adsorption functions seem more challenging. Can you clarify your rationale here?

Third, “positive priming” involves increasing Vmax for all fluxes (modifying Vmax), but would MIMICS give you a positive priming effect via increases in microbial biomass just by increasing SOMa concentrations as a function of biochar additions? That is, do you need to hard wire the positive priming effects, or may they be an emergent property of the model of the existing model structure?

Line 283-305. I’m struck that the control plots (without biochar) are a really good opportunity to validate the non-char calibration that was done in step 1 of the calibration (MIMICS-TSMb). How well are the calibrated parameters from 58 locations doing at what seems like an independent set of observations? Or, I may not be understanding the experimental design accurately? Are the 387 paired measurements included in the 58 locations (e.g. 58 control / calibration plots + a bunch more BC treatment plots (section 2.1)? The introduction and table 1 makes it seem like there are an additional 134 paired measurements being used for the MIMICS-BC evaluation. This seems implied in the conclusion “We further validated MIMICS against field measurements on global croplands without… biochar addition”, but I can’t find these results presented.

Line 311, Fig S6. I’m not really clear how this double exponential model was applied in MIMICS? Or was this just used to extend the “observational record” for sites with < 8 years of data?

Why not add AIC information to Fig 2e (Similar to Fig S7)?

Why are AIC values for calibration data in Table S5 different than those reported in Fig S7?

I might suggest bringing in a clearer model validation figure into the main part of the text (e.g. Fig S8)? Or even combining these 3 datasets (rows in Fig S8) into a single set of plots for each model configuration with statistics reported as in Fig 2.

Line 342, I don’t or maize rice and wheat as cover crops. Maybe replace ‘cover crops’ with crop type?

Section 3.2 and 3.2.1 aren’t really about model evaluation, as this is where you’re calibrating the biochar module?

Section 4.1 & 4.2 Parts of this discussion are a little odd, as results from a bunch of other things the authors tried are all introduced and displayed in the SI. I appreciate including this information and analyses, I think they belong in the paper, and are appropriate for the journal. I also wonder if it’s better presented in the results, with a clearer interpretation of your findings discussed in section 4 of the paper.

Fig 5. I don’t love reporting new results in the discussion. Since this is already described in the methods (line 305-315), consider moving this text to results (e.g., 3.2.3 Sensitivity anlaysis)

447-449 This sentence is phrased in a confusing way “reduced the correlations between model-observation biases and input variables”. Clay also doesn’t need to be capitalized.

Line 460, again I don’t really think the end of the discussion is the place to add a bunch of new results from new model calibrations.

References:
Georgiou, K., Abramoff, R. Z., Harte, J., Riley, W. J., & Torn, M. S. (2017). Microbial community-level regulation explains soil carbon responses to long-term litter manipulations. Nature Communications, 8(1), 1223. doi: 10.1038/s41467-017-01116-z.

Kyker-Snowman, E., Wieder, W. R., Frey, S. D., & Grandy, A. S. (2020). Stoichiometrically coupled carbon and nitrogen cycling in the MIcrobial-MIneral Carbon Stabilization model version 1.0 (MIMICS-CN v1.0). Geoscientific Model Development, 13(9), 4413-4434. doi: 10.5194/gmd-13-4413-2020.

Wieder, W. R., Grandy, A. S., Kallenbach, C. M., Taylor, P. G., & Bonan, G. B. (2015). Representing life in the Earth system with soil microbial functional traits in the MIMICS model. Geoscientific Model Development, 8(6), 1789-1808. doi: 10.5194/gmd-8-1789-2015.

Wieder, W. R., Sulman, B. N., Hartman, M. D., Koven, C. D., & Bradford, M. A. (2019). Arctic Soil Governs Whether Climate Change Drives Global Losses or Gains in Soil Carbon. Geophysical Research Letters, 46(24), 14486-14495. doi: 10.1029/2019gl085543.
Citation: https://doi.org/10.5194/gmd-2023-114-RC2

Mengjie Han et al.

Supplement

https://doi.org/10.5194/gmd-2023-114-supplement

Mengjie Han et al.

Viewed

Total article views: 362 (including HTML, PDF, and XML)

HTML	PDF	XML	Total	Supplement	BibTeX	EndNote
252	92	18	362	29	4	6

HTML: 252
PDF: 92
XML: 18
Total: 362
Supplement: 29
BibTeX: 4
EndNote: 6

Views and downloads (calculated since 10 Jul 2023)

Month	HTML	PDF	XML	Total
Jul 2023	117	41	4	162
Aug 2023	99	29	8	136
Sep 2023	34	18	6	58
Oct 2023	2	4	0	6

Cumulative views and downloads (calculated since 10 Jul 2023)

Month	HTML	PDF	XML	Total
Jul 2023	117	41	4	162
Aug 2023	99	29	8	136
Sep 2023	34	18	6	58
Oct 2023	2	4	0	6

Viewed (geographical distribution)

Total article views: 343 (including HTML, PDF, and XML) Thereof 343 with geography defined and 0 with unknown origin.

Country	#	Views	%

Latest update: 03 Oct 2023

Short summary

Because the impact of biochar on SOC dynamics is not represented in most land carbon models used for assessing land-based climate mitigation, our study develops a biochar model that incorporates our current understanding of biochar effects on SOC based on a soil carbon model (MIMICS) for future predictions at the regional and global scale. The MIMICS model was improved with new processes considered and the developed biochar model can generally reproduce the SOC changes after adding biochar.


Total:	0
HTML:	0
PDF:	0
XML:	0