Accounting for empirical global soil organic characteristics and moisture heterogeneities in soil organic decomposition scheme of land surface models

Salmon, Elodie; Guenet, Bertrand; Ducharne, Agnès

doi:10.5194/egusphere-2025-3511

Preprints

https://doi.org/10.5194/egusphere-2025-3511

Preprints

25 Jul 2025

| 25 Jul 2025

Accounting for empirical global soil organic characteristics and moisture heterogeneities in soil organic decomposition scheme of land surface models

Elodie Salmon, Bertrand Guenet, and Agnès Ducharne

Abstract. Below ground soil organic carbon (OC) decomposition is governed by several biophysical drivers, causing difficulties to accurately capture the spatial patterns of soil OC stock and of CO₂ flux in Earth System models (ESMs). These biophysical drivers influence soil OC decomposition due to the respiration of heterotrophic organisms. Formulation in global scale process-based models of these processes consist of functions that modify the soil OC decay rate and therefore the soil heterotrophic respiration (HR) which modify global soil OC stock estimated by models. Current soil HR modifiers employed in models are a single relationship between soil moisture and the rate of decomposition that are employed for all the ecosystem types. Observational database meta-analysis relationships of SOIL MOISTURE and soil HR has been established considering observed soil physical properties. These relationships serve to define an empirical model that consists of a collection of different relationships based on soil organic carbon content, clay fraction and bulk density in order to uniquely substitute SOIL MOISTURE control on soil HR with a function modifier that reflects soil HR spatial heterogeneity.

In the present study, this empirical model has been embedded in the land surface model Organising Carbon and Hydrology In Dynamic Ecosystems (ORCHIDEE). The effect of the multivariate approach on simulation results has been assessed on soil OC stock and soil HR estimations at global scale. Results show that global soil OC stocks are nearly doubled in the modified model version, which is closer to observations-based products compared with the initial version, while CO₂ emissions, due to soil HR, are unchanged. The latitudinal soil OC distribution is maintained, displaying as much soil OC stock in tropical regions as under higher latitudes. This study demonstrates the significance of secondary drivers in the relationship between SOIL MOISTURE and the soil HR response to enable accounting for soil OC stock and CO₂ fluxes heterogeneous spatial pattern.

Received: 21 Jul 2025 – Discussion started: 25 Jul 2025

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.

Download & links

Elodie Salmon, Bertrand Guenet, and Agnès Ducharne

Status: final response (author comments only)

RC1:
'Comment on egusphere-2025-3511', Anonymous Referee #1, 28 Oct 2025
The authors identified the clear lack in current LSMs and ESMs of using monotonically increasing CMF-functions, although a hill-shaped function is closer to reality by modeling anoxic effects. The proposed integration of a multivariate function by Moyano et. al is clearly a mechanistic improvement, which is argumentatively well explained. Anyway could these advances in the evaluation not clearly be supported, which could maybe done with further evaluation. I propose to consider the manuscript for publication in ESD after major revisions addressing the points below. The additional evaluation requested may, however, reveal structural issues or inconsistencies in the results, which could ultimately justify rejection if major flaws are identified.

Major revisions:
Assumptions on the effect of wetlands in the evaluation products are plausible, but with masking those and the ORCHIDEE(-M) predictions these assumptions should be tested. As there is overestimation of stocks in tropics by the M-version, this downside should be evaluated further, which can only be done by masking. Afterwards should the actual spatial correlation be additionally plotted.

All evaluation products are themselves predictions/generalizations of ground truth measurements. Evaluation should additionally be done on the ground truth dataset behind the products.

As ORCHIDEE(-M) produces subdaily predictions and the effect of moisture on HR can assumed to be a very dynamic one, additional evaluation with subdaily products as (e.g.) Fluxcom Reco could reveal additional insights and integrate additionally tower measurements into evaluation.

The M-version is supposed to model also anoxic conditions. It should be further elucidated, why wetlands are assumed to still be missed.

Minor revisions:
15: actually there are additional HR modifiers.

48: residuals?

51: I would say the moisture product forces, while constraining would mean a traget used to optimized parameters.

58: Q10 formulation itself is not hill-shaped and conceptually the same as Arrhenius.

60-65: I had real difficulties in understanding this paragraph's message.

68: Linear means here factor of 1? Why do ratios lower than 1 indicate non-linear relationships?

77-79: I did not understand the meaning of the last part of the sentence.

90: I would have needed a short introduction into the parameter PRSR.

91-93: I think the order of the sentence is not correct.

100, 114: Texture is a control in the standard version already. It would be interesting to discuss why this is not enough.

109: The moisture could be below 25% but the relationship is constant below 25%.

123-124: I could not identify where the linear regression models are coming from.

134: Confusion about the indices n=1 in brackets and then n-1=0.

135: The mentioned subtraction is not visible in the function above.

Fig. 1: Organic rich is difficult to distinguish from mineral soil values. Furthermore would it be interesting to see the effects of clay vs. OC in the mineral function.

161: If SoilGrids v2.0 was used, then Poggio not Hengl was first author. If not, why was not v2.0 used? Does SoilGrids (v2.0) claim to model peatlands correct?

183: distribution not dispersal.

186: The description of CMF strength is confusing; low CMF implies strong limitation on HR.

193: ‘negative values lead to higher values’ is a little counterintuitive.

Fig. 2: Mention that mean annual values are indicated here.

241: OC stock rate should be stock I guess.

308: Table 2 & 3 would be much clearer if PFT’s would directly be written out. Abbreviations are not necessary.
Citation: https://doi.org/10.5194/egusphere-2025-3511-RC1
- AC1: 'Reply on RC1', Elodie Salmon, 30 Jan 2026
  
  The comment was uploaded in the form of a supplement: https://egusphere.copernicus.org/preprints/2025/egusphere-2025-3511/egusphere-2025-3511-AC1-supplement.pdf
  
  Citation: https://doi.org/10.5194/egusphere-2025-3511-AC1
RC2:
'Comment on egusphere-2025-3511', Anonymous Referee #2, 13 Nov 2025

This study reports an improvement to a carbon model, ORCHIDEE, by replacing the water function in the soil process, and examines the impact of this modification. The authors demonstrate how this improvement affects the estimation of soil carbon and respiration by evaluating the spatial distributions and comparing model outputs with data-driven products. The study shows how improving the soil water function in decomposition processes enhances model performance. However, the manuscript is not easily readable for many readers due to a lack of detailed definitions of variables and functions. As it stands, the manuscript is difficult to follow for those who are not familiar with the Moyano function and the ORCHIDEE model. In addition, I am not sure whether soil moisture in the Moyano function and that in ORCHIDEE are conceptually and definitionally the same.

Major comments
1 Definition and depth of soil moisture
I suspect that the definition of soil moisture in Moyano and ORCHIDEE are not the same. I could not find detailed definitions in the manuscript. This issue is fundamental for incorporating the Moyano function.

2 Soil depth for ORCHIDEE SOC: the assumption of "soil OC content accumulated up to 1 m depth"
In this study, it is assumed that soil OC content is accumulated up to 1 m depth in the model (although officially there is no explicit depth definition in the model). If the SOC functions and decomposition constants in the model are based on CENTURY, the effective soil depth for SOC would be much shallower (around 20-30 cm), because CENTURY was originally developed for agricultural lands. Indeed, Figure 4 shows that the observed SOC (0-100 cm) is much higher than the modeled values. I agree that some assumption is needed to compare model outputs with observational data, but is the 1-m assumption appropriate? Would 0-30 cm be more appropriate? More justification and discussion are needed for this 1-m assumption.

3 Inconsistent and unclear variable definitions
Terms such as litter moisture, soil moisture, organic-rich soil, water saturation, relative soil respiration, control moisture function, and control moisture values are not consistently defined. This inconsistency makes it difficult to follow the manuscript. A thorough revision is needed to ensure clarity and consistency.

4 Soil moisture function descriptions
The descriptions and definitions of the soil moisture functions are also difficult to follow and should be clarified.

Specific comments:
Title / Line 81: "Soil organic characteristics" - this phrase is confusing. Do you mean "soil characteristics" or "organic matter characteristics"? Is "soil organic characteristics" common in this field?

Line 17 / Abstract: "SOIL MOISTURE" - why is this in capital letters? RH is defined, but soil moisture is not.

Line 32-33: "Equilibrium" - this word implies stability or balance, but soil carbon is not at equilibrium. Please revise to avoid misunderstanding.

Line 91: This sentence may need a reference.

Line 109: “theta being the soil moisture in m3/m3 and ranges between 0.25 and 1." - There are many indices representing soil moisture in soil science; please define clearly. Also, why is the minimum value 0.25?

Line 112: "Soil moisture in fraction of saturation" - unclear, please define.

Line 113: "Interval of 0.01" - unclear. Soil moisture is continuous, so why calculate at 0.01 intervals?

Line 119: "Soil height" sounds odd; consider "soil depth" or "soil thickness."

Line 120: "Gridded dependent" - unclear.

Line 122-127: "Maintain integrity" - I cannot follow this process; please clarify.

Line 132-137: Definition of the function - I cannot follow this equation. (1) Why interval? (2) Why is the value for interval n a function of interval n-1? (3) Why is SRini necessary to make a moisture function with a maximum of 1?

Line 166-180: These paragraphs seem to belong to the Methods section.

Figure 3: Litter moisture and soil moisture are not defined. In Line 113, "organic richer soil" is mentioned - does this refer to the same thing?

Line 263: "Yedoma" - unfamiliar term; please define.

Figure 4 / Line 245-247: (1 m depth) - see major comment above; the SOC in ORCHIDEE seems equivalent to a shallower depth.

Line 320-321: Modeled RH vs. data-driven RH - the estimates are similar (within 50-53), so I would not say modeled RH is higher than observed, except for Koning. I would describe them as similar in magnitude.

Line 380-388: RH did not change; NPP is important - more explanation is needed. If moisture function values differ, RH should differ. However, because the model runs to equilibrium, are the RH differences offset by different soil carbon stocks (as RH sources)? Is NPP equivalent to RH? This needs more discussion.

Citation: https://doi.org/10.5194/egusphere-2025-3511-RC2
- AC2: 'Reply on RC2', Elodie Salmon, 30 Jan 2026
  
  The comment was uploaded in the form of a supplement: https://egusphere.copernicus.org/preprints/2025/egusphere-2025-3511/egusphere-2025-3511-AC2-supplement.pdf
  
  Citation: https://doi.org/10.5194/egusphere-2025-3511-AC2

Elodie Salmon, Bertrand Guenet, and Agnès Ducharne

Viewed

Total article views: 1,196 (including HTML, PDF, and XML)

HTML	PDF	XML	Total	BibTeX	EndNote
1,041	123	32	1,196	40	45

HTML: 1,041
PDF: 123
XML: 32
Total: 1,196
BibTeX: 40
EndNote: 45

Views and downloads (calculated since 25 Jul 2025)

Month	HTML	PDF	XML	Total
Jul 2025	38	11	0	49
Aug 2025	172	11	7	190
Sep 2025	666	6	2	674
Oct 2025	45	19	5	69
Nov 2025	42	18	6	66
Dec 2025	24	27	4	55
Jan 2026	47	30	8	85
Feb 2026	7	1	0	8

Cumulative views and downloads (calculated since 25 Jul 2025)

Month	HTML	PDF	XML	Total
Jul 2025	38	11	0	49
Aug 2025	172	11	7	190
Sep 2025	666	6	2	674
Oct 2025	45	19	5	69
Nov 2025	42	18	6	66
Dec 2025	24	27	4	55
Jan 2026	47	30	8	85
Feb 2026	7	1	0	8

Viewed (geographical distribution)

Total article views: 1,191 (including HTML, PDF, and XML) Thereof 1,191 with geography defined and 0 with unknown origin.

Country	#	Views	%

Latest update: 05 Feb 2026

Short summary

Soil organic carbon stockage is a key process to mitigate climate change and is intertwined with soil temperature and moisture and of other secondary soil properties. This study shows the significance of secondary drivers in the relationship between soil moisture and microbial efficiency in soil organic carbon degradation. Using empirical relationships in a global ecosystem model enhanced significantly the heterogeneous spatial pattern of soil organic carbon stock and carbon dioxide fluxes.


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