the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 18 Aug 2025
Uncertainty in Land Carbon Fluxes Simulated by CMIP6 Models from Treatments of Crop Distributions and Photosynthetic Pathways
Abstract. A reliable representation of the diversity of vegetation in terrestrial ecosystems is needed for the accurate simulation of present and future biogeochemical cycling and global climate, particularly as climate change affect different vegetation types differently. We compare the distributions of crops and of C3 versus C4 photosynthetic pathways in both natural vegetation and crops across Earth System Models in the 6th Coupled Model Intercomparison Project (CMIP6). We find a large range in vegetation type for area, gross primary production (GPP) and carbon stock change in both natural vegetation and croplands across the models. Even though 10 of 11 models used Land Use Harmonization (LUH2) crop areas as input data, modeled total crop area ranges from -28 to +10 % of the data-based estimate. The C3 and C4 crop areas were -56 to +15 % and -100 to +38 % of LUH2 for 2014, respectively. The C4 fraction of total vegetation area in the models is 9–25 %, compared to 17 % in observation-based estimates. Total global GPP varies by a factor of two across the models, and the C4 fraction of GPP ranges from 12 to 27 %. Simulated trends in the fraction of GPP by C3 versus C4 vegetation type (-20 to +29 %) would have changed global isotopic discrimination by -0.35 to +0.11 ‰ over 1975–2005, indicating that modeled changes in vegetation type do not account for the +0.7 ‰ increase indicated by atmospheric data. Disparity in vegetation types contributes to uncertainty in land carbon fluxes and further constraints and improvements in models are needed.
-
Notice on discussion status
The requested preprint has a corresponding peer-reviewed final revised paper. You are encouraged to refer to the final revised version.
-
Preprint
(1975 KB)
-
Supplement
(1394 KB)
-
The requested preprint has a corresponding peer-reviewed final revised paper. You are encouraged to refer to the final revised version.
- Preprint
(1975 KB) - Metadata XML
-
Supplement
(1394 KB) - BibTeX
- EndNote
- Final revised paper
Journal article(s) based on this preprint
Interactive discussion
Status: closed
-
RC1: 'Comment on egusphere-2025-3785', Anonymous Referee #1, 22 Sep 2025
The comment was uploaded in the form of a supplement: https://egusphere.copernicus.org/preprints/2025/egusphere-2025-3785/egusphere-2025-3785-RC1-supplement.pdfCitation: https://doi.org/
10.5194/egusphere-2025-3785-RC1 - AC1: 'Reply on RC1', Joseph Ovwemuvwose, 19 Nov 2025
- AC3: 'Reply on RC1: This the second response to the latest comment by referee 1', Joseph Ovwemuvwose, 04 Jun 2026
-
RC2: 'Comment on egusphere-2025-3785', Anonymous Referee #2, 22 Sep 2025
In this manuscript, Ovwemuvwose et al. examined the C3 and C4 fractions in global vegetation cover, gross primary productivity (GPP), and vegetation carbon stocks (Cveg) across 11 CMIP6 models. They report substantial differences among models in these C3/C4 fractions, as well as in their long-term trends. Those trends also do not align well with observations—for instance, the C3/C4 fraction in GPP and the global decline rate in δ¹³C discrimination. I read the paper with great interest because it addresses an important and often overlooked source of inter-model spread in CMIP6. However, several key issues prevent me from supporting publication at this stage.
1. Core assumptions about C3/C4 GPP and Cveg. CMIP6 models do not directly report GPP or Cveg partitioned into C3 and C4 components. The authors therefore estimated C3 and C4 values by multiplying total GPP or Cveg by the areal fraction of C3 or C4 vegetation. As they themselves noted, this is problematic: on average, GPP per unit area is higher for C4 than for C3 vegetation, whereas Cveg per unit area is likely greater for C3 plants (often woody) than for predominantly herbaceous C4 plants. These relationships also vary geographically, shaped by climate and soil. Consequently, it is difficult to place much confidence in downstream analyses of C3/C4 GPP, Cveg, and their trends.
A similar concern applies to the treatment of C4 crops. Several CMIP6 models lack explicit fractions for C4 crops versus natural C4 vegetation, so the authors assume C4 crop fraction = C4 fraction × total crop fraction. This implies that C3/C4 ratios for crops mirror those of natural vegetation at each grid cell—an unlikely scenario. Open data (e.g., Luo et al., 2024) show no such pixel-level pattern.
The LUH C4-crop fraction itself is also problematic: in LUH c4 crop fraction remains static (based on ~2000 data), so the apparent change in C4 crop area likely reflects total cropland expansion or contraction. It is unclear how individual CMIP models treat this issue, and I would not regard the LUH C4 crop trend as an “observation,” nor use it to declare a model (e.g., UKESM1) incorrect (L155, L328).
If the aim is a robust model–data comparison of C3/C4 fractions in area, GPP, and Cveg, I suggest considering the TRENDY DGVM ensemble, which may provide explicit estimates for these variables.
2. Writing quality and over-generalisation. Although I usually refrain from commenting on style, some issues affect the paper’s scientific clarity. I gave a few examples below
Carelessness: L137: “Click or tab here to enter text” is clearly a placeholder. The manuscript contains two “Discussion” sections (4 and 6); Section 6 seems intended as a Conclusion. Please proofread carefully. More examples in minor comments below.
Overstatements:The manuscript repeatedly describes the work as analysing “vegetation diversity” or “vegetation types” (L11, L16, L19, L74, L215, L275). In fact, it focuses on the photosynthetic pathway (C3 vs C4). Terms such as diversity or even vegetation types/PFTs carry broader meaning and could mislead readers. Being precise will not diminish the study’s novelty.
Other comments:
L31: “Land Cover and Land Use Change (LUCC)” is a more standard term.
L38: Molotoks et al.—only one reference is cited; the “a” after 2018 is unnecessary (cf. L286).
L43–44: The description of C3 vs C4 lacks detail on anatomical/structural differences that drive their distinct climate responses—important for motivating later arguments.
L70: The “3” in C3 should be subscript.
L125: You did not describe how C3/C4 Cveg was estimated; I assume you applied the same method as for GPP.
L139: Define Δtot before using it.
L140: Too many simplifying assumptions undermine the credibility of later detailed analyses; the variation in Δ is particularly large.
L186: Please clarify the source of the 10–45 % cropland figures.
L201: “UKESM is quite consistent”—please explain.L224: I remain unconvinced by the C4 GPP and Cveg values.
L315: Instead of stating that COS measurements prove models underestimate CO₂ fertilisation, consider a more cautious phrasing, e.g. “we noticed there is still difference between models and COS on the quantification of eCO2 effect and highlight the uncertainty in future 13C discrimination rate as eCO2 benefits C3 more than C4…”
Citation: https://doi.org/10.5194/egusphere-2025-3785-RC2 - AC2: 'Reply on RC2', Joseph Ovwemuvwose, 19 Nov 2025
Peer review completion
Interactive discussion
Status: closed
-
RC1: 'Comment on egusphere-2025-3785', Anonymous Referee #1, 22 Sep 2025
The comment was uploaded in the form of a supplement: https://egusphere.copernicus.org/preprints/2025/egusphere-2025-3785/egusphere-2025-3785-RC1-supplement.pdf
- AC1: 'Reply on RC1', Joseph Ovwemuvwose, 19 Nov 2025
- AC3: 'Reply on RC1: This the second response to the latest comment by referee 1', Joseph Ovwemuvwose, 04 Jun 2026
-
RC2: 'Comment on egusphere-2025-3785', Anonymous Referee #2, 22 Sep 2025
In this manuscript, Ovwemuvwose et al. examined the C3 and C4 fractions in global vegetation cover, gross primary productivity (GPP), and vegetation carbon stocks (Cveg) across 11 CMIP6 models. They report substantial differences among models in these C3/C4 fractions, as well as in their long-term trends. Those trends also do not align well with observations—for instance, the C3/C4 fraction in GPP and the global decline rate in δ¹³C discrimination. I read the paper with great interest because it addresses an important and often overlooked source of inter-model spread in CMIP6. However, several key issues prevent me from supporting publication at this stage.
1. Core assumptions about C3/C4 GPP and Cveg. CMIP6 models do not directly report GPP or Cveg partitioned into C3 and C4 components. The authors therefore estimated C3 and C4 values by multiplying total GPP or Cveg by the areal fraction of C3 or C4 vegetation. As they themselves noted, this is problematic: on average, GPP per unit area is higher for C4 than for C3 vegetation, whereas Cveg per unit area is likely greater for C3 plants (often woody) than for predominantly herbaceous C4 plants. These relationships also vary geographically, shaped by climate and soil. Consequently, it is difficult to place much confidence in downstream analyses of C3/C4 GPP, Cveg, and their trends.
A similar concern applies to the treatment of C4 crops. Several CMIP6 models lack explicit fractions for C4 crops versus natural C4 vegetation, so the authors assume C4 crop fraction = C4 fraction × total crop fraction. This implies that C3/C4 ratios for crops mirror those of natural vegetation at each grid cell—an unlikely scenario. Open data (e.g., Luo et al., 2024) show no such pixel-level pattern.
The LUH C4-crop fraction itself is also problematic: in LUH c4 crop fraction remains static (based on ~2000 data), so the apparent change in C4 crop area likely reflects total cropland expansion or contraction. It is unclear how individual CMIP models treat this issue, and I would not regard the LUH C4 crop trend as an “observation,” nor use it to declare a model (e.g., UKESM1) incorrect (L155, L328).
If the aim is a robust model–data comparison of C3/C4 fractions in area, GPP, and Cveg, I suggest considering the TRENDY DGVM ensemble, which may provide explicit estimates for these variables.
2. Writing quality and over-generalisation. Although I usually refrain from commenting on style, some issues affect the paper’s scientific clarity. I gave a few examples below
Carelessness: L137: “Click or tab here to enter text” is clearly a placeholder. The manuscript contains two “Discussion” sections (4 and 6); Section 6 seems intended as a Conclusion. Please proofread carefully. More examples in minor comments below.
Overstatements:The manuscript repeatedly describes the work as analysing “vegetation diversity” or “vegetation types” (L11, L16, L19, L74, L215, L275). In fact, it focuses on the photosynthetic pathway (C3 vs C4). Terms such as diversity or even vegetation types/PFTs carry broader meaning and could mislead readers. Being precise will not diminish the study’s novelty.
Other comments:
L31: “Land Cover and Land Use Change (LUCC)” is a more standard term.
L38: Molotoks et al.—only one reference is cited; the “a” after 2018 is unnecessary (cf. L286).
L43–44: The description of C3 vs C4 lacks detail on anatomical/structural differences that drive their distinct climate responses—important for motivating later arguments.
L70: The “3” in C3 should be subscript.
L125: You did not describe how C3/C4 Cveg was estimated; I assume you applied the same method as for GPP.
L139: Define Δtot before using it.
L140: Too many simplifying assumptions undermine the credibility of later detailed analyses; the variation in Δ is particularly large.
L186: Please clarify the source of the 10–45 % cropland figures.
L201: “UKESM is quite consistent”—please explain.L224: I remain unconvinced by the C4 GPP and Cveg values.
L315: Instead of stating that COS measurements prove models underestimate CO₂ fertilisation, consider a more cautious phrasing, e.g. “we noticed there is still difference between models and COS on the quantification of eCO2 effect and highlight the uncertainty in future 13C discrimination rate as eCO2 benefits C3 more than C4…”
Citation: https://doi.org/10.5194/egusphere-2025-3785-RC2 - AC2: 'Reply on RC2', Joseph Ovwemuvwose, 19 Nov 2025
Peer review completion
Journal article(s) based on this preprint
Data sets
CMIP6 models data Output Different groups https://esgf-node.ipsl.upmc.fr/projects/cmip6-ipsl/
Interactive computing environment
Jupyter notebook used for the CMIP6 models analysis and data visualisation Joseph Ovwemuvwose https://zenodo.org/records/16883407
Viewed
| HTML | XML | Total | Supplement | BibTeX | EndNote | |
|---|---|---|---|---|---|---|
| 11,430 | 1,051 | 251 | 12,732 | 533 | 253 | 320 |
- HTML: 11,430
- PDF: 1,051
- XML: 251
- Total: 12,732
- Supplement: 533
- BibTeX: 253
- EndNote: 320
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1
Joseph Ovwemuvwose
Ian Colin Prentice
Heather Graven
The requested preprint has a corresponding peer-reviewed final revised paper. You are encouraged to refer to the final revised version.
- Preprint
(1975 KB) - Metadata XML
-
Supplement
(1394 KB) - BibTeX
- EndNote
- Final revised paper