the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 02 Jul 2025
The Green Sahara Re-desertification transition and its Climate Impacts over Northern Africa, Mediterranean Basin and the Levant
Abstract. This study presents an ensemble of atmosphere-ocean coupled Regional Climate Model (RCM) simulations for the Mid-Holocene (MH) climate across the Middle East, Mediterranean and North Africa (MENA) regions. This ensemble is generated using the Weather Research and Forecast (WRF) model with online coupling to the Regional Ocean Modelling System (ROMS) ocean model, which simulates the dynamics of the entire Mediterranean Sea, utilizing forcing derived from the Mid-Holocene (MH) climate generated from the UofT-CCSM4 General Circulation Model (GCM) with a prescribed Green Sahara (GS), and having a prescribed GS in the land surface in WRF, but allowing the land surface scheme in WRF to compute influence from GS. The results of this ensemble are characterized by an increased precipitation field, similar to previous results, and a 2 m surface temperature that is lower than those in results that have a fully prescribed GS in WRF. Further analysis determined that under the same GS prescription as in a previous study, the land surface scheme used in this study produces higher evaporation and a smaller change in albedo, jointly producing a lower 2 m surface temperature. This conclusion is supported by a new set of sensitivity experiments that modifies the prescribed land surface field, which also showed that the climate over the Middle East is sensitive to land surface states over northern Africa.
- Preprint
(34843 KB) - Metadata XML
- BibTeX
- EndNote
Status: final response (author comments only)
-
RC1: 'Comment on egusphere-2025-2957', Anonymous Referee #1, 07 Aug 2025
Review of the paper The Green Sahara Re-desertification transition and its Climate Impacts over Northern Africa, Mediterranean Basin and the Levant, submitted to CP by Xie et al. 2025.
General
This paper addresses the topic of MH orbitally forced climate and environmental transitions and associated land cover feedback in the MENA region. The topic is relevant for the understanding of climate dynamics at global and regional scales, in the context of ongoing climate change which requires urgent adaptation measures.
Although the overall methodology used to investigate the topic is appropriate, the paper fails in meeting the standards for publication in CP. My main concerns are listed below:
- The paper is very long and not very well organised, with many figures composed by many panels, making difficult to establish the connection between the research questions, the methodology and the results. Material should be selected to only present what is relevant to the conclusions. The use of English needs further improvement.
- The research questions are not clearly outlined.
- The experimental setup seems very complex and the relationship with the research questions is not clear.
- Differences and comparisons are presented without any assessment of statistical significance, and the evaluation of whether the results are quantitatively meaningful is not possible.
- The discussion of the physical processes is often speculative only, not supported by evidence.
- The novelty of the results is questionable: the importance of the land cover processes for correctly simulating MH climate in the MENA region has been already pointed out in many modelling studies: what is the added value of this paper? In particular, what do we learn from the analysis regarding the optimisation of the parametrisation schemes to simulate MH climate in the MENA region?
For the above reasons, I cannot recommend the publication of this paper, which I believe still requires substantial and major modifications. I add below some comments that I hope can be helpful to the authors for improving the paper.
Specific
Title: Where in the paper re-desertification is discussed? The paper is about greening.
Introduction cites the relevant literature and sets the background for the study. However, research questions and objectives of the paper are not declared. I suggest you add at the end of the introduction a statement regarding research questions and objectives of the paper.
Section 2.1: A table summarising the experimental setup would be helpful to visualise the overall experimental approach. Please also clarify what you mean with “sensitivity expansion simulations”.
Section 2.2: The experimental setup is rather complex, description needs improvement. Please clarify that each combination of parametrisation schemes represents an ensemble member (the expression “each ensemble set” at L151 is confusing). Please also explain the rationale behind the definition of the different ensembles and members. Finally, please better clarify the objective of the albedo experiment.
Figures 2 to 6: Why N individual members are not shown? On which longitudinal band precipitation is averaged? How statistical significance of the anomalies is estimated? How proxy-model agreement is assessed? Specifically, how uncertainty in proxy reconstructions compare with the internal variability of model simulations?
Section 3.3: This section should be reorganised to discuss the physical variables presented in Figures 9-13 along with the summer precipitation response presented in Figure 8. This would help the reader to better visualise the physical processes leading to the precipitation response.
Section 3.4: It is not clear why the albedo experiment is only briefly discussed in relation to the Middle East case study. The paper is already very long, and the discussion of the Middle East case seems to deviate from the main story. This is a very relevant topic indeed, and it would deserve a full article, instead of only such a brief discussion.
Minor/Technical
L44: Please check brackets. Same at L67, 176, 201, 204, 223, 224, 231, 241…
L50: I’d also mention Africa-CORDEX.
L61: Can you please add some words about the specificity this experimental setup with respect to the CORDEX protocol?
Figure 1: A legend for the colours associated with the different land covers would be helpful. What do you mean by “in each region”? “Categories” instead of “category”.
L110-113: This sentence is a bit intricated, please rephrase it.
L126: Please declare the variable name first and put the acronym in brackets.
L146-151: please check for typos in this sentence.
L166: I assume RCM should be GCM instead.
L169: Please use clear different labels for RCM and GCM simulations.
L178-186: The phrasing of this paragraph makes difficult to follow the story, please consider rephrasing and streamline it.
L203: Do changes in temperature and precipitation refer to the whole domain or to specific regions?
L206: I’m not sure 100% that you can use Mesopotamia to refer to modern Middle East: https://www.britannica.com/place/Mesopotamia-historical-region-Asia
L207: What L5 member are you referring to?
L226, 262, 294, 298, 337, 368, 540, 554: How statistical significance is assessed?
L267: Where exactly is the increase in precipitation observed?
L300: Where can we see the effect of increased evaporation on precipitation?
L314-318: This part is speculative; we cannot see anywhere evidence of those processes going on.
Citation: https://doi.org/10.5194/egusphere-2025-2957-RC1 -
RC2: 'Comment on egusphere-2025-2957', Anonymous Referee #2, 14 Aug 2025
Recommendation
This manuscript does not clearly articulate a valid research question and although the authors use an advanced regional modelling setup to show variability between schemes, this is already a well established result. The albedo tests labelled ‘A’ are far as I can understand the main advance over the authors’ paper in press, but these are set up in a complex manner which is not well justified. The manuscript is overly long with potentially redundant figures. I am sorry but I do not recommend publication.
Main comments:
1) It remains unclear by the end of section 1 what the purpose of this study? It seems that some simulations are already in press by the authors but they propose to try some other land surface schemes. However it is not clear what the aim of this is or what will be learnt from this.
2) Reading section 2 the main advance proposed here seems to be an incremental modification of albedo and FPAR (fraction of photosynthetically active radiation) in which regional increments are added depending on the vegetation type already prescribed. This therefore seems to add even more ‘greenness’ to a model with green Sahara already prescribed. It is unclear a) what is the justification for this, b) whether the resultant fields are physically plausible and c) what we can learn from this sensitivity test.
3) Given that the land-surface is prescribed it is unclear why using a different land surface scheme should meaningfully change the results. This is hinted at on line 260-262: “This confirms that the “control Mid-Holocene” setup is properly implemented in the L ensemble members, and suggests that choosing a different land surface scheme does not lead to a significant difference in MH climate anomaly in the MH-ref scenario without the GS surface.” but seems to be a major weakness of the experimental setup.,
4) Any dependency of the results on the land surface scheme used is entirely expected given existing model differences. The differences are analysed but there is no connection made to the structure or parameters of the land surface scheme. Therefore, it is unclear what we can learn from this comparison.
5) The principle finding from the albedo modification run A seems to be that it can also influence the climate over Arabia. This is not justifiable as currently written (see comment 2 above). But this finding is also already evident from the non -A sensitivity runs that seem to be in press i.e. from the panels on the left column of figure 14. This aspect needs to be clarified.
Minor Comments:
Lines 6-7: “having a prescribed GS in the land surface in WRF, but allowing the land surface scheme in WRF to compute influence from GS”
This does not make sense to me I’m afraid. Please clarify what this means and why it is important for regional modelling.
Lines 8-10 “Further analysis
determined that under the same GS prescription as in a previous study, the land surface scheme used in this study produces
10 higher evaporation and a smaller change in albedo, jointly producing a lower 2m surface temperature”
At this point it is unclear what the previous study is and how it can be referenced like this in the abstract? Is it another study with this model or a different model? It is also unclear what the importance of this finding really is beyond other users of this model?
Lines 65-72: it is unclear what you are getting at with ‘only one land surface scheme” it would be helpful to elaborate (even if only with model/scheme names).
Lines 123-136: The description of the sensitivity simulation is unclear. Why are you changing albedo in a run that already has a GS and why are you altering FPAR? The resultant albedo and FPAR fields are not shown. Are they physically plausible?
Lines 202: “produced an increase in annual precipitation and 2m surface temperature (T2) with respect to the PI climate”
This is not clear from figure 3?
line 571: “By comparing the absolute annual precipitation over
various sub-regions in the Middle East, this study found that the shift away from GS over northern Africa will lead to increased
aridity over the Middle East, to an extent that causes deforestation in it.”
This is unclear.
In many places the bracketing of references is incorrect. Here is one example:
Line 60: “... and coupled atmosphere-ocean experiments Huo et al. (2021).”
Figure 1: the colour scale should be labelled with classes not numbers.
Citation: https://doi.org/10.5194/egusphere-2025-2957-RC2
Viewed
| HTML | XML | Total | BibTeX | EndNote | |
|---|---|---|---|---|---|
| 524 | 30 | 11 | 565 | 30 | 50 |
- HTML: 524
- PDF: 30
- XML: 11
- Total: 565
- BibTeX: 30
- EndNote: 50
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1