the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 20 Jul 2026
Drivers and uncertainties of regional climate change through overshoot and net-zero greenhouse gas emissions futures
Abstract. Humanity’s emissions of greenhouse gases have warmed the planet to the point that Earth may already be passing the 1.5 °C Paris Agreement threshold. Without stringent reductions in emissions of greenhouse gases, it is becoming increasingly likely that Earth will exceed 2 °C global warming by the end of the century. Should Paris Agreement global warming levels be exceeded, future periods of net-negative carbon dioxide emissions would be required to reverse global warming and satisfy the Paris Agreement goals. However, the implications of that for regional climates remain uncertain. We investigate drivers and uncertainties in regional climate change using the SSP5-3.4 overshoot scenarios that include extended periods of net-zero anthropogenic carbon dioxide emissions after overshoot and net-negative emissions. We find that model differences in evolution of regional climate during the net-zero carbon dioxide emissions period are far greater than model differences during the net-positive and net-negative emissions periods of the overshoot scenario. However, the diversity in models’ climate projections during the post-overshoot net-zero emissions period may be partially explained by changes in large-scale sea surface temperature and atmospheric pressure gradients. By further exploring sources of uncertainty in climate projections throughout the overshoot scenario, we find that the uncertainty from simulated internal variability can far outweigh the differences in climate projections between models, especially during the extended net-zero emissions period of the scenario.
- Preprint
(2212 KB) - Metadata XML
-
Supplement
(4742 KB) - BibTeX
- EndNote
Status: open (until 31 Aug 2026)
- RC1: 'Comment on egusphere-2026-3644', Peter Pfleiderer, 30 Jul 2026 reply
-
RC2: 'Comment on egusphere-2026-3644', Anonymous Referee #2, 31 Jul 2026
reply
Review of "Drivers and uncertainties of regional climate change through overshoot and net-zero greenhouse gas emissions futures"
Overall Evaluation:
The paper examines Earth system model simulations of SSP5-3.4, examining the differences in regional climate change during net-positive, net-negative and net-zero emissions periods. The study finds that greatest diversity in model response during the net-zero phase, with some models resuming global warming, as well as a wide range of regional climate change. Overall, while I have some concerns, the paper is well written and scientifically sound, I would recommend that the paper undergo minor revisions.General Comments:
(1) My understanding is that all of the simulations were concentration driven, even during the 'net-zero' phase. Therefore it is likely that model's diagnosed emissions will deviate from true net-zero during this later period. Thus the authors should compute diagnosed CO2 emissions for this period to see how far models vary from true net-zero. Residual positive or negative emissions may explain some of the variation in response of the models during this late period.(2) Climate models are not independent and thus using formal measures of statistical significance on output from a group of climate models violates the fundamental assumptions of such metrics. Climate models often share components, and are related to one-another in almost a phylogenetic pattern (Knutti et al 2013). This is especially true for your group of models, which includes 3 variants of GISS. I would recommend that you remove all formal metrics of statistical significance and simply describe correlations as 'strong' or 'weak'. Please also note in the methods section the lack of model independence, so the readers can be appropriately cautious when interpreting the correlations.
Specific Comments:
Line 25 to 26: Replace 'of that' with 'of negative emissions' or similar.
Line 41: Two decimal places for the temperature is too high precisions, please round to nearest 0.1ºC.
Line 78 to 82: Please rewrite for clarity.
Line 153 and many other locations: Please add 'year' in front of numbers which are calendar years to improve clarity.
Line 159 and elsewhere: Please use consistent notation for SSP ex SSP5-3.4
Table 1: Please expand captions to describe the table.
Table 1: Write out 'Number' in place of 'N'Line 197: Typo for 'pr'
Line 251 to 252: Please write out 'land to ocean' in place of 'L-O'
Equations 6: Please fix typo in 'member'
Line 317: Be more specific about "changes to melting". Do you mean sea ice or ice sheets, or both?
Figure 3 and other figures: Please add a sentence to caption describing what each anomaly is relative to.
Line 385: Typo
References:
Knutti R, Masson D, Gettelman A. Climate model genealogy: Generation CMIP5 and how we got there. Geophysical Research Letters. 2013 Mar 28;40(6):1194-9.Citation: https://doi.org/10.5194/egusphere-2026-3644-RC2
Viewed
| HTML | XML | Total | Supplement | BibTeX | EndNote | |
|---|---|---|---|---|---|---|
| 98 | 32 | 11 | 141 | 27 | 13 | 9 |
- HTML: 98
- PDF: 32
- XML: 11
- Total: 141
- Supplement: 27
- BibTeX: 13
- EndNote: 9
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1
The authors present an in depth analysis of implications of overshoot scenarios on regional climates. The study focuses on long-term implications and therefore nicely complements earlier studies that only investigate the short-term changes in overshoot scenarios. The available literature is well presented and discussed in the introduction. The methods are clearly explained and the results are well presented although some parts of the results section could benefit from some more interpretation and streamlining (section 3.3). Overall, the methods seem sound and the manuscript is well written. I think that the manuscript would be an important contribution to the climate overshoot literature. Before publication, I would recommend discussing some of the methodological choices (definition of periods) in more detail and expanding the interpretation of some of the results.
Major comments:
Introduction: Good overview of previous studies. The research gap could be stated a bit more clearly. I guess it is the analysis of the “net-zero” period after the overshoot which has not been studied in (most?) previous studies.
Is there a specific reason for the placement of the net-positive and the net-negative climate periods? Is there a reason to not place these periods in the center of the ramp-up and ramp-down instead of just before peak in the warming case and just before net-zero in the cooling phase? If there is a reason, it would be interesting to know why.
After reading the results section and it’s interpretation I’m wondering whether the meaning or interpretation of the selected periods might be misleading. To me it seems as if the periods could be interpreted as follows:
- Historical: pre-industrial climate
- Net-positive: peak warming (or something close to it)
- Net negative: first moment of net-zero (or just before that)
- Net zero: after a long period of net zero
I think that these periods are interesting for the comparison, but to me they do not fully align with their naming. For instance, “net-positive” is indeed a period with net-positive emissions, but the main difference to “historical” is that climate is way warmer. Therefore, I also struggle with the wording of “ramp-up” and “ramp-down” in the conclusions.
This choice of the periods and the way they are analyzed makes it hard to interpret the processes that might explain differences between the maps in Fig. 3 and 4. For instance, there could be different explanations for the result that the differences between “net-negative” and “net-positive” are smaller than the differences between “net-positive” and “historical”: (i) the difference in GMST is larger for “pos” vs “hist”, (ii) in “neg” vs “pos” slow responses to the emissions during the ramp-up phase might still be at play and potentially mask some of the effects of the ramp-down period, (iii) ramp-up was faster than ramp-down. These are just a few explanations that come to my mind. That being said, I agree that in this “not-idealized” it is difficult to find a way to adequately compare the effects of ramping-up and ramping-down. For the article, it would still be helpful to discuss in a bit more detail what the choices of the analyzed periods and their comparison implies.
One additional way to look at the temperature changes would be to plot the differences between selected periods after removing the global mean surface temperature value of each respective period. I think that would nicely reveal the regional differences in the warming (net-pos vs hist), cooling (net-neg vs net-pos) and stabilization (net-zero vs net-neg).
It would be good to extend Figure 2 to 1850 and include the selected periods to see how they differ in terms of GMST.
The correlation analysis in section 3.3 is an interesting way to explore the drivers of regional changes in the net-zero period. I found this section a bit hard to follow as there are numerous different correlations that are computed and discussed. To me, it is not fully clear what is meant by “storylines” in this section. (a) Is a change in correlations between precipitation (or temperature) and one of the large scale indices a storyline? (b) Or is it a strong change in one of the large scale index with implications for precipitation and temperature? When I read the titles of the sections I was thinking that something like (b) would come, but it seems as if (a) was the aim of this analysis. This was confusing to me. Although all the analysis is useful and insightful, some additional interpretation and guidance would be helpful here.
Some of these correlations are only shown in the supplement and there in a pretty raw manner. If possible, I would suggest to show some results from section 3.3.1 in a summarized manner. Or is this data presented in table 2? In line 433-434 this table is introduced as:
“Table 2 lists regional details and large-scale indices used to explore regional climate change after net-zero.” Anyways, why don’t you show these correlations for both climate states in this table?
The comparison of multi-model and multi-member uncertainty is interesting. Since the multi-member uncertainty is only based on ACCESS-ESM1.5 I would recommend to discuss how internal climate variability in this model is represented as compared to other models from CMIP6. For the “net-positive” period the internal variability could be compared to other models. One could also check whether people have discussed such differences between models in the literature.
Minor comments:
l96: I wouldn’t call SSP5-34 a “plausible” scenario. I guess you mean “less idealized” in contrast to CDR-MIP.
l135: We use simulations from the ... (?)
l180: “net-p” in equation
l265: typo “membe”
Figure 2: has no panel b or c.
Figure 3 & 4: I was a bit confused about panel c. Is there a reason to count models with positive anomalies? I think that for fig. 3 bottom middle, the count of negative anomalies would be more interesting. Would a count of “model-agreement” be an option?
l399: Are these anomalies to the “net-negative” period?
l583-595: Minor comment on the wording: While in the results section you discuss periods (a period before peak warming and a period before net-zero) here you write about “ramp-up” and “ramp-down”. This might be misleading as people might interpret it as a trend analysis.
l607-609: I agree, that this analysis “may be useful in understanding diversity in regional climate projections after net-zero”. And you did the analysis, so I think you could expand the interpretation of this analysis and also highlight a few key findings/hypothesis/questions in the conclusions.