the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 08 May 2026
Hot extremes following net-zero CO2 emissions in UKESM: physical drivers and role of vegetation
Abstract. Reaching net-zero CO2 emissions is essential to halt continuing global warming and attempt to stabilise global temperatures. However, large uncertainties remain on the sign and the magnitude of the long-term responses of the climate system following anthropogenic emissions cessation. This study contributes to improving our understanding of the climate system post CO2 emissions cessation, by exploring the global and regional temperature evolution in UKESM1.2 following the TIPMIP protocol. Zero CO2 emission simulations, starting from global warming levels of +1.5 °C to +5 °C above pre-industrial are analysed to understand the impact of historical cumulative emissions and associated global warming level on post zero-emissions trends. We find that the global average surface air temperature (GSAT) keeps increasing in all zero CO2 emission UKESM1.2 projections. The increase is more pronounced at higher warming levels, approaching 0.25 °C per century in the +3.0 °C to +5.0 °C scenarios. Most of the warming occurs in the Southern Hemisphere, particularly in the Southern Ocean, while the Northern Hemisphere land experiences a slight cooling trend. These regional cooling trends are more marked for the annual temperature maxima, with several regions across 45–65° N experiencing cooling of >1 °C per century. We find the strongest regional cooling trend following emissions cessation in the higher warming scenarios. Here, we investigate the drivers behind the cooling trend in northeastern North America, where the cooling magnitude exceeds 1.5 °C per century. We find that the cooling trend is almost completely explained by thermodynamic drivers. We reconcile this finding with the UKESM1.2 dynamic vegetation changes, as the evergreen vegetation cover increases across all regions experiencing substantial cooling in the hot extremes. This finding highlights the significant regional contribution that vegetation changes can have for the attenuation of annual temperature maxima, supporting the case for their careful consideration in future mitigation and adaptation strategies. However, these results also show the limitations of highly idealised scenario protocols like TIPMIP, which set crop and pasture distributions, as well as other anthropogenic forcings, to pre-industrial values, allowing vegetation to expand freely. This highlights the importance of developing new zero emissions protocols considering other forcing agents beyond CO2.
- Preprint
(3380 KB) - Metadata XML
- BibTeX
- EndNote
Status: final response (author comments only)
- RC1: 'Comment on egusphere-2026-1733', Anonymous Referee #1, 29 May 2026
-
RC2: 'Comment on egusphere-2026-1733', Anonymous Referee #2, 23 Jun 2026
Summary:
This study aims to characterise post-net-zero changes in heat extremes and provide a mechanistic interpretation for relevant regional differences. Even though mean temperatures tend to slightly warm over land after emissions cessation, the intensity of heat extremes cools substantially over certain regions, particularly at high global warming levels. Northeastern North America is used as a case study, where a dominant thermodynamic contribution is found. This contribution is distinctly linked to the expansion of broadleaf evergreen vegetation, which also occurs in other regions where heat extremes reduce their intensity.
The study is timely, well-written, and the literature is appropriately considered and cited. The mechanistic approach not only provides a first explanation for post-net-zero changes in heat extremes but also has wide-ranging implications for policymaking, as it links land use planning with mitigation and adaptation benefits.
I recommend the article for publication, subject to consideration of the following smaller comments and suggestions, which are intended to strengthen the manuscript and refine some smaller issues.
General comments:
- Missing supplementary plots: throughout the text, some results are presented accompanied by not shown instead of pointing to a figure. This is the case at least six times, which is too many, in my opinion. I believe that in most cases, the accompanying analyses have been done and would be valuable assets as supplementary figures, so I would encourage the authors to include them.
- Justify/explain the choice of some experiments or regions in analysis: In some cases, some experiments or regions are chosen for part of the analysis, but without really justifying why they were chosen. For instance, three of the experiments are used in Figure 5, or only one in Figure 6. In some of these cases, adding other experiments in the supplementary material would also help in showing that the displayed experiments were not cherry-picked.
- Show dynamic contribution: In the analysis of atmospheric circulation analogues, a small, but not negligible, dynamical contribution is found (Figure 4, second row, third column). Thus, I think it would be important to also show the slp and/or z500 fields, and discuss the changes in these when comparing the late and early periods.
- Robustness of linearity and trends: in many figures, linear fits are used, and trends are computed, but it is not often stated how well the fit actually explains the data. For example, trends in Figure 3 or Figure 5 would benefit from some mention of the reliability of the linear model (either directly on the plot or somewhere in the text).
- Add some more discussion: the discussion of results is comprehensive and clear. Yet, a few more lines could be added. For instance, on the limitations of using stabilised warming levels compared to warming levels in a present-day/future transient climate.
Specific comments:
- Abstract, lines 7-10: three sentences in the abstract are used to present the results on mean temperature changes. Given that these are not the main focus of the article (and not particularly original considering the existing literature), I would recommend condensing them into one sentence or leaving them out entirely to focus on the changes in heat extremes.
- Line 23: Given the wide relevance of the temperature threshold, it should be clarified how 1.5°C above pre-industrial is defined (which, in this case, I guess it is just one year above the threshold).
- Line 31: The wording is long established is a bit misleading. It seems to suggest that the post-net-zero global temperature evolution is better understood, and for a longer time, than the literature suggests (as mentioned in the next sentence). I would recommend providing a more nuanced statement and possibly mentioning some of the early work focusing on net-zero responses (e.g. Mathews & Caldeira, 2008).
- Lines 51-52: Two things here might lead to a minor misinterpretation of the TIPMIP setup. First, it reads as if 80 Gt CO2 are imposed by the protocol, but it instead asks to find the emissions rate that leads to the desired warming trend. Second, while the goal is technically 2°C after a century, the protocol aims for 0.2°C/decade, which is a slightly different goal, as it carries a more specific guideline on the precision of the rate of warming.
- Line 69: maybe mention the method to estimate the linear fit?
- Line 92: I cannot find a previous mention of the piControl run. It might be good to do that when explaining the analysed simulations, and mention the length of the experiment, which is relevant here.
- Line 106: the residual mentioned here can be confused with the residual contribution (thermodynamic) mentioned a few lines later. Also, how large is e, and does it affect the calculation of the dynamically induced temperature component?
- Is it mentioned somewhere with respect to what period exactly the anomalies and trends are calculated? (i.e. is it just the value at emissions cessation, or some window around it?)
- Line 118-121: a few intervals are mentioned here; are these covering the full range across all experiments, or some other measure of spread across the simulations?
- Lines 124-126: it may not be as clear as in the net-zero UKESM simulations, but the simulations from King et al. (2024a) also seem to show trend dependence on the global warming level. This is in part because the trends there are smaller, but also the ACCESS-ESM runs seem to cover a smaller range of global warming levels. I would rephrase the sentence here to better reflect that previous studies also show a similar tendency (i.e. stronger warming trend in warmer equilibrating climates).
- Line 138: the widespread increases over land masses are barely noticeable in the presented figure, due to the colorbar chosen, as changes are often much larger over high latitude ocean regions. Maybe an alternative supplementary figure with only land masses (and a colorbar covering a smaller range) could help, but at least this (the differences in magnitude of change between regions) should be acknowledged more consistently in the text.
- Figure 2: I would label the subpanels and use the labels when referring to the results in the text (e.g. line 149), since it sometimes takes a while to find the plot that is being described.
- Figure 4: it is unclear to me how the repeated subsampling (30 times selecting 20 events in each period, as mentioned in the Methods section) is reflected in this figure. Would there be a way to show the robustness of the found patterns?
- Figure 5: This figure is a little busy, so it might be worth considering separating some of the experiments into different panels. In any case, it might be important to mention the large variability somewhere in the text, since it currently reads as if the relationship is much clearer than what the figure really shows.
- Lines 211-213: Maybe a longer explanation on why this is the case might be relevant. Also, if winter temperatures are most relevant, I wonder if a supplementary plot for winter temperature changes might also be a nice addition.
- Line 214: I think that experiment dn1.0 has not been properly introduced in the methods beyond saying that experiments with negative emissions are also used.
- Lines 223-224: It reads at the end as if the negative emissions values go fully back to the values in the ramp-up, but the figure suggests that there is still a large fraction of hysteresis.
- Figure 8: the legend of the left panel is partly cut out.
Typos/language suggestions:
- Abstract, line 6: comma missing after pre-industrial.
- Line 29: both CO2 emission and emissions are used throughout the text. I would stick to emissions.
- Line 30: How are references ordered? I would recommend either temporally or alphabetically, but none seem to be used consistently in the manuscript.
- Line 37: I would say could have instead of has.
- Line 53: net-negative projections can be misleading; use net-negative emissions projections instead.
- Line 70: The them could refer here to both the heat fluxes and the temperature extremes.
- Line 113: change events to event’s.
- Line 190: observed might be misinterpreted here, as if it would refer to observations.
- Figure 3 caption: change wit to with.
- Line 199: correct A2 to A1 (and vice versa in line 250).
- Line 202: striking is a bit too subjective; I would prefer more neutral language.
References:
- Matthews, H. D., & Caldeira, K. (2008). Stabilizing climate requires near-zero emissions. Geophysical Research Letters, 35(4). https://doi.org/10.1029/2007GL032388
Citation: https://doi.org/10.5194/egusphere-2026-1733-RC2
Data sets
UKESM TIPMIP Ensemble Terrafirma consortium https://gws-access.jasmin.ac.uk/public/ukesm/TerraFIRMA
Viewed
| HTML | XML | Total | BibTeX | EndNote | |
|---|---|---|---|---|---|
| 199 | 55 | 11 | 265 | 21 | 18 |
- HTML: 199
- PDF: 55
- XML: 11
- Total: 265
- BibTeX: 21
- EndNote: 18
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1
General comment:
This study addresses how heat extremes change after net-zero CO2 emissions in climate-stabilizing scenarios. The authors find decreasing trends in mean temperature and in the temperature of the hottest days, which are associated with heat waves, in parts of North America and East Asia. After decomposing this cooling signal into dynamic and thermodynamic components, they attribute the thermodynamic component to changes in vegetation type. The study's storyline is timely, and the findings are scientifically interesting and useful for society (though the reality and regional impact with a coarse-resolution model have some limitations). The methodology in the analysis is also reasonable. While I generally have a positive impression, some clarification is needed before I can recommend acceptance.
Major comment:
Minor comment: