General comments

This paper uses a new set of climate model simulations to explore climate stabilization under zero green gas emissions. This is a very welcome and novel study, which provides an insight into how climate change might evolve, and a useful comparison with many other studies which focus on much shorter simulations with rapidly increasing emissions. The paper has some important findings, for example highlighting that delaying mitigation by even 5 years could have implications for hundreds of years. The paper also illustrates how the impacts of global warming change over time and highlights the need for more research to explore the impacts of different mitigation pathways. The paper is generally fluently written with useful illustrations.

I have a few broader comments which might help improve the manuscript:

• I would like to see a bit more discussion of the relevance of these simulations to the real world/ climate policy. The idea that we could instantly stop greenhouse gas emissions is of course hypothetical, and in reality even very strong mitigation would be associated with a decline in emissions over time. This complicates matters as it is more likely to lead to overshoot scenarios, which could have different implications from the scenarios used here. It would be useful to add a short discussion of this.

• The terminology “net zero” is often used to refer to mitigation which is compliant with the Paris Agreement (i.e. 1.5⁰C or 2⁰C global warming), whereas in this study some of the net zero scenarios exceed 3⁰ I found this a bit confusing, and I had to keep reminding myself that “net zero” didn’t necessarily imply low emissions. I also found it a bit confusing to compare the 3⁰C transient sample with the 3⁰C stable samples, as I believe they have different cumulative emissions? In some cases a “reduction” (for example in heat extremes) is noted between the transient and stable cases, and it is unclear whether this is a reduction over time in the same scenario, or rather a difference between a rapidly warming 3⁰C world with high emissions and stabilized world which has received fewer anthropogenic greenhouse gas emissions and more slowly reached 3⁰C. I have added a few comments below that might help make the comparison between these clearer. In addition, I think it would be useful to calculate the cumulative emissions for each GWL condition. Perhaps this could be added to Table 1? I also wonder whether it might be more straightforward to refer to the scenarios as “zero emission scenarios” rather than “net zero” since I think the scenarios imply that emissions from CO₂ sources stop, rather than there being any change in CO₂ sinks or CO₂ drawdown.

• The manuscript is generally very well written, but some sections of the results are a bit complicated to read. I have provided line by line comments to help improve clarity.

• The paper is quite long, and below you will see I have suggested adding a couple more panels to some of the figures! As a reader I didn’t find it too long, but if it needs to be shortened I am not sure if Figure 5 and 6a are both needed. Also, the methods section has quite a bit of detail which I found a little hard to follow without the results, some of the explanation could come during the results figure by figure. The introduction is also quite long, although it is very clearly written and useful.

Line by line comments

Line 26 – “differ greatly” – differ with reference to what? Higher greenhouse gas scenarios? Or is this about difference between regions? Please clarify.

Line 45 – “will result” – suggest to change to “would result” since it is a scenario.

Line 284 – reference to Joshi et al. 2008. Please make clear how this reference supports – did they find something similar and with what kind of simulation?

Line 291 – please clarify what you mean by a “fast rate”.

Line 292-295 – suggest to clarify sentence – “even under the lowest global warming simulation broadly aligning with the Paris Agreement” – which simulation do you mean? And, in the latter half of the sentence I think the “beyond” could be removed? Not sure what is meant by “beyond” here.

Line 299-300 – “relatively small” – relative to the point of zero emissions yes, but there is a change relative to preindustrial. Please clarify.

Line 303-304 – I think it would be a good idea to put this in the context of the change during the transient simulation. You could say that the models are showing strong persistence in the change in sea ice, since the decrease in sea ice extent experienced during the transient simulation is then broadly maintained (albeit with substantial variability) for 1000 years.

Line 305-6 – Do we know why the decline continues after emissions cessation in the Antarctic? Is it due to slower ocean changes in the Antarctic region? It would be nice to comment or perhaps give a hint that you will explore this later in the paper.

Line 314-315 – might be helpful to add “over time” for example “Under net-zero emissions, in the Antarctic, there is an increasing change of sea ice free events over time…”

Line 330-331 – why might it be related to Southern Ocean warming? Could you give a brief indication (quite interesting!).

Line 345-347 – Yes, if the rate of emissions reduction is the same.

Line 356 – “Africa” – more helpful to say “Northern and Central Africa”

Line 372 – and also dependence on the cumulative emissions? Does the SSP5-8.5 3⁰C have higher cumulative emissions than the late stable 3⁰C?

Figure 8 and 9 – I wonder whether it would be useful to show and describe the warming patterns for the early stable and late stable periods as well? i.e. the warming relative to preindustrial rather than the difference between SSP5-8.5 and the stable periods? It might be nice to make the point clearly that a 3⁰C “transient” world looks very different from a 3⁰C “stabilized” world – i.e. (I think) – 3⁰C transient has huge warming over continents and some warming over sea ice regions, 3⁰C transient has more uniform warming? Are the continents showing warming similar to the global mean of 3⁰C? As a reader I’d quite like to visualize this.

418-419 – “suggest very large areas of the world may exhibit some return towards pre-industrial levels of seasonal-average precipitation.” – very interesting. Again, could you show this? Even as an appendix.

480 – “marked reduction” – is there a decrease over time as the climate stabilizes, or is this because a different simulation is used for the “early stable 1.5⁰C” vs the “late stable 1.5⁰C”. If the extremes actually decrease over time, despite the cumulative emissions being the same, this is really interesting and could be highlighted more. Either way it would be nice to comment on this.

500-505 – Please could you clarify whether you expect that the stabilization scenarios show a decrease after emissions cessation? Or is the reduction between the 21^st century and the net zero simulations shown in Figure 15 a result of the different emissions (where this 21^st century simulation includes emissions and temperatures which exceed any of the net zero stabilization scenarios?)

522 – “reduced land temperature means and extremes” – is there a reduction over time? Or is it a reduction compared to a transient world with the same GWL (and higher emissions)?

Figure 14 caption – check references to figure letters. I think “e, f” should be “d, e”

1. Does the paper address relevant scientific questions within the scope of ESD? Yes
2. Does the paper present novel concepts, ideas, tools, or data? Yes
3. Are substantial conclusions reached? Yes
4. Are the scientific methods and assumptions valid and clearly outlined? Yes
5. Are the results sufficient to support the interpretations and conclusions? Yes
6. Is the description of experiments and calculations sufficiently complete and precise to allow their reproduction by fellow scientists (traceability of results)? Yes
7. Do the authors give proper credit to related work and clearly indicate their own new/original contribution? Yes
8. Does the title clearly reflect the contents of the paper? Yes
9. Does the abstract provide a concise and complete summary? Yes
10. Is the overall presentation well structured and clear? Yes
11. Is the language fluent and precise? Yes, generally, I’ve made suggestions for places where it could be clearer.
12. Are mathematical formulae, symbols, abbreviations, and units correctly defined and used? Yes
13. Should any parts of the paper (text, formulae, figures, tables) be clarified, reduced, combined, or eliminated? Yes
14. Are the number and quality of references appropriate? Yes
15. Is the amount and quality of supplementary material appropriate? Yes

General comments

This manuscript analyzes a new set of climate change experiments utilizing the ACCESS-ESM-1.5 Earth system model to simulate climate evolution under net-zero emissions. Seven simulations spanning 1000 years were branched of the SSP5-8.5 simulation at 5-year intervals starting in 2030 to assess climate stabilization in line with the Paris Agreement and higher warming scenarios. Significant findings are provided for continued Southern Ocean warming, reversal of mid-latitude drying trends, and varying sea ice extent trajectories in the Arctic and Antarctic. The study also examines the El Niño-Southern Oscillation (ENSO), noting reduced amplitude and frequency of events under climate stabilization. The results show that regional changes persist for centuries post-emission cessation, indicating long-term climate impacts despite net-zero emissions. The results are presented by temporal evolution and by global warming levels, where climatic changes in early and late periods after emission cessation are compared against transient warming under SSP5-8.5. The manuscript provides relevant and novel results such as insights into the climatic effects of delayed emission cessation, regional timing of peak warming after emission cessation, and temperature and precipitation trend differences between transient and stabilized global warming levels.

I do have a few questions and comments, which I consider overall minor, that I would like the authors to address.

The manuscript is well written, structured and generally reads fluently. While the manuscript is quite long, the flow of the text does help its understanding. Specific suggestion for sections to be adjusted are listed below.

This study provides a unique dataset as the stabilization simulations are run in emission-driven mode and because they are provided for an entire millennium, which is specifically relevant for ‘slow’ climate subsystems such as the ocean or cryosphere elements. The authors highlight an important point in line 96, that is relevant for the consideration of climate change experiments for policy under emission budget frameworks. Hence, the experimental design provided here should be inspiration for other modeling groups to at least replicate a sub-ensemble of the experiments described here for a better inter-model comparison.

Given that these simulations are emission driven, I was a bit surprised that no carbon cycle response was analyzed. I understand that ocean and land carbon cycle responses will affect atmospheric CO₂ concentration (which is provided) and therefore its effect on temperature (and precipitation) is shown – ignoring physical feedbacks here, but given the length of these simulations, there would be an opportunity to analyze long-term carbon cycle responses under emission cessation. Since the paper is already long in itself by focusing on temperature and precipitation, I would not expect any additional analysis here, as the authors also refer to further analysis in the last sentence of Section 4. Nonetheless, I would appreciate one or two sentences discussing carbon cycle implications in these emission-driven simulations in the Conclusions Section.

Further do I miss a bit a more refined discussion of the results in the frame of other ESM responses (such as under ZECMIP). While the authors make clear that the results can only be interpreted under the specific climate response of ACCESS-ESM-1.5, its differences to other models, or e.g., the ZECMIP multi-model mean, can affect the interpretation of the results presented here. I suggest to either add a few comments in appropriate places in the main text or discuss this issue a bit more in the Conclusions Section.

An overall limitation of this study is that a single climate model is used, which is very relevant for and interpretation of the result provided here. This is acknowledged by the authors in various places of the main text. However, I think this is a key part to keep in mind when assessing the results of this study. I made specific comments below where I think the manuscript could benefit from some more detail in the framing of the results in the context of other climate models.

Specific comments

From the abstract, I find it is not entire clear whether the focus of the analysis is on the comparison of stabilization vs. transient climate responses, or long-term stabilization at different warming levels only. The reader might appreciate if this can be better clarified.

Line 25 – ‘mid-latitude drying trends’ – it might be useful for the reader to know here that this refers to extremes provided in the analysis.

Line 42 – ‘near-linear relationship between cumulative carbon dioxide emissions’ – there are indications that this might only be true for transient positive emissions.

Line 46 – ‘near-zero global mean temperature change for the following century’ – that’s on a multi-model mean, but individual models can show quite large positive or negative ZEC after emissions cessation.

Figure 1 – I find it not quite intuitive to present results in panels c) and specifically d) together with the conceptual panels of the experiments in a) and b). I would further like to see a panel showing cumulative emissions in these experiments. I therefore suggest moving panel d) to Figure 3 as the first panel, and exchange it with a panel for cumulative emissions in Figure 1. Also, the panels c) and d) have no x-axis labels, while panels a) and b) do.

Lines 115-116 – ‘… with the starting point chosen…’ – As I understand it, I am not convinced by the transferability of this approach to other ESMs because: 1) each ESM will have different long-term climate response – positive or negative ZEC, which vastly affects the branching points from the transient warming experiment, and thus may have different policy implications for when net-zero need to be reached, and 2) reaching temperature targets under non-stable climate response under zero emissions will strongly depend on the intended simulation length, as there is no stabilization (as illustrated in Fig. 1d).

Line 150 – ‘… representation of different ENSO flavors appears to be worse in ACCESS-ESM-1.5 than in other CMIP6 models’ – given that a part of the analysis focuses on the ENSO response, I miss some discussion of this point. How does it differ from other ESM? This part could be provided at this point of the main text. And in which way does this affect the results later discussed in the main Results Section and provided in Figure 15? This part could be incorporated in Section 3.4 line 495, where this is already referred to by the authors.

Line 159 – ‘… mean surface temperature (GMST) slightly decreases in the first 20-50 years…’ – could the authors please add an explanation as to why there is this initial drop?

Line 160 ‘reduced non-CO2 and aerosol concentrations has taken place…’ – The combination of these sentences is slightly confusing as non-CO2 and aerosols were set to 1850 levels as indicated in line 154. Please adjust for clarity.

Lines 157-168 – These lines describe results provided in Figure 1 (see my suggestion on Figure 1 above). I think this paragraph is better placed in the Results Section, e.g. line 281, which would go along nicely with moving Figure 1 panel d) to Figure 3. Please adjust for clarity of the Methods Section.

Line 172 – ‘quasi-stabilized’ – I find this a bit misleading, as emissions are entirely stabilized, but temperatures are clearly not. Perhaps remove.

Section 2.3.1 – This entire section is a listing of analyses that have been performed. This is a bit a question of taste, but I would suggest removing this Section entirely. The reader might have difficulties remembering methodological details when reading about the results later. Rather I would like to see the relevant bits incorporated into the results section where appropriate, which would make following and interpreting the results much easier.

Line 189 – ‘… with the first period…’ – disconnected sentence that sound strange, please revise.

Line 193 – given that there is often a focus on reaching 1.5°C, the reader might appreciate an extra Figure A3 that is as Figure A2 but keeping NZ2030 as reference (i.e., NZ2035-NZ2030, NZ2040-NZ2030, …).

Figure 2 – the definition of ‘early stable’ and ‘late stable’ should be indicated more clearly to be based on stable emissions, as temperature are clearly not stable – both in the Figure caption and in the text between lines 220-223.

Figure 2 – The reader might appreciate horizontal reference lines for the three warming levels of 1.5, 2 and 3 °C. Also, to me these periods highlighted in bold do not average out at these warming levels, which I would explain by the fact that a range of +/-0.2°C was taken for the definition of the target GWL in these timeseries.

Figure 2 – If the intention of defining warming levels from the time series was to collect as many years as possible from the available simulation that comply with the +/-0.2°C range around the target GWL, I don’t quite understand why simulated period are only chosen from specific simulations. There are clearly periods from other simulations, that reach into this range, but the specific years/decades were not considered here apparently. For example, the simulation branched at 2040 (third grade of orange) drops into the range of temperatures in the bold period of the simulation branched at 2035 (second grade of orange), but the respective years were not considered in the analysis. Such instances can be found for almost all periods highlighted in bold. Could the authors please clarify why this approach has been taken. If it was simply for simplicity of selecting the timeseries for the analysis, this should be made transparent in the main text.

Lines 224-226 – ‘…, but results may… GWL ensembles.’ – I think this a key part in the assessment and interpretation of the results given the methodology used here. The reader might benefit from the authors providing more detail (or some speculation) on how the described differences may look like and how relevant they are with respect to the results presented here.

Line 227 – ‘The method…’ – I suggest moving this sentence to line 216.

Line 228 – ‘The use of…’ – I suggest moving this sentence to line 218.

Figure 3 – apart from adding the global mean temperature pane as indicated in the comments above, I would suggest plotting panel a as the land-ocean ratio instead of anomalies. Further, I would plot panels c) and d) as anomalies.

Line 284 – ‘, and increase…’ – worth noting that is seems to even stabilize at a non-zero difference.

Lines 222-223 – ‘…, suggest that there are complex changes occurring through the next-zero simulations…’ – In line with the interpretation provided in lines 341-342, could the authors please comment on – and perhaps incorporate in the main text – whether this is purely because of the regional temperature changes shown in Fig. 5 (i.e. relatively stable Arctic Amplification, and much accelerated Antarctic regional warming due to slow Southern Ocean response), or whether there is indication and/or contribution of other irreversible processes such as non-linear ice sheet responses and feedbacks. Comments in that direction are made in the Conclusions Section, but it might be worth to also discuss this here.

Lines 471-472 – This sentence is a repetition of lines 462-464.

Line 495 – see comment for line 150. It might be good to add some discussion on how ACCESS-ESM-1.5 differs to other models in the context of the ENSO response and how relevant this is for the analysis here.

Technical corrections

Axes and/or legend labels are too small in Figures 1, 5, 7, 8, 9, 10, 11, 13, 14, 15. Figure 3 has no x-axis labels, but I assume they would appear too small if they were there, too.

Line 160 – subscript ‘2’ in ‘CO2’.

Line 189 – ‘… temperature difference values …’ – add ‘between the two periods’.

Line 199 – ‘The changing pattern’ – remove ‘changing’.

Line 305 – ‘… in the Arctic, also continues…’ – remove ‘also’.

Line 361 and Line 362 – ‘Figure 4X-X’ should be ‘Figure 7X-X’.

Line 465 – ‘G)’ should be ‘g)’.

Line 503 – ‘are not projected’ – please check the presence of ‘not’ in this sentence. For how I understand this sentence, it should be removed.

Other

1. Does the paper address relevant scientific questions within the scope of ESD?
   
   Yes
2. Does the paper present novel concepts, ideas, tools, or data?
   
   Yes
3. Are substantial conclusions reached?
   
   Yes
4. Are the scientific methods and assumptions valid and clearly outlined?
   
   Yes
5. Are the results sufficient to support the interpretations and conclusions?
   
   Yes
6. Is the description of experiments and calculations sufficiently complete and precise to allow their reproduction by fellow scientists (traceability of results)?
   
   Yes
7. Do the authors give proper credit to related work and clearly indicate their own new/original contribution?
   
   Yes
8. Does the title clearly reflect the contents of the paper?
   
   Yes
9. Does the abstract provide a concise and complete summary?
   
   Yes
10. Is the overall presentation well structured and clear?
    
    Yes, see some specific recommendations for the Methods Section
11. Is the language fluent and precise?
    
    Yes
12. Are mathematical formulae, symbols, abbreviations, and units correctly defined and used?
    
    Yes
13. Should any parts of the paper (text, formulae, figures, tables) be clarified, reduced, combined, or eliminated?
    
    Yes, see some specific recommendations for the Methods Section
14. Are the number and quality of references appropriate?
    
    Yes
15. Is the amount and quality of supplementary material appropriate?
    
    Yes, with the suggestion to add one panel