the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 13 Aug 2024
Regional non-reversibility of mean and extreme climate conditions in CMIP6 overshoot scenarios linked to large-scale temperature asymmetries
Abstract. Overshoot scenarios, in which the forcing reaches a peak before starting to decline, show non-symmetric changes during the CO2 increasing and decreasing phases, producing persistent changes on climate. Non-reversibility mechanisms, associated among others with lagged responses of climate components, changes in ocean circulation and heat transport and changes in the ice cover, bring hysteresis to the climate system. These mechanisms generally have an impact in global scales, potentially generating hemispheric temperature changes and alterations of the Intertropical Convergence Zone (ITCZ). This work analyzes simulations from the Coupled Model Intercomparison Project Phase 6 (CMIP6) to explore the relevance of these mechanisms in overshoot scenarios with different forcing conditions (SSP5-3.4OS and SSP1-1.9) and the impact of these large-scale mechanisms on regional climates, with a particular focus on the degree to which changes in regional extremes are reversible. Results show that non-reversibility of temperature and precipitation extremes mostly occurs during the transition period around the global temperature maximum, when a decoupling between regional extremes and global temperature generates persistent changes at regional level. These changes mainly impact temperature extremes in extratropical regions and precipitation extremes in tropical regions around the ITCZ. In scenarios with strong forcing changes like SSP5-3.4OS, regional non-reversibility can be mostly linked to a temperature asymmetry between Northern and Southern Hemisphere, associated with ITCZ shifts. This asymmetry may be associated with persistent changes in the heat transport and with a different thermal inertia depending on the region, leading regionally to a different timing of the temperature maximum. In scenarios with lower forcing changes like SSP1-1.9, the contribution of this mechanism is more limited and other factors like ice melting may also have a relevant role.
- Preprint
(11595 KB) - Metadata XML
- BibTeX
- EndNote
Status: open (until 24 Sep 2024)
-
RC1: 'Comment on esd-2024-26', Peter Pfleiderer, 27 Aug 2024
reply
The authors present a comprehensive analysis of regional climate signals in the overshoot scenarios SSP5-34OS and SSP1-19. The study provides valuable insights about overshoot implications and is surely of interest for a wider audience. The scope of the presented study is very similar with Pfleiderer et al. 2024 and, as the authors write in their discussion, many of the findings are consistent. Due to different methodological choices, the results of this study are still very relevant and definitely worth publishing. It would, however, be useful to discuss the methodological differences between the two studies and implications in more detail.
The study is well written and I would suggest publication after some clarifications and potential changes in the framing.The authors have chosen to work with ensemble means which is a common approach but has some subtle implications when studying overshoot scenarios. Most importantly, different climate models reach peak warming in different years (fig A1) and in the same way, the periods when GMT stabilizes and the respective period before peak warming differs between climate models. Therefore, some of the signal seen in fig 5, 6 (...) might be influenced by considerably different GMT levels. Example: UK-model is considerably warmer in 2090 as compared to 2030 (see fig A1.). The meaning and interpretation of "ensemble mean" becomes a bit complicated in this case as many different effects (regional hysteresis, differences in GMT trajectories, ...) will be merged in one number. It would be important to discuss these effects and if possible estimate how important they are in comparison.
Besides these technical implications of focusing on the ensemble mean, it is questionable whether ensemble mean differences are useful to inform potential risks. As Pfleiderer et al. 2024 shows (also visible in the appendix of the manuscript) the regional response differs considerably between climate models. Going into the details of model differences might be beyond the scope of this paper, but the authors could consider to show one exemplary regional difference between two models to highlight uncertainty when it comes to overshoot scenarios. In my opinion the uncertainty of climate projections in a cooling climate deserve some special attention as there is no observational data with a forced cooling trend to compare with.
Last comment on the ensemble mean: is there a reason you don't use the ensemble median? I would find the ensemble median more appropriate as it does not give additional weight to single models with strong reactions. I would suggest to check the sensitivity to the choice of ensemble mean / ensemble median and briefly discuss this.
The authors do not mention potential effects of aerosol reductions. On a regional level, changes in aerosol emissions can considerably influence precipitation and temperature. In SSP5-34OS and SSP1-19, besides changes in GHG, aerosol emissions change and some of the regional changes around peak warming might be influenced by aerosols. The authors should at least discuss this caveat/feature of the analyzed scenarios and it's implications on the findings.
Specific comments:
Stippling in all figures: Is this a test performed on the ensemble mean? If yes, it would also be interesting to show model agreement.
L72-73: Pfleiderer et al. 2024 focuses on exactly that question.
Fig 1: Why is there no brown line in the bottom of panel a & c?
L95-102: Just out of interest: do you have these extreme indices for all the models and runs listed in table 1? From my experience, daily data that is required for the computation of these indices is not (easily?) available for all simulations for which monthly tas and pr exists.
L103-107: I would expect that all the results would be quite sensitive to the choice of this period. Therefore, some sensitivity testing and some more discussion of the implications of different GMT levels between pre-overshoot period and the stabilization period would be helpful.
L141-149: For the comparison of the two scenarios, it would be useful to merge fig. 2 and fig. 3. Since the GMT trajectory differs between the two scenarios, it is a bit unclear what conclusions can be drawn from the comparison of the two scenarios when fixed periods are used (as you also show in fig 4).
Fig 4: "obtained as the year after the maximum in which temperature reaches the same value as in the period 2290-2300" -> what is the tolerance for the temperature differences? And are you comparing 20-year periods with the 10-year period at the end (2290-2300)? I think that for this comparison, the period should have the same length. Similar question for e) and f): are you comparing single years, or 20-year periods?
L175-177: How do you interpret that the ITCZ shift is lower in SSP1-19 but precipitation differences are higher?
L193-195: Does the word "being" belong here?
L199-201: Although the slope looks similar, there is a different TXx - GMT relationship after the overshoot. At the same GMT level after the overshoot one would expect a lower TXx value in EN, right?
L251-253: Again, I'm not sure if I would agree. Isn't this statement contradicting L234-236?
Citation: https://doi.org/10.5194/esd-2024-26-RC1
Viewed
| HTML | XML | Total | BibTeX | EndNote | |
|---|---|---|---|---|---|
| 162 | 60 | 13 | 235 | 6 | 6 |
- HTML: 162
- PDF: 60
- XML: 13
- Total: 235
- BibTeX: 6
- EndNote: 6
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1