the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 10 Oct 2023
The interaction of Solar Radiation Modification and Earth System Tipping Elements
Abstract. The avoidance of hitting tipping points is often considered a key benefit of Solar Radiation Modification (SRM) techniques, however, the physical science underpinning this has thus far not been comprehensively assessed. This review assesses the available evidence for the interaction of SRM with a number of earth system tipping elements in the cryosphere, the oceans, the atmosphere and the biosphere , with a particular focus on the impact of SAI. We review the scant available literature directly addressing the interaction of SRM with the tipping elements or for closely related proxies to these elements. However, given how limited this evidence is, we also identify and describe the drivers of the tipping elements, and then assess the available evidence for the impact of SRM on these. We then briefly assess whether SRM could halt or reverse tipping once feedbacks have been initiated. Finally, we suggest pathways for further research. We find that SRM mostly reduces the risk of hitting tipping points relative to same emission pathway scenarios without SRM, although this conclusion is not clear for every tipping element, and large uncertainties remain.
-
Notice on discussion status
The requested preprint has a corresponding peer-reviewed final revised paper. You are encouraged to refer to the final revised version.
-
Preprint
(1217 KB)
-
The requested preprint has a corresponding peer-reviewed final revised paper. You are encouraged to refer to the final revised version.
- Preprint
(1217 KB) - Metadata XML
- BibTeX
- EndNote
- Final revised paper
Journal article(s) based on this preprint
Interactive discussion
Status: closed
-
RC1: 'Comment on egusphere-2023-1753', Anonymous Referee #1, 26 Nov 2023
The paper “The interaction of solar radiation modification with Earth System Tipping Elements” by Futerman et al., explores the current state of the literature of the effects of solar radiation modification (SRM) on climate tipping elements. This is a very timely review that should indeed be carried out in order to inform future research on SRM; therefore, I believe that the authors have a very good reasoning for their review article at this critical point in time. While I can see the time and effort that the authors put in this review article, I found the rationale of the article difficult to grasp and think that several changes are necessary before this article can be considered for publication in Earth System Dynamics. I apologize for my critical feedback and I hope the authors will find the comments below helpful for their decision on how to proceed.
Key points:
- Streamlining of the article: Throughout the article, each of the tipping elements is split up into four parts: (i) Current state of the tipping element; (ii) Drivers and Feedbacks; (iii) Impacts of SRM; (iv) Further research. While I think that the core of the paper is in (iii), I do not think that another review on the current state of the tipping elements as well as drivers and feedbacks is necessary (not the main topic of this paper, is it?) or helpful (the current state of the tipping elements is done in the global tipping points report and the many other papers of this special issue). Therefore, I suggest to strongly cut those parts (maybe put the long version in an SI if the authors would like to keep this information). For instance, the section on Mountain Glaciers (2.3) had a good balance between the subsections in my view.
This is not to say that I didn’t find this information valuable and interesting to read, but it distracts from the main purpose of this article (SRM) in my view. - Rationale of the article: The above (my point 1) is partially reflected in the abstract where the authors state that they review the literature on SRM but given that there is not a lot of literature they state that additionally a review of the current state of tipping elements as well as their drivers and feedbacks was carried out. I think this rationale is invalid and needs to be sharpened. In my view, a review on SRM on tipping elements is warranted even if the current state of literature is still immature and sparse in some places. This should be reflected in the abstract and the text overall. It can, however, not be an excuse to add only partially related parts to a paper because literature evidence of a certain aspect is weak.
- Main results/Table 1: I think this is the main outcome of the paper and I like it very much (with some smaller suggestions below). This table should be placed in the beginning and then elaborated in the specific sections. This gives the reader a clear picture of the main results early on in the article instead of after 47 pages of text – I believe this table is also not announced before so that the reader could anticipate. I also strongly recommend to add at least one further figure that graphically represents the table, e.g. on a world map populated with the tipping elements and their reaction to SRM.
Overall, I believe that the manuscript would profit a lot from a clearer focus on SRM. Some concrete points below.
Major points:
- I find that section 1.3 (Solar Radiation Modification and Tipping Elements) could be streamlined strongly.
- While figure 1 is illustrative, I am unsure whether this is really useful in this manuscript because it is not a result from this paper but a reproduction.
- In my view, this is similar for figure 2 as long as there are no studies that discuss how SRM directly alters Atlantic Ocean circulations; if there are, please add them to the figure.
- Figure 3 is a general figure of S-shapes that allows for different types of tipping (forcing, noise, rate, …). As such, this figure is not specific to the AMOC and should either be removed or moved up to the introduction. It may actually be a good figure to discuss threshold-free feedbacks as opposed to different types of tipping in the introduction.
- Page 12, l 322-346: Can be strongly shortened in my opinion. In particular if there are not many studies that discuss SRM, then this should be stated, and additional CDR studies (e.g. Garbe et al., 2020) should be kept very brief as this is not the main contribution of the paper as I understand.
- Pages 32-35: The impact of SRM on ecological systems in general: As opposed to the other sections which are excellently referenced, this section is not. Further, my overall feeling is that this section can be condensed to around 20-35% of text lengths.
- Section 5.2: Dipterocarp Forests: This is not a global tipping element. Why is it discussed in this manuscript? Do the authors suggest that it should be considered a climate tipping element because it is relevant on the global scale? For me, it sounds regionally very relevant but more like a super-regional regime shift rather than a global tipping element (e.g. see Rocha et al., 2018, Science: 1126/science.aat7850). Maybe this section can be moved to the SI or removed.
- Section 5.5: Indian subcontinent biodiversity hotspots: It is unclear to me why this section is included because (i) it is not a climate tipping element and (ii) there are also hotspots of biodiversity in Africa, Indonesia and particular in South America. Therefore, I suggest to move this section in the SI of the paper or remove it.
- Comments to Table 1, which is really helpful:
- Column: Overall confidence of what is meant here? Overall confidence of SRM being able to reverse tipping? Overall confidence (=agreement) of the literature on column “Can SAI reverse tipping”
- Rows “Dipterocarp Forest” and “Indian Subcontinent Biodiversity Hotspots”: How can SRM help once tipping is completed (as noted in “b. Likely” in column 4). Once biodiversity is lost or a forest has died back, SRM cannot help to restore these systems as they have developed and adapted over millions of years.
- Section 6.1 (Further Research): I suggest to keep these sections short because for each of the tipping elements, it is already discussed where future research can broaden knowledge.
There is a number of smaller points - Minor points:
- Abstract: Please write SAI in the abstract out at first occurrence
- Page 2, line 45: Also cite Levermann et al., 2012: https://doi.org/10.1007/s10584-011-0126-5
On page 846 is a definition of tipping elements - Page 2, line 53-56: The definition of ecological tipping elements is unclear. In Armstrong McKay, tipping of ecosystems means a large-scale state shift of the Amazon rainforest, boreal forests, or coral reefs. The death of single species does not constitute a tipping in the Earth system sense. Please clarify.
- Page 2, l 59: Can you directly here give an example of a threshold-free feedback? (maybe this is a good spot for figure 3)
- Page 3, l 79: “… has been exceeded for sufficiently long times” What you mean are so-called overshoots. Replace citations by
- Ritchie et al., 2021: https://doi.org/10.1038/s41586-021-03263-2
- Wunderling et al., 2023: https://doi.org/10.1038/s41558-022-01545-9
- Page 4, l 99: Irvine et al., 2019 does not exist in the reference list but only Irvine et al., 2018. Please check.
- Page 7, l 184, cite Levermann, Winkelmann, 2016, The Cryosphere: https://doi.org/10.5194/tc-10-1799-2016
- Page 13, l 369: Did the authors really mean 2089, not 2099?
- Page 26/27: Formatting changes twice, please check.
- Page 31, l 894: The Schneider et al., 2019 paper on a 10°C warming due to cloud changes is speculative (given its huge temperature feedback). This should be stated somewhere around these lines.
- Page 37, chapter 5.3. Amazon basin: In the introductory paragraph, the combined adverse influence of deforestation, human-made fires, and climate change on the Amazon rainforest could be discussed more directly.
- Somewhere in this section, the new Bochow et al., 2023, Science Advances paper should be cited: 1126/sciadv.add9973
- I believe MCB (probably Marine Cloud Brightening) is only used as an abbreviation
- Page 46: I think this study by Rao et al., 2023 in Communications Earth and Environment should be discussed briefly in this section (https://doi.org/10.1038/s43247-023-00910-6)
Citation: https://doi.org/10.5194/egusphere-2023-1753-RC1 - AC1: 'Reply on RC1', Gideon Futerman, 26 Jan 2024
- Streamlining of the article: Throughout the article, each of the tipping elements is split up into four parts: (i) Current state of the tipping element; (ii) Drivers and Feedbacks; (iii) Impacts of SRM; (iv) Further research. While I think that the core of the paper is in (iii), I do not think that another review on the current state of the tipping elements as well as drivers and feedbacks is necessary (not the main topic of this paper, is it?) or helpful (the current state of the tipping elements is done in the global tipping points report and the many other papers of this special issue). Therefore, I suggest to strongly cut those parts (maybe put the long version in an SI if the authors would like to keep this information). For instance, the section on Mountain Glaciers (2.3) had a good balance between the subsections in my view.
-
RC2: 'Comment on egusphere-2023-1753', David Armstrong McKay, 13 Dec 2023
Review of "The interaction of Solar Radiation Modification with Earth System Tipping Elements" (egusphere-2023-1753) for Earth System Dynamics
Paper summary:
In this paper, the authors present a review of scientific literature around the potential for solar radiation modification (SRM) to prevent tipping points (TPs) in the Earth system, and in particular climate tipping points (CTPs). While the risks posed by CTPs have been suggested as a motivation for considering SRM, studies on the specifics of how SRM might interact with tipping dynamics in parts of the climate system liable to tip (i.e. tipping elements/systems) have only recently started to be undertaken. In many cases no such specific studies exist, although more general research on how SRM affects feedbacks involved in tipping might be available instead. The authors collate direct or indirect studies relating to potential SRM impacts on tipping systems and their ability to mitigate CTPs, covering commonly proposed tipping systems in the cryosphere, ocean, atmosphere, and biosphere. They then assess whether the collated evidence supports SRM being able to prevent tipping in those systems to some extent, as well as whether tipping can be countered once in progress. They find that while many systems have some evidence that tipping drivers can be partly compensated for using SRM, confidence is generally low and research is currently lacking in many areas, for which future research avenues are suggested for improved study scenarios and models.
General comments:
In general this is a timely paper, given increasing attention on the risks posed by climate tipping points and proposals that this necessitates SRM, but only a small (but recently growing) pool of targeted research in this area. Reviewing current literature and identifying gaps to target with further research is therefore a welcome exercise that will hopefully spur on more research in this area, and easily falls within the scope of this journal.
Tipping systems and their hypothesised dynamics are in general accurately portrayed, though in places could variously do with more detail or being more concise (see specific comments for details). There are some places too where the latest papers should be included (e.g. on Antarctica) as the pace is picking up in this field now with some relevant papers published since submission. In several places tighter writing, reducing repetition, and more consistent structuring would help improve flow, which would also make the paper more concise (it’s not overly long at present, but ideally it shouldn’t get longer, and shorter is preferable for readability if possible).
It’s good to have put SRM discourse in context upfront, and the complexities around SRM not cancelling out all climate change as easily as GMST. However, I think there could be more discussion of how applying SRM in ESMs (particularly for slightly older SRM studies) that lack key tipping-relevant processes and feedbacks means we may be missing some key SRM-CTP interactions, such as the effects of SRM-induced precipitation or ecological changes. For example, ESMs are likely biased towards AMOC stability (per IPCC), lack spatial heterogeneity in ecohydrological dynamics in the Amazon (where many ESMs disagree on even sign of precip change too), or lack abrupt thaw or interactions with surface vegetation in permafrost. In some cases this could offset part of the SRM compensation of tipping, and in some cases potentially even worsen tipping risk (e.g. that one study shows worsening of Amazon droughts with SRM vs. emission scenario without SRM). To me this means the conclusion in the abstract that “We find that SRM mostly reduces the risk of hitting tipping points relative to same emission pathway scenarios without SRM” needs additional clarification that this is mostly relative to temperature, with other climatic factors that are not so well understood likely complicating and potentially undermining this risk reduction.
Beyond the often discussed biospheric tipping systems of Amazon rainforest, boreal forests, and warm-water coral reefs, the authors also select dipterocarp forests in Southeast Asia and various ecosystems in South Asia as potential tipping systems. However, for some examples such as dipterocarp forests, eastern Himalayas, or the Western Ghats their potential for self-sustaining regime shifts is not currently clear from the presented evidence. Additionally, it’s not clear why for example the Sundarbans are highlighted but not the mangrove biome more widely, or why other biomes like savannahs/grasslands or non-coral ocean ecosystems aren’t analysed. If this is because these are not covered by existing studies then that’s OK but should be made clear, otherwise it seems a few examples of highly biodiverse ecosystems have been picked and the possibility of tipping in them speculated. I suggest being clearer on reasons for inclusion and their potential tipping dynamics, or adjusting review framing to be explicitly about whether SRM can prevent tipping dynamics plus biodiversity loss in some selected examples. The latter could be a whole other (and interesting) paper though, so I would suggest focusing on where tipping dynamics can be shown to be possible or likely.
Specific comments:
Line no.
12-13: Is preventing CTPs considered a key benefit of SRM that often in literature? My reading of literature is that it has increasingly been suggested in some papers/proposals, but I'm not sure preventing CTPs is so often made central yet.
16: SAI not yet defined.
20-22: Given the large uncertainties include non-temperature factors such as precipitation changes potentially countering or even cancel out tipping-compensation by SRM (particularly for biosphere, where hydrological effects makes SRM very uncertain and e.g. some experiments show worsening of Amazon droughts with SRM), I think this statement could do with some clarification. To me, this review shows that while direct SRM-CTP studies are few or missing for many systems, where studies have been done it appears SRM can reduce tipping risk specifically with respect to regional or global temperature (with high uncertainties), but current model limitations mean SRM's influence on tipping dynamics via other climatic factors could reduce or even nullify risk reductions.
29: AM22 is specifically about climate TPs rather than wider Earth system TPs, which is likely the case here too given regional/global temperature trigger focus.
50: I think self-sustaining or self-perpetuating is more accurate - the changes can be quite steady for a long time rather than accelerating in some cases (e.g. ice sheet collapse).
55: Is the TP here the extinction itself or events leading to it as TP? I'm not sure the former could be seen as a TP, more a general threshold response, as there's no clear self-sustaining change dynamic there (beyond maybe a population bottleneck / functional extinction effect in final stages). Extinction can certainly lead to ecosystem regime shifts though, which can be a TP if the regime shift is self-sustaining in some regard (e.g. such that hysteresis occurs).
73-75: This is a key point (along with e.g. lines 143-146), and links to my hesitation elsewhere about confidence in SRM’s compensation for CTPs being somewhat temperature-centric.
79: A more specific reference here for tipping overshoot would be Ritchie et al (2021) [https://www.nature.com/articles/s41586-021-03263-2].
105-107: Good to have put SRM in wider context upfront.
113-114: Presumably there’s not so much research out there for attempts at regional cooling.
115: Is “resembling” here these references, or is this a sentence fragment?
183: "Greenland Ice Sheet " would be clearer than “Greenland tipping element”.
200-203: I think this would be better rearranged so explanation of rebound comes first.
208-210: I suspect the ice sheet components in the models likely used for these studies were lacking in terms of representing enough ice sheet dynamics to capture tipping, which is a limitation (and taps in to my general comment that SRM studies using ESMs that lack key tipping dynamics won’t be getting the full picture of how SRM effects tipping dynamics).
227-228: Specify by 2100 – much more sea level rise beyond this even on low pathways.
230-231: Recent research from Li et al (2023) [https://www.nature.com/articles/s41586-023-05762-w] would be a good reference inclusion here.
232: “currently driven” – air temperature could become more important in future, depending on circulation changes.
238: East Antarctica is conventionally split in to marine basins and land-based, as they face different dynamics that lead to quite different tipping thresholds and timescales.
240-243: This sentence is a bit convoluted to me, consider rearranging.
251-252: Missing here is the sense of where the tipping point might be, which is generally the point at which retreating grounding line reaches main retrograde slope leading to accelerating discharge (subject to buttressing, pinning, etc.).
276-278: Probably a bit higher warming needed for East Antarctic basins than WAIS based on e.g. Garbe et al. 2020 (more like 3-6C for Wilkes land for example, though I’d agree the risk starts growing from 2C).
287: A key reference on issues around MICI is Edwards et al 2019 [https://doi.org/10.1038/s41586-019-0901-4]
290: Clearer to say “both are preceded by” ice shelf disintegration.
209: 32cm sea level rise by when? It would also be useful to put these values in context by stating WAIS total sea level equivalent (3+m) earlier.
308: Given e.g. Goddard et al. 2023 is discussed below and there are some systems with fewer studies, "relatively few" seems more accurate now. Another recent release to consider in this section too is Sutter et al. (2023) [https://www.nature.com/articles/s41558-023-01738-w].
309-310: Presumably this result is specific to surface air temperature – useful to specify this in context of limited short to medium term impact on ocean warming.
326: Garbe et al. (2020) is not specifically a CDR experiment – might need to rephrase above for clarity.
329: True, but how close it recovers by present day levels depends on experiment type (equilibrium vs “quasi-static”). Also, do you mean current warming of ~1.2C or pre-industrial by present levels?
334-335: Feels like this should be discussed earlier in this subsection given it's such a critical point, before moving on to what changing SAT might be able to do as a second-order driver and then what reversal/CDR studies show.
358-359: Could probably do with a little more detail on and beyond melt-elevation and melt-albedo feedbacks here (these two can signpost back to ice sheets for more info), e.g. role of changing snow patterns and black carbon/dust on albedo, thermokarst interaction, slope instabilities, negative feedbacks from retreat to higher altitudes or debris insulation, etc., as well as that unlike ice sheets these feedbacks happen on a largely local to regional scale.
368-369: Are G3 and G4 defined before this point (for readers not familiar with GeoMIP terminology or context)?
375-377: A Himalayan example is given here, but can anything be said so far about how heterogenous SRM precipitation impacts may affect different glacier regions, e.g. from GeoMIP? If not, then can be explicit in that.
397: SAM not defined before now.
422: Specify what sort of temperature, e.g. surface air, sea surface, regional/global etc.
447-448: Could clarify here that while some positive feedbacks exist that have been suggested could drive tipping dynamics (e.g. https://link.springer.com/article/10.1007/s10584-011-0126-5; https://journals.ametsoc.org/view/journals/clim/34/11/JCLI-D-20-0558.1.xml) the explanation given here is generally more favoured.
476-479: Presumably this is relative to a scenario of continued warming without SRM, rather than versus no further warming in above example? (otherwise this statement could be seen as conflicting with a reduction in March extent above.)
485: Missing "which" after citation
498-500: I think this could do with clarification - compared with continued warming on same emissions pathway but without SRM?
512: Clarify as (presumably) soil temperature.
523: Perhaps useful to clarify this as melt of ice blocks/wedges, and add thermokarst/talik development (which is not necessarily ice melt dependent).
536: Perhaps worth clarifying that in AM22 this was an additional threshold beyond the more general (and higher confidence) widespread localised abrupt thaw above ~1.5C.
549: Missing in this section (reflecting the point made under further research) is a discussion of how current generation models don't capture non-gradual thaw processes, so currently these conclusions primarily relate to gradual, non-tipping thaw only, or how changes in e.g. precipitation interact with abrupt thaw or eco-hydrological processes.
580-581: SRM is also unlikely to reverse abrupt thaw processes like thermokarst/talik development once started, but can limit formation of new taliks.
604: Could specify very long timescales is necessary because of the centennial-millennial timescale of ocean heat uptake and circulation.
616: A useful figure, but why no equivalent for cryosphere or biosphere systems? A generic schematic showing e.g. elevation/albedo feedbacks might help illustrate key feedbacks there. There could also more focus specifically on how SRM might intervene in those feedbacks to make figures more specific to this article (rather than textbook-style background).
622: Deep convection should also be marked in Southern Ocean, as it’s not separate to sinking there.
628: Might need to clarify deep convection/sinking, which is illustrated separately in figure but are really part and parcel (with "sinking" a simplification of deep convection).
640-641: Relevant here is the IPCC's assessment of collapse unlikely before 2100, but also that CMIP models are biased towards stability and lack key aspects (e.g. ice sheet meltwater) that likely lead to under-estimates of weakening and collapse. This is why unrealistic hosing is used in some studies (like Jackson et al. 2023 below) to overcome this tendency. This is mentioned in next paragraph, but think it’s useful context within this paragraph.
652: Jackson et al. 2022 is the preprint, can update to final 2023 paper now.
660: Liu et al 2017 relevant here too [https://www.science.org/doi/10.1126/sciadv.1601666]
662: Another angle beyond palaeo and model evidence not covered here is observations, with the debate over current slowdown (with IPCC AR6 having low confidence in some weakening) and of potential early warning signals suggestive of destabilisation [e.g. https://www.nature.com/articles/s41558-021-01097-4; https://www.nature.com/articles/s41467-022-32704-3; https://www.nature.com/articles/s41467-023-39810-w]
677-678: Ice sheet runoff is quite the important missing factor though, which is missing from most ESM simulations of AMOC weakening/collapse (and therefore SRM studies using those too).
700: One way of exploring possible SRM effects on AMOC would be to explore the role of past/current aerosol emission effects on AMOC, which have been suggested to generally reduce AMOC tipping likelihood (and is being partly reversed over North Atlantic over recent years) and so could effectively be a smaller-scale SRM analogue.
705: There’s a question mark over whether precip changes from SRM would help or hinder though (though admittedly it seems to be a small factor, model limitations permitting).
721-723: This is a key point, and relates to my general comment on uncertainty inherent in using models with limited tipping dynamics representation to project SRM effects on CTPs.
728-729: This is a very good point, and may be applicable to a other tipping points too, as while many have generic warming level thresholds estimated several are likely to have rate-dependent too [e.g. Ritchie et al. 2023 https://esd.copernicus.org/articles/14/669/2023/]. Could be something to highlight need for further study of in discussion.
733-738: This paragraph largely repeats previous one.
748-749: Indeed – a subset of this risk is that if SRM was initiated to protect the AMOC but it turns out tipping has already commenced (as exactly when thresholds are crossed is uncertain and there's a time lag in tipping dynamics) then SRM might have to be phased down or out to avoid over-cooling North Atlantic region, but at potential cost of unprotecting other regions and CTPs. Relates to potential tricky balancing of how to target mitigating multiple CTPs and other climate impacts without worsening some over others.
761: Missing parenthesis after “injection points”.
808: Given similar issues to AMOC could merge with further research there for a clearer structure, in a similar way to how land ice further research is grouped together.
823: Missing fullstop before citation?
831: Relevant here is new paper Li et al (2023) [https://www.nature.com/articles/s41586-023-05762-w] which projects strong weakening with further warming (RCP8.5 in that study, but likely similar magnitude decline for RCP4.5 on longer timescales)
834: Sentence would be clearer if what the mechanism is for is specified (overturning weakening? collapse?)
839: Contrast Li et al, who suggest wind has marginal effect on projected weakening (though possibly model dependent).
862: Only marine stratocumulus clouds are tackled in this section, and not e.g. monsoons which are sometimes considered atmospheric TPs, though possibly more due to aerosols than climate change proper. Given the relation of monsoon tipping to aerosols, and recent work on interhemispheric AOD difference as a key safe boundary (Rockström et al. 2023), their relation to SRM would be an interesting avenue to explore if resources permit.
894: Warming values repeated from past paragraph (not bad, but unnecessary to make hysteresis point).
920: What does this mean - the feedback is fully counteracted, or literally turned off?
948: This general intro to ecosystem TPs & SRM is useful in general, but is repetitive in parts and could be tightened.
953-954: In my experience extinction is not normally thought of as a TP / regime shift in ecology (although extinction or replacement of a keystone species could help trigger a wider regime shift).
965-967: Clarify – the impact/event itself was not a TP (just an abrupt exogenous forcing with an equally abrupt response), but it likely drove many localised/regionalised ecosystem regime shifts (some but not all of which might have featured self-sustaining tipping dynamics).
1010-1013: This sentence can be simplified for clarity.
1024: To me this is conceptually confused – "forests" and "biodiversity" are very different categories of things (one a collection of objects forming a system, the other a property of such a system).
1025-1030: These sentences are a bit confusing and could do with tightening.
1051-1052: I'd like to see more evidence presented on how these forests are susceptible to tipping (via regeneration failure) that justifies pulling it out as a specific unit to analyse. In analyses of tropical rainforest-rainfall feedbacks, rainforests in maritime Southeast Asia are normally found to be more robust due to plentiful ocean rainfall sources, so it can't be that mechanism, and syncrhonised seeding occurs in many forest biomes, particularly tropical but including to some extent temperate too. Numata et al. is a key citation here, but it's not apparent that regime shifts per se are projected in that study. An alternative would be too merge in with discussion of other tropical forests.
1073-1075: I couldn't find this in the given reference.
1078: Clarify where.
1082-1083: Assumes there is a TP in this system – if this is uncertain though could reframe this section as about protecting a particular biome that may or may not feature tipping dynamics (but then a question is why this ecosystem and not various others).
1086: Missing a "Further Research" subheading here?
1097: This surprises me, as Amazon is normally considered to be one of the most biodiverse biomes (when considered as a biome unit). The cited analysis doesn't highlight Amazon because the "hotspots" are a product both of endemism and threat level, with Amazonia seemingly left out because of lower threat rather than lower endemism (i.e. hotspots in Myers et al 2000 don't mean that's where most biodiversity is, but where areas of high biodiversity are most threatened – it also excludes invertebrates, and any newly identified species in past 23 years). Also, this citation is not in the reference list (I’m not sure it adds too much to discussion anyway).
1101-1102: 2-6C is for just climate change without anthropogenic co-drivers like deforestation – including those could lower it further.
1104: More recent studies highlight that degraded forest is a likely alternative state (possibly more common than savannah).
1123-1124: Some discussion of where dieback is more likely – in the drier south and east, as per bistable areas in Staal et al. 2020 – would be useful too for understanding the feedbacks and areas most at risk.
1143: Could make a link here to AMOC collapse too, which is projected to shift ITCZ southwards and so cause drying in Northern Amazonia (likely similar to in Younger Dryas). Also some link to monsoon shifts and global aerosol patterns.
1150-1151: Even here GCMs often struggle to agree on even the sign of regional precip change, which is one of the reasons ESMs tend to disagree on likelihood and scale of Amazon dieback.
1165: "the utility... is of limited utlity" – rephrase to e.g. "Thus, existing studies...".
1194: I think representing land use change, and spatial heterogeneity in plant traits adaptivity across the Amazon is also key for better model representation.
1205-1206: This probably needs a citation [e.g. https://journals.plos.org/plosone/article?id=10.1371/journal.pone.0025026]
1249: Critical too in some regions like the Caribbean is disease and invasive species spread (often facilitated by warming and globalised trade).
1268: Some numbers here would be useful to demonstrate this if possible.
1278: Recurrence in between hurricanes?
1291: While the basis for SRM effects on coral bleaching is better understood than for some tipping systems, given the fair few uncertainties explored above relating e.g. to extreme events, complex co-drivers, or variation in SRM impacts between different coral regions makes it seem not quite so strong to me. Not a huge amount of dedicated studies are cited above either. Personally I'd like to see studies with up-to-date ESMs exploring where and how much SRM is necessary to e.g. keep below specific Degree Heating Week or recurrence thresholds, how this affects storms, and the spatial heterogeneity of these.
1303-1304: Are these are all vulnerable to tipping? Not all ecosystems have alternative stable states to tip to – they can just become gradually degraded – so it's important to present or cite evidence for self-sustaining regime shifts. The Sundarbans are the clearest example here, with coastal erosion and salinity effects likely to drive tipping dynamic across mangrove biome (GTPR, 2023), but given they're at risk of regime shifts across the tropics why is this the only locality discussed?
1311: Here as in above in text?
1315-1317: Maybe better discussed in SRM effects subsection below?
1329: In other subsections the drivers and mechanisms of a specific tipping system is analysed, while here it's for multiple different systems within the same region, which is somewhat inconsistent in paper structuring. Additionally there's some overlap with other sections, e.g. for glaciers.
1334-1336: Specify what system this tipping point is in – downstream ecosystems? Peri-glacial environment around glaciers?
1343-1353: This paragraph reads rather fragmented to me. Additionally, I don’t think the TPs on line 1345-1346 have been explicitly described yet.
1374-1378: Possibly worth noting that this hypothesis (or at least magnitude of effect) is debated.
1388-1390: Worth noting that biogeophysical effects can have opposite regional climate impacts to the global climate effect (which could potentially interact with SRM) – e.g. boreal dieback to steppe increasing albedo and regional cooling, versus carbon released by dieback increasing global warming.
1394: Given complex eco-hydrological mechanisms in boreal forest dynamics this subsection could do with mention of how (likely uncertain) impacts of SRM on precip might complicate temperature-based considerations.
1403: It seems that there's no specific SRM studies re. boreal forests – if so, should highlight that as a knowledge gap.
1417-1422: Minor point, but not sure these sentences are necessary as they mostly repeat the Introduction.
1426-1428: Useful point to mention –potentially relevant to highlight in abstract.
Table 1: Is b) in the 4th column linked to a specific timescale? GrIS, AIS, and permafrost here are Not Reversible by SAI once tipping is complete, which is true on centennial timescales but not necessarily very long timescales, while AMOC collapse is Yes with hysteresis, when on long enough timescales that could apply to the former examples too. I’m also not entirely sure if AMOC/SPG collapse can likely be reversed once tipping has begun (depends on feedback dynamics, which we don’t have a great hold on right now). Also, is “Uncertain” in last row 4th column for both a and b?
1441-1442: These partial compensations are mostly with regards to temperature though – I would hazard that greater uncertainty / lower confidence would hold on this partial compensation once accounting for impacts on other climate factors such as precipitation. At the moment the Discussion and Table 1 doesn't highlight this.
1444-1447: These points are repeated two paragraphs down.
1445: I don’t think peak-shaving hasn’t been mentioned prior to this point, so could do with an explanatory citation on first instance.
1449-1453: This overlaps discussion of emergency use and hysteresis in paragraph below – could probably tighten discussion to be more concise.
1475: Could specify biosphere/ecosystem TPs here (e.g. biomass plantation based CDR making forest dieback or savannah degradation more likely).
1477: A key issue and source of uncertainty for me is that we know most about how SRM might effect CTPs via temperature, but much less about other climatic factors like precipitation, which in many cases could reduce or even overwhelm the compensation delivered by reducing warming (e.g. less precip delivered to ice sheets, or cooling accompanied by droughts in tropical forests). This relates to the research approaches given below – improving modelling of Earth system tipping dynamics is critical to be able to give a fuller assessment of how different SRM schemes might affect tipping processes.
1478-1481: This is not very clear to me at present – is this saying that these Qs wouldn't be relevant to people who think SRM research shouldn't be pursued at all because of moral hazard issues, or more the issues raised in the penultimate paragraph?
Dr. David A. McKay
Citation: https://doi.org/10.5194/egusphere-2023-1753-RC2 - AC2: 'Reply on RC2', Gideon Futerman, 26 Jan 2024
Interactive discussion
Status: closed
-
RC1: 'Comment on egusphere-2023-1753', Anonymous Referee #1, 26 Nov 2023
The paper “The interaction of solar radiation modification with Earth System Tipping Elements” by Futerman et al., explores the current state of the literature of the effects of solar radiation modification (SRM) on climate tipping elements. This is a very timely review that should indeed be carried out in order to inform future research on SRM; therefore, I believe that the authors have a very good reasoning for their review article at this critical point in time. While I can see the time and effort that the authors put in this review article, I found the rationale of the article difficult to grasp and think that several changes are necessary before this article can be considered for publication in Earth System Dynamics. I apologize for my critical feedback and I hope the authors will find the comments below helpful for their decision on how to proceed.
Key points:
- Streamlining of the article: Throughout the article, each of the tipping elements is split up into four parts: (i) Current state of the tipping element; (ii) Drivers and Feedbacks; (iii) Impacts of SRM; (iv) Further research. While I think that the core of the paper is in (iii), I do not think that another review on the current state of the tipping elements as well as drivers and feedbacks is necessary (not the main topic of this paper, is it?) or helpful (the current state of the tipping elements is done in the global tipping points report and the many other papers of this special issue). Therefore, I suggest to strongly cut those parts (maybe put the long version in an SI if the authors would like to keep this information). For instance, the section on Mountain Glaciers (2.3) had a good balance between the subsections in my view.
This is not to say that I didn’t find this information valuable and interesting to read, but it distracts from the main purpose of this article (SRM) in my view. - Rationale of the article: The above (my point 1) is partially reflected in the abstract where the authors state that they review the literature on SRM but given that there is not a lot of literature they state that additionally a review of the current state of tipping elements as well as their drivers and feedbacks was carried out. I think this rationale is invalid and needs to be sharpened. In my view, a review on SRM on tipping elements is warranted even if the current state of literature is still immature and sparse in some places. This should be reflected in the abstract and the text overall. It can, however, not be an excuse to add only partially related parts to a paper because literature evidence of a certain aspect is weak.
- Main results/Table 1: I think this is the main outcome of the paper and I like it very much (with some smaller suggestions below). This table should be placed in the beginning and then elaborated in the specific sections. This gives the reader a clear picture of the main results early on in the article instead of after 47 pages of text – I believe this table is also not announced before so that the reader could anticipate. I also strongly recommend to add at least one further figure that graphically represents the table, e.g. on a world map populated with the tipping elements and their reaction to SRM.
Overall, I believe that the manuscript would profit a lot from a clearer focus on SRM. Some concrete points below.
Major points:
- I find that section 1.3 (Solar Radiation Modification and Tipping Elements) could be streamlined strongly.
- While figure 1 is illustrative, I am unsure whether this is really useful in this manuscript because it is not a result from this paper but a reproduction.
- In my view, this is similar for figure 2 as long as there are no studies that discuss how SRM directly alters Atlantic Ocean circulations; if there are, please add them to the figure.
- Figure 3 is a general figure of S-shapes that allows for different types of tipping (forcing, noise, rate, …). As such, this figure is not specific to the AMOC and should either be removed or moved up to the introduction. It may actually be a good figure to discuss threshold-free feedbacks as opposed to different types of tipping in the introduction.
- Page 12, l 322-346: Can be strongly shortened in my opinion. In particular if there are not many studies that discuss SRM, then this should be stated, and additional CDR studies (e.g. Garbe et al., 2020) should be kept very brief as this is not the main contribution of the paper as I understand.
- Pages 32-35: The impact of SRM on ecological systems in general: As opposed to the other sections which are excellently referenced, this section is not. Further, my overall feeling is that this section can be condensed to around 20-35% of text lengths.
- Section 5.2: Dipterocarp Forests: This is not a global tipping element. Why is it discussed in this manuscript? Do the authors suggest that it should be considered a climate tipping element because it is relevant on the global scale? For me, it sounds regionally very relevant but more like a super-regional regime shift rather than a global tipping element (e.g. see Rocha et al., 2018, Science: 1126/science.aat7850). Maybe this section can be moved to the SI or removed.
- Section 5.5: Indian subcontinent biodiversity hotspots: It is unclear to me why this section is included because (i) it is not a climate tipping element and (ii) there are also hotspots of biodiversity in Africa, Indonesia and particular in South America. Therefore, I suggest to move this section in the SI of the paper or remove it.
- Comments to Table 1, which is really helpful:
- Column: Overall confidence of what is meant here? Overall confidence of SRM being able to reverse tipping? Overall confidence (=agreement) of the literature on column “Can SAI reverse tipping”
- Rows “Dipterocarp Forest” and “Indian Subcontinent Biodiversity Hotspots”: How can SRM help once tipping is completed (as noted in “b. Likely” in column 4). Once biodiversity is lost or a forest has died back, SRM cannot help to restore these systems as they have developed and adapted over millions of years.
- Section 6.1 (Further Research): I suggest to keep these sections short because for each of the tipping elements, it is already discussed where future research can broaden knowledge.
There is a number of smaller points - Minor points:
- Abstract: Please write SAI in the abstract out at first occurrence
- Page 2, line 45: Also cite Levermann et al., 2012: https://doi.org/10.1007/s10584-011-0126-5
On page 846 is a definition of tipping elements - Page 2, line 53-56: The definition of ecological tipping elements is unclear. In Armstrong McKay, tipping of ecosystems means a large-scale state shift of the Amazon rainforest, boreal forests, or coral reefs. The death of single species does not constitute a tipping in the Earth system sense. Please clarify.
- Page 2, l 59: Can you directly here give an example of a threshold-free feedback? (maybe this is a good spot for figure 3)
- Page 3, l 79: “… has been exceeded for sufficiently long times” What you mean are so-called overshoots. Replace citations by
- Ritchie et al., 2021: https://doi.org/10.1038/s41586-021-03263-2
- Wunderling et al., 2023: https://doi.org/10.1038/s41558-022-01545-9
- Page 4, l 99: Irvine et al., 2019 does not exist in the reference list but only Irvine et al., 2018. Please check.
- Page 7, l 184, cite Levermann, Winkelmann, 2016, The Cryosphere: https://doi.org/10.5194/tc-10-1799-2016
- Page 13, l 369: Did the authors really mean 2089, not 2099?
- Page 26/27: Formatting changes twice, please check.
- Page 31, l 894: The Schneider et al., 2019 paper on a 10°C warming due to cloud changes is speculative (given its huge temperature feedback). This should be stated somewhere around these lines.
- Page 37, chapter 5.3. Amazon basin: In the introductory paragraph, the combined adverse influence of deforestation, human-made fires, and climate change on the Amazon rainforest could be discussed more directly.
- Somewhere in this section, the new Bochow et al., 2023, Science Advances paper should be cited: 1126/sciadv.add9973
- I believe MCB (probably Marine Cloud Brightening) is only used as an abbreviation
- Page 46: I think this study by Rao et al., 2023 in Communications Earth and Environment should be discussed briefly in this section (https://doi.org/10.1038/s43247-023-00910-6)
Citation: https://doi.org/10.5194/egusphere-2023-1753-RC1 - AC1: 'Reply on RC1', Gideon Futerman, 26 Jan 2024
- Streamlining of the article: Throughout the article, each of the tipping elements is split up into four parts: (i) Current state of the tipping element; (ii) Drivers and Feedbacks; (iii) Impacts of SRM; (iv) Further research. While I think that the core of the paper is in (iii), I do not think that another review on the current state of the tipping elements as well as drivers and feedbacks is necessary (not the main topic of this paper, is it?) or helpful (the current state of the tipping elements is done in the global tipping points report and the many other papers of this special issue). Therefore, I suggest to strongly cut those parts (maybe put the long version in an SI if the authors would like to keep this information). For instance, the section on Mountain Glaciers (2.3) had a good balance between the subsections in my view.
-
RC2: 'Comment on egusphere-2023-1753', David Armstrong McKay, 13 Dec 2023
Review of "The interaction of Solar Radiation Modification with Earth System Tipping Elements" (egusphere-2023-1753) for Earth System Dynamics
Paper summary:
In this paper, the authors present a review of scientific literature around the potential for solar radiation modification (SRM) to prevent tipping points (TPs) in the Earth system, and in particular climate tipping points (CTPs). While the risks posed by CTPs have been suggested as a motivation for considering SRM, studies on the specifics of how SRM might interact with tipping dynamics in parts of the climate system liable to tip (i.e. tipping elements/systems) have only recently started to be undertaken. In many cases no such specific studies exist, although more general research on how SRM affects feedbacks involved in tipping might be available instead. The authors collate direct or indirect studies relating to potential SRM impacts on tipping systems and their ability to mitigate CTPs, covering commonly proposed tipping systems in the cryosphere, ocean, atmosphere, and biosphere. They then assess whether the collated evidence supports SRM being able to prevent tipping in those systems to some extent, as well as whether tipping can be countered once in progress. They find that while many systems have some evidence that tipping drivers can be partly compensated for using SRM, confidence is generally low and research is currently lacking in many areas, for which future research avenues are suggested for improved study scenarios and models.
General comments:
In general this is a timely paper, given increasing attention on the risks posed by climate tipping points and proposals that this necessitates SRM, but only a small (but recently growing) pool of targeted research in this area. Reviewing current literature and identifying gaps to target with further research is therefore a welcome exercise that will hopefully spur on more research in this area, and easily falls within the scope of this journal.
Tipping systems and their hypothesised dynamics are in general accurately portrayed, though in places could variously do with more detail or being more concise (see specific comments for details). There are some places too where the latest papers should be included (e.g. on Antarctica) as the pace is picking up in this field now with some relevant papers published since submission. In several places tighter writing, reducing repetition, and more consistent structuring would help improve flow, which would also make the paper more concise (it’s not overly long at present, but ideally it shouldn’t get longer, and shorter is preferable for readability if possible).
It’s good to have put SRM discourse in context upfront, and the complexities around SRM not cancelling out all climate change as easily as GMST. However, I think there could be more discussion of how applying SRM in ESMs (particularly for slightly older SRM studies) that lack key tipping-relevant processes and feedbacks means we may be missing some key SRM-CTP interactions, such as the effects of SRM-induced precipitation or ecological changes. For example, ESMs are likely biased towards AMOC stability (per IPCC), lack spatial heterogeneity in ecohydrological dynamics in the Amazon (where many ESMs disagree on even sign of precip change too), or lack abrupt thaw or interactions with surface vegetation in permafrost. In some cases this could offset part of the SRM compensation of tipping, and in some cases potentially even worsen tipping risk (e.g. that one study shows worsening of Amazon droughts with SRM vs. emission scenario without SRM). To me this means the conclusion in the abstract that “We find that SRM mostly reduces the risk of hitting tipping points relative to same emission pathway scenarios without SRM” needs additional clarification that this is mostly relative to temperature, with other climatic factors that are not so well understood likely complicating and potentially undermining this risk reduction.
Beyond the often discussed biospheric tipping systems of Amazon rainforest, boreal forests, and warm-water coral reefs, the authors also select dipterocarp forests in Southeast Asia and various ecosystems in South Asia as potential tipping systems. However, for some examples such as dipterocarp forests, eastern Himalayas, or the Western Ghats their potential for self-sustaining regime shifts is not currently clear from the presented evidence. Additionally, it’s not clear why for example the Sundarbans are highlighted but not the mangrove biome more widely, or why other biomes like savannahs/grasslands or non-coral ocean ecosystems aren’t analysed. If this is because these are not covered by existing studies then that’s OK but should be made clear, otherwise it seems a few examples of highly biodiverse ecosystems have been picked and the possibility of tipping in them speculated. I suggest being clearer on reasons for inclusion and their potential tipping dynamics, or adjusting review framing to be explicitly about whether SRM can prevent tipping dynamics plus biodiversity loss in some selected examples. The latter could be a whole other (and interesting) paper though, so I would suggest focusing on where tipping dynamics can be shown to be possible or likely.
Specific comments:
Line no.
12-13: Is preventing CTPs considered a key benefit of SRM that often in literature? My reading of literature is that it has increasingly been suggested in some papers/proposals, but I'm not sure preventing CTPs is so often made central yet.
16: SAI not yet defined.
20-22: Given the large uncertainties include non-temperature factors such as precipitation changes potentially countering or even cancel out tipping-compensation by SRM (particularly for biosphere, where hydrological effects makes SRM very uncertain and e.g. some experiments show worsening of Amazon droughts with SRM), I think this statement could do with some clarification. To me, this review shows that while direct SRM-CTP studies are few or missing for many systems, where studies have been done it appears SRM can reduce tipping risk specifically with respect to regional or global temperature (with high uncertainties), but current model limitations mean SRM's influence on tipping dynamics via other climatic factors could reduce or even nullify risk reductions.
29: AM22 is specifically about climate TPs rather than wider Earth system TPs, which is likely the case here too given regional/global temperature trigger focus.
50: I think self-sustaining or self-perpetuating is more accurate - the changes can be quite steady for a long time rather than accelerating in some cases (e.g. ice sheet collapse).
55: Is the TP here the extinction itself or events leading to it as TP? I'm not sure the former could be seen as a TP, more a general threshold response, as there's no clear self-sustaining change dynamic there (beyond maybe a population bottleneck / functional extinction effect in final stages). Extinction can certainly lead to ecosystem regime shifts though, which can be a TP if the regime shift is self-sustaining in some regard (e.g. such that hysteresis occurs).
73-75: This is a key point (along with e.g. lines 143-146), and links to my hesitation elsewhere about confidence in SRM’s compensation for CTPs being somewhat temperature-centric.
79: A more specific reference here for tipping overshoot would be Ritchie et al (2021) [https://www.nature.com/articles/s41586-021-03263-2].
105-107: Good to have put SRM in wider context upfront.
113-114: Presumably there’s not so much research out there for attempts at regional cooling.
115: Is “resembling” here these references, or is this a sentence fragment?
183: "Greenland Ice Sheet " would be clearer than “Greenland tipping element”.
200-203: I think this would be better rearranged so explanation of rebound comes first.
208-210: I suspect the ice sheet components in the models likely used for these studies were lacking in terms of representing enough ice sheet dynamics to capture tipping, which is a limitation (and taps in to my general comment that SRM studies using ESMs that lack key tipping dynamics won’t be getting the full picture of how SRM effects tipping dynamics).
227-228: Specify by 2100 – much more sea level rise beyond this even on low pathways.
230-231: Recent research from Li et al (2023) [https://www.nature.com/articles/s41586-023-05762-w] would be a good reference inclusion here.
232: “currently driven” – air temperature could become more important in future, depending on circulation changes.
238: East Antarctica is conventionally split in to marine basins and land-based, as they face different dynamics that lead to quite different tipping thresholds and timescales.
240-243: This sentence is a bit convoluted to me, consider rearranging.
251-252: Missing here is the sense of where the tipping point might be, which is generally the point at which retreating grounding line reaches main retrograde slope leading to accelerating discharge (subject to buttressing, pinning, etc.).
276-278: Probably a bit higher warming needed for East Antarctic basins than WAIS based on e.g. Garbe et al. 2020 (more like 3-6C for Wilkes land for example, though I’d agree the risk starts growing from 2C).
287: A key reference on issues around MICI is Edwards et al 2019 [https://doi.org/10.1038/s41586-019-0901-4]
290: Clearer to say “both are preceded by” ice shelf disintegration.
209: 32cm sea level rise by when? It would also be useful to put these values in context by stating WAIS total sea level equivalent (3+m) earlier.
308: Given e.g. Goddard et al. 2023 is discussed below and there are some systems with fewer studies, "relatively few" seems more accurate now. Another recent release to consider in this section too is Sutter et al. (2023) [https://www.nature.com/articles/s41558-023-01738-w].
309-310: Presumably this result is specific to surface air temperature – useful to specify this in context of limited short to medium term impact on ocean warming.
326: Garbe et al. (2020) is not specifically a CDR experiment – might need to rephrase above for clarity.
329: True, but how close it recovers by present day levels depends on experiment type (equilibrium vs “quasi-static”). Also, do you mean current warming of ~1.2C or pre-industrial by present levels?
334-335: Feels like this should be discussed earlier in this subsection given it's such a critical point, before moving on to what changing SAT might be able to do as a second-order driver and then what reversal/CDR studies show.
358-359: Could probably do with a little more detail on and beyond melt-elevation and melt-albedo feedbacks here (these two can signpost back to ice sheets for more info), e.g. role of changing snow patterns and black carbon/dust on albedo, thermokarst interaction, slope instabilities, negative feedbacks from retreat to higher altitudes or debris insulation, etc., as well as that unlike ice sheets these feedbacks happen on a largely local to regional scale.
368-369: Are G3 and G4 defined before this point (for readers not familiar with GeoMIP terminology or context)?
375-377: A Himalayan example is given here, but can anything be said so far about how heterogenous SRM precipitation impacts may affect different glacier regions, e.g. from GeoMIP? If not, then can be explicit in that.
397: SAM not defined before now.
422: Specify what sort of temperature, e.g. surface air, sea surface, regional/global etc.
447-448: Could clarify here that while some positive feedbacks exist that have been suggested could drive tipping dynamics (e.g. https://link.springer.com/article/10.1007/s10584-011-0126-5; https://journals.ametsoc.org/view/journals/clim/34/11/JCLI-D-20-0558.1.xml) the explanation given here is generally more favoured.
476-479: Presumably this is relative to a scenario of continued warming without SRM, rather than versus no further warming in above example? (otherwise this statement could be seen as conflicting with a reduction in March extent above.)
485: Missing "which" after citation
498-500: I think this could do with clarification - compared with continued warming on same emissions pathway but without SRM?
512: Clarify as (presumably) soil temperature.
523: Perhaps useful to clarify this as melt of ice blocks/wedges, and add thermokarst/talik development (which is not necessarily ice melt dependent).
536: Perhaps worth clarifying that in AM22 this was an additional threshold beyond the more general (and higher confidence) widespread localised abrupt thaw above ~1.5C.
549: Missing in this section (reflecting the point made under further research) is a discussion of how current generation models don't capture non-gradual thaw processes, so currently these conclusions primarily relate to gradual, non-tipping thaw only, or how changes in e.g. precipitation interact with abrupt thaw or eco-hydrological processes.
580-581: SRM is also unlikely to reverse abrupt thaw processes like thermokarst/talik development once started, but can limit formation of new taliks.
604: Could specify very long timescales is necessary because of the centennial-millennial timescale of ocean heat uptake and circulation.
616: A useful figure, but why no equivalent for cryosphere or biosphere systems? A generic schematic showing e.g. elevation/albedo feedbacks might help illustrate key feedbacks there. There could also more focus specifically on how SRM might intervene in those feedbacks to make figures more specific to this article (rather than textbook-style background).
622: Deep convection should also be marked in Southern Ocean, as it’s not separate to sinking there.
628: Might need to clarify deep convection/sinking, which is illustrated separately in figure but are really part and parcel (with "sinking" a simplification of deep convection).
640-641: Relevant here is the IPCC's assessment of collapse unlikely before 2100, but also that CMIP models are biased towards stability and lack key aspects (e.g. ice sheet meltwater) that likely lead to under-estimates of weakening and collapse. This is why unrealistic hosing is used in some studies (like Jackson et al. 2023 below) to overcome this tendency. This is mentioned in next paragraph, but think it’s useful context within this paragraph.
652: Jackson et al. 2022 is the preprint, can update to final 2023 paper now.
660: Liu et al 2017 relevant here too [https://www.science.org/doi/10.1126/sciadv.1601666]
662: Another angle beyond palaeo and model evidence not covered here is observations, with the debate over current slowdown (with IPCC AR6 having low confidence in some weakening) and of potential early warning signals suggestive of destabilisation [e.g. https://www.nature.com/articles/s41558-021-01097-4; https://www.nature.com/articles/s41467-022-32704-3; https://www.nature.com/articles/s41467-023-39810-w]
677-678: Ice sheet runoff is quite the important missing factor though, which is missing from most ESM simulations of AMOC weakening/collapse (and therefore SRM studies using those too).
700: One way of exploring possible SRM effects on AMOC would be to explore the role of past/current aerosol emission effects on AMOC, which have been suggested to generally reduce AMOC tipping likelihood (and is being partly reversed over North Atlantic over recent years) and so could effectively be a smaller-scale SRM analogue.
705: There’s a question mark over whether precip changes from SRM would help or hinder though (though admittedly it seems to be a small factor, model limitations permitting).
721-723: This is a key point, and relates to my general comment on uncertainty inherent in using models with limited tipping dynamics representation to project SRM effects on CTPs.
728-729: This is a very good point, and may be applicable to a other tipping points too, as while many have generic warming level thresholds estimated several are likely to have rate-dependent too [e.g. Ritchie et al. 2023 https://esd.copernicus.org/articles/14/669/2023/]. Could be something to highlight need for further study of in discussion.
733-738: This paragraph largely repeats previous one.
748-749: Indeed – a subset of this risk is that if SRM was initiated to protect the AMOC but it turns out tipping has already commenced (as exactly when thresholds are crossed is uncertain and there's a time lag in tipping dynamics) then SRM might have to be phased down or out to avoid over-cooling North Atlantic region, but at potential cost of unprotecting other regions and CTPs. Relates to potential tricky balancing of how to target mitigating multiple CTPs and other climate impacts without worsening some over others.
761: Missing parenthesis after “injection points”.
808: Given similar issues to AMOC could merge with further research there for a clearer structure, in a similar way to how land ice further research is grouped together.
823: Missing fullstop before citation?
831: Relevant here is new paper Li et al (2023) [https://www.nature.com/articles/s41586-023-05762-w] which projects strong weakening with further warming (RCP8.5 in that study, but likely similar magnitude decline for RCP4.5 on longer timescales)
834: Sentence would be clearer if what the mechanism is for is specified (overturning weakening? collapse?)
839: Contrast Li et al, who suggest wind has marginal effect on projected weakening (though possibly model dependent).
862: Only marine stratocumulus clouds are tackled in this section, and not e.g. monsoons which are sometimes considered atmospheric TPs, though possibly more due to aerosols than climate change proper. Given the relation of monsoon tipping to aerosols, and recent work on interhemispheric AOD difference as a key safe boundary (Rockström et al. 2023), their relation to SRM would be an interesting avenue to explore if resources permit.
894: Warming values repeated from past paragraph (not bad, but unnecessary to make hysteresis point).
920: What does this mean - the feedback is fully counteracted, or literally turned off?
948: This general intro to ecosystem TPs & SRM is useful in general, but is repetitive in parts and could be tightened.
953-954: In my experience extinction is not normally thought of as a TP / regime shift in ecology (although extinction or replacement of a keystone species could help trigger a wider regime shift).
965-967: Clarify – the impact/event itself was not a TP (just an abrupt exogenous forcing with an equally abrupt response), but it likely drove many localised/regionalised ecosystem regime shifts (some but not all of which might have featured self-sustaining tipping dynamics).
1010-1013: This sentence can be simplified for clarity.
1024: To me this is conceptually confused – "forests" and "biodiversity" are very different categories of things (one a collection of objects forming a system, the other a property of such a system).
1025-1030: These sentences are a bit confusing and could do with tightening.
1051-1052: I'd like to see more evidence presented on how these forests are susceptible to tipping (via regeneration failure) that justifies pulling it out as a specific unit to analyse. In analyses of tropical rainforest-rainfall feedbacks, rainforests in maritime Southeast Asia are normally found to be more robust due to plentiful ocean rainfall sources, so it can't be that mechanism, and syncrhonised seeding occurs in many forest biomes, particularly tropical but including to some extent temperate too. Numata et al. is a key citation here, but it's not apparent that regime shifts per se are projected in that study. An alternative would be too merge in with discussion of other tropical forests.
1073-1075: I couldn't find this in the given reference.
1078: Clarify where.
1082-1083: Assumes there is a TP in this system – if this is uncertain though could reframe this section as about protecting a particular biome that may or may not feature tipping dynamics (but then a question is why this ecosystem and not various others).
1086: Missing a "Further Research" subheading here?
1097: This surprises me, as Amazon is normally considered to be one of the most biodiverse biomes (when considered as a biome unit). The cited analysis doesn't highlight Amazon because the "hotspots" are a product both of endemism and threat level, with Amazonia seemingly left out because of lower threat rather than lower endemism (i.e. hotspots in Myers et al 2000 don't mean that's where most biodiversity is, but where areas of high biodiversity are most threatened – it also excludes invertebrates, and any newly identified species in past 23 years). Also, this citation is not in the reference list (I’m not sure it adds too much to discussion anyway).
1101-1102: 2-6C is for just climate change without anthropogenic co-drivers like deforestation – including those could lower it further.
1104: More recent studies highlight that degraded forest is a likely alternative state (possibly more common than savannah).
1123-1124: Some discussion of where dieback is more likely – in the drier south and east, as per bistable areas in Staal et al. 2020 – would be useful too for understanding the feedbacks and areas most at risk.
1143: Could make a link here to AMOC collapse too, which is projected to shift ITCZ southwards and so cause drying in Northern Amazonia (likely similar to in Younger Dryas). Also some link to monsoon shifts and global aerosol patterns.
1150-1151: Even here GCMs often struggle to agree on even the sign of regional precip change, which is one of the reasons ESMs tend to disagree on likelihood and scale of Amazon dieback.
1165: "the utility... is of limited utlity" – rephrase to e.g. "Thus, existing studies...".
1194: I think representing land use change, and spatial heterogeneity in plant traits adaptivity across the Amazon is also key for better model representation.
1205-1206: This probably needs a citation [e.g. https://journals.plos.org/plosone/article?id=10.1371/journal.pone.0025026]
1249: Critical too in some regions like the Caribbean is disease and invasive species spread (often facilitated by warming and globalised trade).
1268: Some numbers here would be useful to demonstrate this if possible.
1278: Recurrence in between hurricanes?
1291: While the basis for SRM effects on coral bleaching is better understood than for some tipping systems, given the fair few uncertainties explored above relating e.g. to extreme events, complex co-drivers, or variation in SRM impacts between different coral regions makes it seem not quite so strong to me. Not a huge amount of dedicated studies are cited above either. Personally I'd like to see studies with up-to-date ESMs exploring where and how much SRM is necessary to e.g. keep below specific Degree Heating Week or recurrence thresholds, how this affects storms, and the spatial heterogeneity of these.
1303-1304: Are these are all vulnerable to tipping? Not all ecosystems have alternative stable states to tip to – they can just become gradually degraded – so it's important to present or cite evidence for self-sustaining regime shifts. The Sundarbans are the clearest example here, with coastal erosion and salinity effects likely to drive tipping dynamic across mangrove biome (GTPR, 2023), but given they're at risk of regime shifts across the tropics why is this the only locality discussed?
1311: Here as in above in text?
1315-1317: Maybe better discussed in SRM effects subsection below?
1329: In other subsections the drivers and mechanisms of a specific tipping system is analysed, while here it's for multiple different systems within the same region, which is somewhat inconsistent in paper structuring. Additionally there's some overlap with other sections, e.g. for glaciers.
1334-1336: Specify what system this tipping point is in – downstream ecosystems? Peri-glacial environment around glaciers?
1343-1353: This paragraph reads rather fragmented to me. Additionally, I don’t think the TPs on line 1345-1346 have been explicitly described yet.
1374-1378: Possibly worth noting that this hypothesis (or at least magnitude of effect) is debated.
1388-1390: Worth noting that biogeophysical effects can have opposite regional climate impacts to the global climate effect (which could potentially interact with SRM) – e.g. boreal dieback to steppe increasing albedo and regional cooling, versus carbon released by dieback increasing global warming.
1394: Given complex eco-hydrological mechanisms in boreal forest dynamics this subsection could do with mention of how (likely uncertain) impacts of SRM on precip might complicate temperature-based considerations.
1403: It seems that there's no specific SRM studies re. boreal forests – if so, should highlight that as a knowledge gap.
1417-1422: Minor point, but not sure these sentences are necessary as they mostly repeat the Introduction.
1426-1428: Useful point to mention –potentially relevant to highlight in abstract.
Table 1: Is b) in the 4th column linked to a specific timescale? GrIS, AIS, and permafrost here are Not Reversible by SAI once tipping is complete, which is true on centennial timescales but not necessarily very long timescales, while AMOC collapse is Yes with hysteresis, when on long enough timescales that could apply to the former examples too. I’m also not entirely sure if AMOC/SPG collapse can likely be reversed once tipping has begun (depends on feedback dynamics, which we don’t have a great hold on right now). Also, is “Uncertain” in last row 4th column for both a and b?
1441-1442: These partial compensations are mostly with regards to temperature though – I would hazard that greater uncertainty / lower confidence would hold on this partial compensation once accounting for impacts on other climate factors such as precipitation. At the moment the Discussion and Table 1 doesn't highlight this.
1444-1447: These points are repeated two paragraphs down.
1445: I don’t think peak-shaving hasn’t been mentioned prior to this point, so could do with an explanatory citation on first instance.
1449-1453: This overlaps discussion of emergency use and hysteresis in paragraph below – could probably tighten discussion to be more concise.
1475: Could specify biosphere/ecosystem TPs here (e.g. biomass plantation based CDR making forest dieback or savannah degradation more likely).
1477: A key issue and source of uncertainty for me is that we know most about how SRM might effect CTPs via temperature, but much less about other climatic factors like precipitation, which in many cases could reduce or even overwhelm the compensation delivered by reducing warming (e.g. less precip delivered to ice sheets, or cooling accompanied by droughts in tropical forests). This relates to the research approaches given below – improving modelling of Earth system tipping dynamics is critical to be able to give a fuller assessment of how different SRM schemes might affect tipping processes.
1478-1481: This is not very clear to me at present – is this saying that these Qs wouldn't be relevant to people who think SRM research shouldn't be pursued at all because of moral hazard issues, or more the issues raised in the penultimate paragraph?
Dr. David A. McKay
Citation: https://doi.org/10.5194/egusphere-2023-1753-RC2 - AC2: 'Reply on RC2', Gideon Futerman, 26 Jan 2024
Peer review completion
Journal article(s) based on this preprint
Viewed
| HTML | XML | Total | BibTeX | EndNote | |
|---|---|---|---|---|---|
| 1,629 | 518 | 62 | 2,209 | 84 | 81 |
- HTML: 1,629
- PDF: 518
- XML: 62
- Total: 2,209
- BibTeX: 84
- EndNote: 81
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1
Cited
4 citations as recorded by crossref.
- Kicking the can down the road: understanding the effects of delaying the deployment of stratospheric aerosol injection E. Brody et al. 10.1088/2752-5295/ad53f3
- Accounting for transience in the baseline climate state changes the surface climate response attributed to stratospheric aerosol injection A. Duffey & P. Irvine 10.1088/2752-5295/ad9f91
- Low‐Altitude High‐Latitude Stratospheric Aerosol Injection Is Feasible With Existing Aircraft A. Duffey et al. 10.1029/2024EF005567
- ‘Tipping points’ confuse and can distract from urgent climate action R. Kopp et al. 10.1038/s41558-024-02196-8
Discussed
Gideon Futerman
Mira Adhikari
Alistair Duffey
Yuanchao Fan
Peter Irvine
Jessica Gurevitch
Claudia Wieners
The requested preprint has a corresponding peer-reviewed final revised paper. You are encouraged to refer to the final revised version.
- Preprint
(1217 KB) - Metadata XML