Background-Greenland collapse

323 views
Skip to first unread message

Andrew Lockley

unread,
Aug 15, 2020, 3:40:37 AM8/15/20
to geoengineering
If this study is correct https://www.nature.com/articles/s43247-020-0001-2
Then it appears to back up a point that I have been making for a long time: geoengineering is required, if we are to keep our coastal cities. I do not see economic or political feasibility for large scale CDR to tackle historic emissions, and thus the task must fall to SRM. 

Nobody has managed to rause an objection to this argument to date. I'd be grateful if those who might disagree were to raise counter arguments now. 

If the situation is as I understand it, prevarication has no clear benefits, and we should thus move quickly to readiness for deployment. 

Andrew 


Douglas MacMartin

unread,
Aug 15, 2020, 7:03:43 AM8/15/20
to andrew....@gmail.com, geoengineering

What is not correct in the media report is this sentence: “This process, however, would take decades.”  Well, I guess arguably that’s true, it’s just it would take a LOT of decades.  Melt rate is currently of order 1-2mm/yr equivalent SLR, so to get the 6m from melting all of Greenland would take a few thousand years.  Obviously it can speed up a lot, but “hey, it’s losing mass” does not remotely imply “therefore we only have a few decades before we lose our coastal cities”.  So no, you can’t use this study to claim that geoengineering is required to keep our coastal cities.  The problem with relying on mitigation+CDR is time-scale, but this study doesn’t prove that our response time-scale needs to be faster than what CDR can (at least hypothetically) provide. 

d

--
You received this message because you are subscribed to the Google Groups "geoengineering" group.
To unsubscribe from this group and stop receiving emails from it, send an email to geoengineerin...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/geoengineering/CAJ3C-07%2BJc0dmp26W2_H08Rsrs1V_sjEs1pkG6xuZZ75OcTa%2Bw%40mail.gmail.com.

Andrew Lockley

unread,
Aug 15, 2020, 7:20:07 AM8/15/20
to Douglas MacMartin, geoengineering
The more advanced the process, the more momentum it has, and the harder it is to stop or reverse. It's unclear whether we could arrest mass loss, without driving the earth into a glacial period (colloquially, an ice age). I don't think there has been any serious modelling work done on ice loss reversal, or even if the models are capable of doing this with any useful accuracy.

Stephen Salter

unread,
Aug 15, 2020, 7:59:48 AM8/15/20
to geoengi...@googlegroups.com

Hi All

The attached note has estimates for ice loss numbers for Arctic ocean ice, Greenland, Antarctic and for deep ocean temperature rise. Greenland is not the biggest.  Some of you may have seen this before and values for input assumptions have been made.  Suggestions for changes to the input assumptions would be welcome.

It would be exciting to see if the buoyancy of melt water below 3000 metres of ice could give enough lubrication for a thick chunk to slide into the sea and make a spectacular tsunami but I expect that there would be objections from the environmental community against practical tests.

It would be useful to compare estimates for low probability of dangerous  events with our choice to fly on an aircraft with the same crash probability.

Stephen

Emeritus Professor of Engineering Design. School of Engineering, University of Edinburgh, Mayfield Road, Edinburgh EH9 3DW, Scotland S.Sa...@ed.ac.uk, Tel +44 (0)131 662 1180 WWW.homepages.ed.ac.uk/shs, YouTube Jamie Taylor Power for Change
Sea level rise 4.pdf

Douglas MacMartin

unread,
Aug 15, 2020, 8:05:04 AM8/15/20
to Andrew Lockley, geoengineering

True that, whether one uses lots of CDR and/or SRM, it is possible that neither of those options would arrest mass loss without cooling beyond preindustrial, and that over the ensuing millennium we would eventually lose our coastal cities no matter what we do, but that we would delay that if we start solar geo sooner.  But (a) we don’t know and I think your second statement is correct, that we don’t have the modeling capability for that, (b) I’m more concerned about what we do in the next 50 years, myself, so questions of whether the cities are ultimately un-savable on timescales of centuries I don’t think matters for near-term decisions (c) personally I’m a lot more worried about Antarctic than Greenland, (d) nothing you said justifies your original claim (insofar as solar geo doesn’t guarantee keeping coastal cities over long timescales either).  I agree that starting solar geo today would result in less SLR by 2100 than not starting it, if emissions stayed the same, but the only thing we know in that sentence is the word “less”.

 

Your original claim was also based on the premise of throwing out all other options.  Yes, generally in any situation in life, if you refuse to consider all of the options but one, then yes, you have only one option.  That’s true, but only trivially.  Yes, if you rule out large-scale CDR to tackle historical emissions, then it is unequivocally true that SRM would be the only way to reduce temperatures beyond the current level.  That’s not exactly a profound observation…

Jasmin S. A. Link

unread,
Aug 15, 2020, 8:40:51 AM8/15/20
to geoengi...@googlegroups.com

"The more advanced the process, the more momentum it has, and the harder it is to stop or reverse." The same can count for social dynamics. Self-reinforcing processes with the tendency towards a lock-in are path-dependent processes. That is, why it is not easy to just change carbon intensive behavior towards low carbon emission behavior. Many social, economical, and technical processes have the production of carbon emissions as side effects which accumulate during the intensifying self-reinforcing processes (cp. Figure 4 p. 63 in https://ediss.sub.uni-hamburg.de/volltexte/2020/10431/pdf/Dissertation.pdf ).

But what about the side effects of SRM that would accumulate over time while deployment in an intensity of "driving the earth into a glacial period"?

It is necessary to decouple self-reinforcing processes from producing carbon emissions as a side effect. And try to potentially include the reduction of carbon emissions as side effects of new self-reinforcing processes that show the dynamic of "the more advanced the process, the more momentum it has, and the harder it is to stop or reverse".

If you would plan the potential deployment of SRM (especially on a large-scale) you would in the same way have to consider the potential side effects beforehand to assure to decouple a deployment of SRM from these potential side effects. Otherwise you would risk the accumulation in acidification of the oceans, self-reinforcing the reduction in biodiversity, loss in coral reefs, accumulating in social tensions (just to name a few of the interconnected potential side effects, cp. Figure 3.1.,p.71 in the transdisciplinary network of potential side effects in: https://www.iass-potsdam.de/sites/default/files/2018-06/EuTRACE_report_digital_second_edition.pdf ).

Furthermore, with SRM you address the global mean temperature. If this would work at all, this would not imply the globally even reduction of temperature, regional effects may differ a lot, potentially increasing also the climate related risks for coastal cities, which you have mentioned as your main concern to safe.

Best,

Jasmin

To view this discussion on the web visit https://groups.google.com/d/msgid/geoengineering/CAJ3C-04pLr9SO5mJQXfUQUXxwUorOSNT1EgdO2AWLVv9Ba4h%2Bg%40mail.gmail.com.
-- 

Dr. Jasmin S. A. Link
Research Group Climate Change and Security
Center for Earth System Research and Sustainability
University of Hamburg
eMail: jasmi...@uni-hamburg.de
www.clisec-hamburg.de
www.cen.uni-hamburg.de

The CEN combines expertise from the natural and social sciences
and unites nine institutes and research centers at Universität Hamburg.

Michael MacCracken

unread,
Aug 15, 2020, 12:54:14 PM8/15/20
to dgm...@cornell.edu, andrew....@gmail.com, geoengineering

Hi Doug et al.--I'm a bit late to this particular conversation, but I am astonished by the suggestion that Greenland can only cause such a small potential rate of rise in sea level. There was just a kerfuffle with the IPCC authors on their draft projections of rates (see attached letter). While surface melt rate may be relatively slow as often calculated, it is not the main loss of mass process--ice stream flow is very likely the major loss rate once it gets going and the calculations that are done in most models do not include this term, nor do they include the effects of ice shelf thinning that is going on. From the peak of the last interglacial to 8 ka, sea level rose at an average rate of a meter per century while global average temperature rose at an average rate of a degree C per 2000 years, and the CO2 concentration was less than 300 ppm. The documentary "Chasing Ice" shows how fast ice can disappear, and not just in the ice stream calving that is the most amazing aspect of that film. And paleo evidence also makes very clear that ice sheets go away much faster than they build up.

And the question is not so much when the cities will be under water as when it will become inevitable that they will be under water--given that consideration and the paleo sensitivity being something like 15-20 meters per degree C warming (and this is not just me saying this, but see Eric Rignot talk to the NAS last year--see https://vimeo.com/332486918 ).

Based on this sensitivity, we're already past the point where it would be good to have climate intervention underway if we want to avoid significant and early risk to our cities with a very high likelihood (and this is the criterion that is often used in building infrastructure--avoiding 1 in 100 year events or even rarer ones--consider the Dutch for their levees--1 in 10,000 year storms).

Mike MacCracken

Ronal Larson

unread,
Aug 15, 2020, 1:48:06 PM8/15/20
to jasmi...@uni-hamburg.de, via geoengineering
Dr.  Link and list:

Thanks for alerting us to your recent thesis.  I have only skimmed so far, but intend to read it carefully - as a way of promoting CDR (and one of its 7-8 main approaches).  

 For others, here is the final paragraph of Dr.  Link's first (main) chapter, with my own emphases added.
This cumulative dissertation uses mathematical sociology in a new way to develop an innovative path dependence perspective of agency, which is combined with conflict and consensus theories and applied to the climate-conflict nexus. The result is a mathematically sociologically rigorous analysis of historical, current and potential future behavior, which is induced by path dependence and applied in the context of climate change. While statistical analyses may produce new extrapolations every millisecond in the era of massive big data, this mathematical sociological analysis of induced social behavior allows to get an idea on what dynamics prevail in a new situation, a situation of change. The new focus looks at the underlying social dynamics rather than merely extrapolating data points. The dissertation shows that even in areas with large complexity and large uncertainty an extensive logical thinking can reveal interdependencies and an understanding of dynamics that can be helpful in connecting disciplines, shaping future research and potentially even societal or political action

The geoengineering list is well suited for discussion of your thesis.  But the now larger breakaway group called CDR also needs the help you might provide through emphasizing path dependency.  Can I ask that you also submit there?. Or  do you feel your thesis is better discussed only on this list?

You had some excellent tables and figures - that I think will help CDR proponents better sort out best future “paths”.  One below.

Again, thanks for letting us know of your thesis completion.  Also few inserts below.

Ron



On Aug 15, 2020, at 6:32 AM, Jasmin S. A. Link <jasmi...@uni-hamburg.de> wrote:

"The more advanced the process, the more momentum it has, and the harder it is to stop or reverse." The same can count for social dynamics. Self-reinforcing processes with the tendency towards a lock-in are path-dependent processes. That is, why it is not easy to just change carbon intensive behavior towards low carbon emission behavior. Many social, economical, and technical processes have the production of carbon emissions as side effects which accumulate during the intensifying self-reinforcing processes (cp. Figure 4 p. 63 in https://ediss.sub.uni-hamburg.de/volltexte/2020/10431/pdf/Dissertation.pdf ).


[RWL:   And there are more, especially related to (the here illegible) lower right corner.


But what about the side effects of SRM that would accumulate over time while deployment in an intensity of "driving the earth into a glacial period"?

[RWL:  See Fig. 3.1 cited below for quite a few “side effects”.

It is necessary to decouple self-reinforcing processes from producing carbon emissions as a side effect. And try to potentially include the reduction of carbon emissions as side effects of new self-reinforcing processes that show the dynamic of "the more advanced the process, the more momentum it has, and the harder it is to stop or reverse".

[RWL:   Is CDR immune from this concern?

If you would plan the potential deployment of SRM (especially on a large-scale) you would in the same way have to consider the potential side effects beforehand to assure to decouple a deployment of SRM from these potential side effects. Otherwise you would risk the accumulation in acidification of the oceans, self-reinforcing the reduction in biodiversity, loss in coral reefs, accumulating in social tensions (just to name a few of the interconnected potential side effects, cp. Figure 3.1.,p.71 in the transdisciplinary network of potential side effects in: https://www.iass-potsdam.de/sites/default/files/2018-06/EuTRACE_report_digital_second_edition.pdf ).

[RWL:   Figure 3.1 is coauthored by yourself on SRM and looks complete.  Have one like it for CDR?  The EuTRACE volume seems to have been written in 2015;  none of the 8 biochar cites are after 2012.

Furthermore, with SRM you address the global mean temperature. If this would work at all, this would not imply the globally even reduction of temperature, regional effects may differ a lot, potentially increasing also the climate related risks for coastal cities, which you have mentioned as your main concern to safe.

[RWL:   You have proven that Social Science has a good bit to offer on SRM;  hope you can do the same for CDR - especially relative to "paths”.

Ron

Douglas MacMartin

unread,
Aug 15, 2020, 5:35:19 PM8/15/20
to Michael MacCracken, andrew....@gmail.com, geoengineering

Hi Mike,

 

-          I don’t know (and I’m not sure anyone really does) how much Greenland ice sheet mass loss can accelerate, but agree that it won’t stay at the current rate.

-          Mainly my point was that the media link Andrew sent was silly by implying that the 6m could happen in decades (even if it didn’t technically say that), and the paper doesn’t make any claims about how much the loss rate will accelerate… and to support the claim that geoengineering is the *ONLY* way to avoid catastrophic sea level rise requires evidence that we need to intervene now rather than over the rest of the century.  I do agree that there’s good reason to suspect that Andrew’s claim may be true, but it is certainly not supported by the paper he was referring to, and I don’t think we can prove that the claim is true.

-          Paleo evidence makes it clear that staying at even the current CO2 levels for millennia would be catastrophic.  It doesn’t do a great job of constraining how rapidly we need to change, e.g. if CDR over the rest of the century would also be an adequate alternative to SRM.  (And paleo evidence also shows that it is possible to get ~5m of SLR in a single century, but that’s coming out of the last glacial when there was a lot more available to be melted, and that doesn’t say that it is possible to get anything close to that this century.)

-          Arguably this is simply quibbling over whether we can prove we’re past the point where even aggressive CDR would work, or whether there’s simply a risk that we’re past that point, in which case arguing over whether that is 20% or 50% or 10% might not matter for policy.  I do agree that we are gambling with the climate, and with odds that no-one would accept in any other circumstance. 

-          Personally, given how little research has been done, I don’t think there’s strong justification for saying that the long-term climate consequences of waiting another 10-15 years for research (and to develop governance capacity) will be so bad that we should go ahead and deploy something now without doing the research (though if we did deploy something now, I’d worry more about the societal response than the physical issues).  But, just like the US didn’t use the first few months of this year to prepare for covid when it knew it was coming, it would be truly awful to not do the research now, leaving us in the same boat yet another decade later.

doug

 

From: Michael MacCracken <mmac...@comcast.net>
Sent: Saturday, August 15, 2020 12:54 PM
To: Douglas MacMartin <dgm...@cornell.edu>; andrew....@gmail.com; geoengineering <geoengi...@googlegroups.com>
Subject: Re: [geo] Background-Greenland collapse

 

Hi Doug et al.--I'm a bit late to this particular conversation, but I am astonished by the suggestion that Greenland can only cause such a small potential rate of rise in sea level. There was just a kerfuffle with the IPCC authors on their draft projections of rates (see attached letter). While surface melt rate may be relatively slow as often calculated, it is not the main loss of mass process--ice stream flow is very likely the major loss rate once it gets going and the calculations that are done in most models do not include this term, nor do they include the effects of ice shelf thinning that is going on. From the peak of the last interglacial to 8 ka, sea level rose at an average rate of a meter per century while global average temperature rose at an average rate of a degree C per 2000 years, and the CO2 concentration was less than 300 ppm. The documentary "Chasing Ice" shows how fast ice can disappear, and not just in the ice stream calving that is the most amazing aspect of that film. And paleo evidence also makes very clear that ice sheets go away much faster than they build up.

Douglas MacMartin

unread,
Aug 15, 2020, 5:40:10 PM8/15/20
to jasmi...@uni-hamburg.de, geoengi...@googlegroups.com

Thanks Jasmin,

 

Agree that all the side effects of SRM need to be considered, and need to evaluate options holistically.

 

That said, SRM does not cause ocean acidification; CO2 causes ocean acidification.  So one should never list ocean acidification as a side effect.  (It is true that SRM doesn’t solve ocean acidification, but it also doesn’t solve car accidents… and no-one lists that as a reason not to consider SRM.)  Implicit in listing acidification in any discussion of SRM is an assumption that somehow we’re required to choose between reducing CO2 or using SRM, in much the same way that we have to choose whether to drive safely or wear a seat belt, but we’re not allowed to do both. 

 

doug

 

From: geoengi...@googlegroups.com <geoengi...@googlegroups.com> On Behalf Of Jasmin S. A. Link
Sent: Saturday, August 15, 2020 8:32 AM
To: geoengi...@googlegroups.com
Subject: Re: [geo] Background-Greenland collapse

 

"The more advanced the process, the more momentum it has, and the harder it is to stop or reverse." The same can count for social dynamics. Self-reinforcing processes with the tendency towards a lock-in are path-dependent processes. That is, why it is not easy to just change carbon intensive behavior towards low carbon emission behavior. Many social, economical, and technical processes have the production of carbon emissions as side effects which accumulate during the intensifying self-reinforcing processes (cp. Figure 4 p. 63 in https://ediss.sub.uni-hamburg.de/volltexte/2020/10431/pdf/Dissertation.pdf ).

Russell Seitz

unread,
Aug 15, 2020, 10:01:07 PM8/15/20
to geoengineering
Mike, Douglas thinks:
 "it is possible that neither of those options would arrest mass loss without cooling beyond preindustrial, and that over the ensuing millennium we would eventually lose our coastal cities no matter what we do, "

But it is curious how may former seaports have ended up stranded inland by sediment transport-  Troy and Pisa fpr instance.We need to consider  how many of "our" major coastal cities existed in recognizable form a hundred, let alone a thousand years ago? Modern demographics have driven the rise of skyscrapers out of former  swamps from Miami to Singapore, and experience architectural rollover on a time scale reckoned in generations, not centuries. 

Coastal plains may be doomed, but  San Francisco  and Hong Kong are a ar cry from Houston and Lagos. Has the IPCC  mapped and counted  the fraction of current  coastal city  populations that can gain a meter of elevation by walking a kilometer inland from the neighborhoods in which they presently reside?  There's more at issue than urban islands like Venice. Cities have a life of their own , and wherever waterfronts  can migrate inland, waterfront properties will continue to be built behind them

Kevin Lister

unread,
Aug 15, 2020, 10:08:33 PM8/15/20
to Douglas MacMartin, Michael MacCracken, Andrew Lockley, geoengineering
Doug,

If it where only the Greenland ice sheet that was suffering catastrophic and irreversible change,  we might be able to relax and wait 10 to 15 years,  but every part of the world and all it's ecosystems are facing similar destructive change,  from the coral  reefs  to the subsea permafrost.  This is happening as the global population is heading towards 10 billion with almost everyone already at each throats.

If we delay SRM then the scale of the necessary intervention will grow with the delay,  most likely exponentially,  and the risks of unintended consequences or out right failure will grow accordingly. At the same time,  society could be at a point of break down making any coordinated plan  virtually impossible  and debates about agreeing on global governance irrelevant. 

Maybe some elements of the paper that was posted can be quibbled over,  but the need to make an urgent start on SRM can't be. 

Kevin 

Jasmin S. A. Link

unread,
Aug 15, 2020, 10:09:10 PM8/15/20
to dgm...@cornell.edu, geoengi...@googlegroups.com

Thank you Doug,

> If you would plan the potential deployment of SRM (especially on a large-scale) you would in the same way have to consider the potential side effects beforehand to assure to decouple a deployment of SRM from these potential side effects. Otherwise you would risk the accumulation in acidification of the oceans, self-reinforcing the reduction in biodiversity, loss in coral reefs, accumulating in social tensions (just to name a few of the interconnected potential side effects, cp. Figure 3.1.,p.71 in the transdisciplinary network of potential side effects in: https://www.iass-potsdam.de/sites/default/files/2018-06/EuTRACE_report_digital_second_edition.pdf ). <

Regarding ocean acidification and acid rain: Yes, putting up mirrors would not directly cause ocean acidification but SAI might cause acid rain (depending on the aerosol used). And no, it still is related via an indirect effect in terms of social behavior: If people comprehend SRM in a way as that there is some sort of technical compensation happening for their carbon emissions, they are likely to rather not reduce their own carbon emissions, but instead increase their own carbon emissions. This is marked in Figure 3.1. as "less necessity for direct emission reduction?" connected to a "rise of CO2 emissions", which causes multiple feedbacks such as the necessity to further increase the SAI deployment. I think, there have been more recent simulation studies at MIT on this kind of behavior that support the relevance of this potential indirect effect.

In my argument on the potential accumulation in acidification of the oceans, I had this indirect effect of the SRM deployment on the rather increased carbon emission behavior in mind: Officially applying a method to reduce global warming which might be understood as "fixing the problem with engineering" might rather reduce than increase individual mitigation efforts. (Of course, we know that the problem is not really fixed, but try to explain that to usual consumers who would know at some point that their government would regularly spend large sums of money on SRM and who might feel more comfortable to stick and increase their former behavior than to really change it, due to path dependence).

Best,

Jasmin

Jasmin S. A. Link

unread,
Aug 15, 2020, 10:09:51 PM8/15/20
to Ronal Larson, via geoengineering

Dear Ron,

Thank you very much, for your interest an for emphasizing the relevance of my thesis in the context of SRM and CDR.

Please, feel free to add me to the CDR list as well.

The detailed potential side effects networks I have published only for SAI (Figure 3.1.) and BECCS (Figure 3.2. in the same EuTRACE final report) so far. I have made one for afforestion, but that has not been as requested and intriguing for discussion yet.

I think, some of the potential indirect side effects of SRM - such as a potential reduction in individual mitigation efforts - might also apply to a large-scale CDR. Depending on the method of CDR and its public presence (like artificial trees), this might also be true for even relatively small applications of CDR. Thus, it might be important to estimate in advance, if the contruction of one artifical tree rather decreases carbon emissions, by capturing them, or in total rather increases carbon emissions, by triggering passengers to feel free to emit more carbon.

Best regards,

Jasmin

Douglas MacMartin

unread,
Aug 16, 2020, 6:59:45 AM8/16/20
to Jasmin S. A. Link, geoengi...@googlegroups.com

Thanks – I agree completely that moral hazard is a serious risk, perhaps the biggest risk.  (But I also think it is important to be more explicit about one’s assumptions.)

 

There’ve been a few studies trying to look at moral hazard, with inconclusive results even on the sign of the effect – very hard to make predictions about, but, of course, very clear that it is a serious possibility.

 

(And re acid rain, that’s not significant in terms of ocean acidification.)

 

doug

Andrew Lockley

unread,
Aug 16, 2020, 7:50:54 AM8/16/20
to Douglas MacMartin, Jasmin S. A. Link, geoengineering
Whenever this issue comes up, I feel obliged to draw attention to the fact that moral(e) hazard is actually two separate things: recklessness (morale) and malfeasance (moral).


Jasmin S. A. Link

unread,
Aug 16, 2020, 11:15:30 AM8/16/20
to geoengi...@googlegroups.com

Yes. I am not sure, whether moral hazard or even morale hazard, as Andrew has pointed out, exactly name what I try to describe. Moral hazard or morale hazard seem to me both have at least a deliberate decision, at least concious, perhaps even rational or optimizing in some sense, of action making in common. Regarding emission behavior, people usually do not deliberately choose to emit CO2 directly or indirectly. Usally, the emissions are just a side-effect of their consumption, their way of living. Thus, a side-effect of their routines. For many people it is even difficult to reflect on how much carbon they emit via which choice of consumption. Consequently, an individual attempt to reduce carbon emissions is a great effort and usually a large step out of the comfort zone for a single person. Calling it moral hazard, if they "just relax" and fall back a bit in their routines, or enjoy intensifying them, sounds a bit hard. This kind of behavioral decision-making processes may be far from what you might expect as rational decision-making.

In my earlier referred thesis (Link 2018) I coin the term "path-dependent behavior", which is the result of the influence of existing path-dependent processes and is rather a following behavior, following routines/standards/institutions, the masses, neighbors etc. It is an efficient way of decision-making, resulting from a short-cut in the brain (applying the least-effort-principle from social psychology). But the resulting action may be a different one than the result of a utility function would be. Like following the advice of your computer expert (Herbert Simon has also described the phenomenon of personal experts) does not mean that you have optimized the variety of choices yourself. But if the choice you have made that way is producing more carbon emissions along the way than a different one would have, that you might have not even thought about when only relying on your personal expert: Can you be accused for moral(e) hazard?

Best regards,

Jasmin

Andrew Lockley

unread,
Aug 16, 2020, 11:24:17 AM8/16/20
to Jasmin S. A. Link, geoengineering
To clarify, morale hazard is lazy recklessness (I'll put off swapping my SUV for a tesla, because geoengineering may sort it out) moral hazard is manipulative (I'll dangle the carrot of geoengineering, to get approval for keystone). Personal profit is inherent in moral hazard, knowingly harming others. 

Sorry if this is wandering off topic. 

Michael MacCracken

unread,
Aug 16, 2020, 1:32:22 PM8/16/20
to geoengineering

Hi Doug--While the thread has headed off onto issues of moral hazard, etc., getting back to the situation with Greenland, here is the latest paper on this: https://scitechdaily.com/climate-scientists-sound-the-alarm-warming-greenland-ice-sheet-passes-point-of-no-return/   And I think Eric Rignot would suggest some of the ice streams for Antarctic ice sheet are looking less than stable.

I would also note that the last deglaciation melted back about two-thirds of the maximum ice--we have one-third left and it is now starting to go. So, saying the past amplified rate was due to having lots more ice around needs to be more carefully stated--there was more at mid-latitudes, but now mi-latitude weather is being pushed to higher latitudes (consider Siberia this summer), so I see no reason for reassurance about the stability of the ice.

There really is no time to wait to at least getting started in dealing with the issue. It is true that the situation may already be too far along to stop significant loss, but I don't know of any real benefits from sea level rise for any nation. And, as I've argued for some time, I think one might start with some modest actions focused on high latitudes (yes, heat does continue to come in from lower latitudes and that has to be accounted for, but I'd suggest there are things that could be quite possibly be done--see  doi:10.1002/2016EF000450

I'm all for research, and actually getting started on moderating the Arctic amplification, etc. that is threatening the ice sheets (and note my suggestions on initial efforts focus on tropospheric approaches, mainly during the peak solar radiation period and not an all-year approach, so not initially a long-lasting influence and less likely to have significant impacts on hemispheric and global atmospheric circulations). I also think that it would best be viewed as part of an iterative learning activity (scientifically, technologically, and with regard to governance, etc.) that would help build insight into the potential for global scale intervention (computer studies are very useful, but I think it time, or very nearly, time to be out there trying to alleviate a very serious and growing risk). And in that such regional efforts are not really dealing with the full issue, I think the moral hazard argument is not quite as significant a concern--indeed, getting a better sense of what intervention would need to look like might well lead to recognition of the need for more intense action.

Right now, the situation is reminding more and more of the frogs in the pot being heated just hesitating and hesitating thinking that the information about jumping out is just not yet adequate.

Mike

Douglas MacMartin

unread,
Aug 16, 2020, 2:47:45 PM8/16/20
to andrew....@gmail.com, Jasmin S. A. Link, geoengineering

Certainly the term moral hazard as used in other literature doesn’t imply any conscious thought process as a requirement.  (Examples might be better termed risk compensation; someone is more likely to drive close to a cyclist who is wearing a helmet, for example.)

 

Perhaps better to simply use the term mitigation deterrence and not make any implied judgment.

Ronal Larson

unread,
Aug 16, 2020, 2:56:00 PM8/16/20
to via geoengineering, jasmi...@uni-hamburg.de
Geo and cc Jasmin

I have sent this message (and added a bit) over to the CDR list.

Ron



On Aug 15, 2020, at 5:54 PM, Jasmin S. A. Link <jasmi...@uni-hamburg.de> wrote:

Dear Ron,

Thank you very much, for your interest an for emphasizing the relevance of my thesis in the context of SRM and CDR.

Please, feel free to add me to the CDR list as well.

The detailed potential side effects networks I have published only for SAI (Figure 3.1.) and BECCS (Figure 3.2. in the same EuTRACE final report) so far. I have made one for afforestion, but that has not been as requested and intriguing for discussion yet.

I think, some of the potential indirect side effects of SRM - such as a potential reduction in individual mitigation efforts - might also apply to a large-scale CDR. Depending on the method of CDR and its public presence (like artificial trees), this might also be true for even relatively small applications of CDR. Thus, it might be important to estimate in advance, if the contruction of one artifical tree rather decreases carbon emissions, by capturing them, or in total rather increases carbon emissions, by triggering passengers to feel free to emit more carbon.

Best regards,

Jasmin


Am 15.08.2020 um 19:47 schrieb Ronal Larson:
Dr.  Link and list:

<snip>

Reply all
Reply to author
Forward
0 new messages