Gregory Benford raised the issue of reaching agreement on the global
scale and how problematic it would be. If for example the arctic
council of nations agreed to initiate a decade-long field trial to
cool the arctic by 1 C, how could they deal with complaints from
nations who felt that this experiment had induced a negative change in
their climate?
In any given year there are floods, droughts, heatwaves, etc. but
during this field trial a fraction of these events would be attributed
(rightly or wrongly, partially or fully) to the intervention. Two
questions arise:
Could it be determined if any changes in the climate (outside of the
target area) had occurred as a result of the intervention during a 10
year trial?
How would aggrieved nations or peoples seek reparations for perceived
negative impacts (scientifically proven or otherwise)?
best regards,
Pete Irvine
PhD student
School of Geographical Sciences
University of Bristol
It may be possible to use one run of a climate model to get an
everywhere-to-everywhere transfer function for the effects of cloud
albedo spray. If several different models can agree then we should be
able to get the probability that a climate event has been caused by the
intervention or by chance.
The technique runs the climate models with pseudo-random changes of
spray from ON to OFF and then correlates the spray pattern with model
results at different places in the world. It looks as though a single
run can identify the transfer function to better than 2 % of the natural
variation. This will tell us where we should, and more importantly
SHOULD NOT spray.
I hope shortly to release an example of how it works in a simulated
extraction of small signals from a real climate record and will then be
trying to get climate modelers to use it.
Stephen
Emeritus Professor of Engineering Design
School of Engineering and Electronics
University of Edinburgh
Mayfield Road
Edinburgh EH9 3JL
Scotland
tel +44 131 650 5704
fax +44 131 650 5702
Mobile 07795 203 195
S.Sa...@ed.ac.uk
http://www.see.ed.ac.uk/~shs
--
The University of Edinburgh is a charitable body, registered in
Scotland, with registration number SC005336.
--
You received this message because you are subscribed to the Google Groups "geoengineering" group.
To post to this group, send email to geoengi...@googlegroups.com.
To unsubscribe from this group, send email to geoengineerin...@googlegroups.com.
For more options, visit this group at http://groups.google.com/group/geoengineering?hl=en.
Stephen, this transfer function is for tropospheric spraying, yes? The
strato cases are not nearly so predictable. I'd expect the tropo sprays
don't propagate far, anyway.
National legislation to protect geo experiments from lawsuits is
crucial. As David says, international treaties will help, but I'd
estimate they will take far longer, and delay is our enemy. I'd be
happy with the Arctic Council defending their experiments, as they have
clout and standing.
Gregory Benford
It seems that there is an assumption that cooling the arctic is
essential and trumps any potential impacts in other parts of the
world.
I'm not sure how it would be achieved but I believe that an in-depth
cost-benefit assesment of any proposed geoengineering scheme is
essential. For example, If the suspected impact of an arctic
intervention was a cooling of the arctic but also a change in
hydrology in another part of the world; surely to justify the
intervention it would have to be proved that the overall effect was
positive?
I can imagine their are a lot of tricky things to quantify, the
impacts of runaway climate change for tundra melt, the subjective
value of any predicted change in the climate, etc. (and a lot of
unquantifiable things besides) but I believe the principle should
remain.
Do you agree that this principle of a calculated or expected positive
global impact should be satisfied before any geoengineering
intervention s undertaken? The alternative seems to justify regions or
countries modifying regional climate for their benefit regardless of
detrimental impacts elsewhere.
regards,
Pete Irvine
> For more options, visit this group athttp://groups.google.com/group/geoengineering?hl=en.
Josh Horton
> For more options, visit this group athttp://groups.google.com/group/geoengineering?hl=en.
Your comments are precisely why having the Arctic Council manage it
will help deflect such UN-level chaos.
No nation is going to try to muscle the Council: they're too powerful.
Let them own the issue, because they have the most at stake. Worry not
about farmers who, in my experience on farms, will blame everything on
the weather (or the government of their own nation).
Meanwhile, take the high road of rational risk analysis, as John says.
Gregory
-----Original Message-----
From: Andrew Lockley <andrew....@gmail.com>
To: j...@cloudworld.co.uk
Cc: p.j.i...@googlemail.com; geoengineering
<geoengi...@googlegroups.com>; Mark Serreze
<ser...@kryos.colorado.edu>
Sent: Sun, Feb 21, 2010 2:56 pm
Subject: Re: [geo] Re: SRM geoengineering: how to deal with the losers?
I don't agree with John's reasoning. Whilst a rational and
all-powerful government might use such an approach, this is not likely
to be the case in practice. Realpolitik suggests that people will look
at only the downside of geoengineering when it affects them personally.
If you're a farmer with a failed crop, the benefit of constraining
methane releases won't seem a worthwhile trade off. Your government
may well agree.
It's the fine detail of the handling of these matters, and proper
governance surrounding them, that will prevent governments 'sinking the
cloud ships', either literally or metaphorically. If these concerns
aren't handled properly, I could even imagine countries deliberately
worsening the greenhouse effect to counteract geoengineering schemes.
Releasing HFCs/CFCs would be a very simple way to do this.
A
On 21 February 2010 14:50, John Nissen <j...@cloudworld.co.uk>
wrote:
Hello Peter,
To answer your question on cost/benefit, we need to analyse the
balanceof the risks of geoengineering to cool the Arctic against the
risks ofnot geoengineering. This balancing of risks needs to take into
accountthe timescale of processes in the Arctic which give rise to the
risksof not geoengineering. For example, if these processes were
veryunlikely to build up significant positive climate forcing or sea
levelrise over the next few decades, one could be relaxed
aboutgeoengineering. But there is data about the Arctic sea ice
retreat,shown in a single graph, that gives much cause for concern.
The graphhas been reproduced in the Copenhagen diagnosis report [1].
The graph is derived from one in this 2007 paper [2], by
Stroeve,Serreze et al, with additions for 2007 and 2008. Note that
this paperwas submitted before the abrupt retreat of sea ice to achieve
a recordlow September 2007.
In the light of such developments in the Arctic, and revised
timingestimates, we should prepare a risk analysis, but it will be in
aseparate thread.
Note that, in terms of probabilistic risk assessment (PRA) [3], we
haverisk defined thus:
In a PRA, risk is characterized by two quantities: the magnitude
(severity) of the possible adverseconsequence(s), and
the likelihood (probability) of occurrence of each consequence.
Conventional cost-benefit analysis [4] is not quite what we are
lookingfor, because it assumes the status quo is acceptable:
Cost–benefit analysis is typically used by governments to
evaluatethedesirability of a given intervention. It is heavily used in
today'sgovernment. It is an analysis of the cost effectiveness of
differentalternatives in order to see whether the benefits outweigh the
costs.The aim is to gauge the efficiency of the intervention relative
tothestatus quo. The costs and benefits of the impacts of
aninterventionare evaluated in terms of the public's willingness to pay
for them(benefits) or willingness to pay to avoid them (costs).
However we can try to "gauge the efficiency of the intervention"
relativeto no intervention in our analysis. We also need to
takeprobabilities and timescale into account, in order to compare
immediateintervention with delayed intervention. Of course, the
"costs" ofintervention must include the risks of adverse consequences.
Cheers from Chiswick,
John
[1] See fig 13 page 30:
http://www.ccrc.unsw.edu.au/Copenhagen/Copenhagen_Diagnosis_LOW.pdf
[2] Published in Geophysical Research Letters and reproduced here:
http://www.smithpa.demon.co.uk/GRL%20Arctic%20Ice.pdf
[3] http://en.wikipedia.org/wiki/Probabilistic_risk_assessment
[4] http://en.wikipedia.org/wiki/Cost-benefit_analysis
---
p.j.irvine wrote: dear Gregory Benford,It seems that there is an
assumption that cooling the arctic isessential and trumps any potential
impacts in other parts of theworld.I'm not sure how it would be
achieved but I believe that an in-depthcost-benefit assesment of any
proposed geoengineering scheme isessential. For example, If the
suspected impact of an arcticintervention was a cooling of the arctic
but also a change inhydrology in another part of the world; surely to
justify theintervention it would have to be proved that the overall
effect waspositive?I can imagine their are a lot of tricky things to
quantify, theimpacts of runaway climate change for tundra melt, the
subjectivevalue of any predicted change in the climate, etc. (and a lot
ofunquantifiable things besides) but I believe the principle
shouldremain.Do you agree that this principle of a calculated or
expected positiveglobal impact should be satisfied before any
geoengineeringintervention s undertaken? The alternative seems to
justify regions orcountries modifying regional climate for their
benefit regardless ofdetrimental impacts elsewhere.regards,Pete
IrvineOn Feb 20, 4:44 pm, xbenf...@aol.com wrote: All:Stephen,
this transfer function is for tropospheric spraying, yes? Thestrato
cases are not nearly so predictable. I'd expect the tropo spraysdon't
propagate far, anyway.National legislation to protect geo experiments
from lawsuits iscrucial. As David says, international treaties will
help, but I'destimate they will take far longer, and delay is our
enemy. I'd behappy with the Arctic Council defending their experiments,
as they haveclout and standing.Gregory Benford-----Original
Message-----From: David Schnare <dwschn...@gmail.com>Cc:
geoengineering <geoengi...@googlegroups.com>Sent: Sat, Feb
20, 2010 7:56 amSubject: Re: [geo] SRM geoengineering: how to deal with
the losers?Pete: My sense is that who will be a winner or loser is an
unknowableunknown. While I agree with Professor Salter that we could
take a stabat it, and indeed others on this list have made some
preliminaryattempts to look at precipitation issues, the sad reality is
thatmodels are dreadfully inadequate for regional analyses and
greatercomputing power will not be enough to make them much more
useful. In the absence of quality information, humankind has used a
best guess,iterative approach, achieved through international treaty.
Mysuspicion is that will be what will have to happen here to
addresswinners and losers. I am concerned, however, that such an
approachwill not work if potential "losers" demand stipulated
reparations. SchnareOn Sat, Feb 20, 2010 at 9:38 AM,
p.j.irvine<p.j.irv...@googlemail.com> wrote:It is likely that any
SRM geoengineering intervention will createwinners and losers and some
nations may always be against anyintervention whatsoever.Gregory
Benford raised the issue of reaching agreement on the globalscale and
how problematic it would be. If for example the arcticcouncil of
nations agreed to initiate a decade-long field trial tocool the arctic
by 1 C, how could they deal with complaints fromnations who felt that
this experiment had induced a negative change intheir climate?In any
given year there are floods, droughts, heatwaves, etc. butduring this
field trial a fraction of these events would be attributed(rightly or
wrongly, partially or fully) to the intervention. Twoquestions
arise:Could it be determined if any changes in the climate (outside of
thetarget area) had occurred as a result of the intervention during a
10year trial?How would aggrieved nations or peoples seek reparations
for perceivednegative impacts (scientifically proven or otherwise)?best
regards,Pete IrvinePhD studentSchool of Geographical SciencesUniversity
of Bristol--You received this message because you are subscribed to the
GoogleGroups "geoengineering" group.To post to this group, send email
to geoengi...@googlegroups.com.To unsubscribe from this group, send
email togeoengineeri...@googlegroups.com.For more options,
visit this group
athttp://groups.google.com/group/geoengineering?hl=en.--David W.
SchnareCenter for Environmental Stewardship--You received this message
because you are subscribed to the GoogleGroups "geoengineering"
group.To post to this group, send email to
geoengi...@googlegroups.com.To unsubscribe from this group, send
email togeoengineeri...@googlegroups.com.For more options,
Josh
On Feb 21, 6:03 pm, xbenf...@aol.com wrote:
> Andrew:
>
> Your comments are precisely why having the Arctic Council manage it
> will help deflect such UN-level chaos.
>
> No nation is going to try to muscle the Council: they're too powerful.
> Let them own the issue, because they have the most at stake. Worry not
> about farmers who, in my experience on farms, will blame everything on
> the weather (or the government of their own nation).
>
> Meanwhile, take the high road of rational risk analysis, as John says.
>
> Gregory
>
>
>
> -----Original Message-----
> From: Andrew Lockley <andrew.lock...@gmail.com>
> To: j...@cloudworld.co.uk
>
> Cc: p.j.irv...@googlemail.com; geoengineering
> <geoengi...@googlegroups.com>; Mark Serreze
> <serr...@kryos.colorado.edu>
> to geoengineer...@googlegroups.com.To unsubscribe from this group, send
> email togeoengineering+unsubscr...@googlegroups.com.For more options,
> visit this group
> athttp://groups.google.com/group/geoengineering?hl=en.--DavidW.
> SchnareCenter for Environmental Stewardship--You received this message
> because you are subscribed to the GoogleGroups "geoengineering"
> group.To post to this group, send email to
> geoengineer...@googlegroups.com.To unsubscribe from this group, send
> email togeoengineering+unsubscr...@googlegroups.com.For more options,
The probabilistic risk assessment sounds like a good idea but from the
wiki entry on the subject it seems to be limited to a single metric, I
guess future global warming potential is a reasonable choice in this
case. I may have misinterpreted this but with a single metric surely
you would miss important things outside of your target such as
regional precipitation changes. Could you clarify this part?
cheers,
Pete
On Feb 21, 2:50 pm, John Nissen <j...@cloudworld.co.uk> wrote:
> Hello Peter,
> To answer your question on cost/benefit, we need to analyse the balance of the risks of geoengineering to cool the Arctic against the risks of not geoengineering. This balancing of risks needs to take into account the timescale of processes in the Arctic which give rise to the risks of not geoengineering. For example, if these processes were very unlikely to build up significant positive climate forcing or sea level rise over the next few decades, one could be relaxed about geoengineering. But there is data about the Arctic sea ice retreat, shown in a single graph, that gives much cause for concern. The graph has been reproduced in the Copenhagen diagnosis report [1].
> The graph is derived from one in this 2007 paper [2], by Stroeve, Serreze et al, with additions for 2007 and 2008. Note that this paper was submitted before the abrupt retreat of sea ice to achieve a record low September 2007.
> In the light of such developments in the Arctic, and revised timing estimates, we should prepare a risk analysis, but it will be in a separate thread.
> Note that, in terms of probabilistic risk assessment (PRA) [3], we have risk defined thus:In a PRA, risk is characterized by two quantities:the magnitude (severity) of the possible adverse consequence(s), andthe likelihood (probability) of occurrence of each consequence.
> Conventional cost-benefit analysis [4] is not quite what we are looking for, because it assumes the status quo is acceptable:Cost–benefit analysis is typically used by governments to evaluate the desirability of a given intervention. It is heavily used in today's government. It is an analysis of the cost effectiveness of different alternatives in order to see whether the benefits outweigh the costs. The aim is to gauge the efficiency of the interventionrelative to the status quo. The costs and benefits of the impacts of an intervention are evaluated in terms of the public's willingness to pay for them (benefits) or willingness to pay to avoid them (costs).
> However we can try to "gauge the efficiency of the intervention"relative to no interventionin our analysis. We also need to take probabilities and timescale into account, in order to compare immediate intervention with delayed intervention. Of course, the "costs" of intervention must include the risks of adverse consequences.
> Cheers from Chiswick,
> John
> [1] See fig 13 page 30:http://www.ccrc.unsw.edu.au/Copenhagen/Copenhagen_Diagnosis_LOW.pdf
> [2] Published in Geophysical Research Letters and reproduced here:http://www.smithpa.demon.co.uk/GRL%20Arctic%20Ice.pdf
> [3]http://en.wikipedia.org/wiki/Probabilistic_risk_assessment
> [4]http://en.wikipedia.org/wiki/Cost-benefit_analysis
> ---
> p.j.irvine wrote:dear Gregory Benford, It seems that there is an assumption that cooling the arctic is essential and trumps any potential impacts in other parts of the world. I'm not sure how it would be achieved but I believe that an in-depth cost-benefit assesment of any proposed geoengineering scheme is essential. For example, If the suspected impact of an arctic intervention was a cooling of the arctic but also a change in hydrology in another part of the world; surely to justify the intervention it would have to be proved that the overall effect was positive? I can imagine their are a lot of tricky things to quantify, the impacts of runaway climate change for tundra melt, the subjective value of any predicted change in the climate, etc. (and a lot of unquantifiable things besides) but I believe the principle should remain. Do you agree that this principle of a calculated or expected positive global impact should be satisfied before any geoengineering intervention s undertaken? The alternative seems to justify regions or countries modifying regional climate for their benefit regardless of detrimental impacts elsewhere. regards, Pete Irvine On Feb 20, 4:44 pm,xbenf...@aol.comwrote:All: Stephen, this transfer function is for tropospheric spraying, yes? The strato cases are not nearly so predictable. I'd expect the tropo sprays don't propagate far, anyway. National legislation to protect geo experiments from lawsuits is crucial. As David says, international treaties will help, but I'd estimate they will take far longer, and delay is our enemy. I'd be happy with the Arctic Council defending their experiments, as they have clout and standing. Gregory Benford -----Original Message----- From: David Schnare<dwschn...@gmail.com>Cc: geoengineering<geoengi...@googlegroups.com>Sent: Sat, Feb 20, 2010 7:56 am Subject: Re: [geo] SRM geoengineering: how to deal with the losers? Pete: My sense is that who will be a winner or loser is an unknowable unknown. While I agree with Professor Salter that we could take a stab at it, and indeed others on this list have made some preliminary attempts to look at precipitation issues, the sad reality is that models are dreadfully inadequate for regional analyses and greater computing power will not be enough to make them much more useful. In the absence of quality information, humankind has used a best guess, iterative approach, achieved through international treaty. My suspicion is that will be what will have to happen here to address winners and losers. I am concerned, however, that such an approach will not work if potential "losers" demand stipulated reparations. Schnare On Sat, Feb 20, 2010 at 9:38 AM, p.j.irvine <p.j.irv...@googlemail.com> wrote: It is likely that any SRM geoengineering intervention will create winners and losers and some nations may always be against any intervention whatsoever. Gregory Benford raised the issue of reaching agreement on the global scale and how problematic it would be. If for example the arctic council of nations agreed to initiate a decade-long field trial to cool the arctic by 1 C, how could they deal with complaints from nations who felt that this experiment had induced a negative change in their climate? In any given year there are floods, droughts, heatwaves, etc. but during this field trial a fraction of these events would be attributed (rightly or wrongly, partially or fully) to the intervention. Two questions arise: Could it be determined if any changes in the climate (outside of the target area) had occurred as a result of the intervention during a 10 year trial? How would aggrieved nations or peoples seek reparations for perceived negative impacts (scientifically proven or otherwise)? best regards, Pete Irvine PhD student School of Geographical Sciences University of Bristol -- You received this message because you are subscribed to the Google Groups "geoengineering" group. To post to this group, send email togeoeng...@googlegroups.com. To unsubscribe from this group, send email togeoengineeri...@googlegroups.com. For more options, visit this group athttp://groups.google.com/group/geoengineering?hl=en. -- David W. Schnare Center for Environmental Stewardship -- You received this message because you are subscribed to the Google Groups "geoengineering" group. To post to this group, send email togeoeng...@googlegroups.com. To unsubscribe from this group, send email togeoengineeri...@googlegroups.com. For more options, visit this group athttp://groups.google.com/group/geoengineering?hl=en.
If our climate models were good enough (at a local level) to predict
inequities, offsetting compensation might indeed be possible even in
an international context. (For more see the link to Earth Times at
www.bunzl.org). Assume they are not good enough (at least for now).
Assume also, as I and others have argued (see "A Test for
Geoengineering?" at www.bunzl.org ), that testing for such inequities
is not possible to do within a reasonable amount of time. Then the
issue becomes one of negotiating ahead of time for unkown inequeities
assumming one only want to proceed with international aquiesence. I
see this as a tough condition to satisfy because it involves a
decisoin not only under uncertainty but also ignorance. We don't know
the probablity or the scale of potential inequities or untoward
effects.
Martin Bunzl
Professor of Philosophy,
Rutgers University
What am I missing?
Paul