ArticlePDF Available

The Risk of Termination Shock From Solar Geoengineering


Abstract and Figures

If solar geoengineering were to be deployed so as to mask a high level of global warming, and then stopped suddenly, there would be a rapid and damaging rise in temperatures. This effect is often referred to as termination shock, and it is an influential concept. Based on studies of its potential impacts, commentators often cite termination shock as one of the greatest risks of solar geoengineering. However, there has been little consideration of the likelihood of termination shock, so that conclusions about its risk are premature. This paper explores the physical characteristics of termination shock, then uses simple scenario analysis to plot out the pathways by which different driver events (such as terrorist attacks, natural disasters, or political action) could lead to termination. It then considers where timely policies could intervene to avert termination shock. We conclude that some relatively simple policies could protect a solar geoengineering system against most of the plausible drivers. If backup deployment hardware were maintained and if solar geoengineering were implemented by agreement among just a few powerful countries, then the system should be resilient against all but the most extreme catastrophes. If this analysis is correct, then termination shock should be much less likely, and therefore much less of a risk, than has previously been assumed. Much more sophisticated scenario analysis—going beyond simulations purely of worst-case scenarios—will be needed to allow for more insightful policy conclusions.
Content may be subject to copyright.
Earth’s Future
The Risk of Termination Shock From Solar Geoengineering
Andy Parker1and Peter J. Irvine2
1Institute for Advanced Sustainability Studies, Potsdam, Germany, 2John A. Paulson School of Engineering and Applied
Sciences, Harvard University, Cambridge, MA, USA
Abstract If solar geoengineering were to be deployed so as to mask a high level of global warming,
and then stopped suddenly, there would be a rapid and damaging rise in temperatures. This effect is often
referred to as termination shock, and it is an influential concept. Based on studies of its potential impacts,
commentators often cite termination shock as one of the greatest risks of solar geoengineering. However,
there has been little consideration of the likelihood of termination shock, so that conclusions about its
risk are premature. This paper explores the physical characteristics of termination shock, then uses simple
scenario analysis to plot out the pathways by which different driver events (such as terrorist attacks, nat-
ural disasters, or political action) could lead to termination. It then considers where timely policies could
intervene to avert termination shock. We conclude that some relatively simple policies could protect a
solar geoengineering system against most of the plausible drivers. If backup deployment hardware were
maintained and if solar geoengineering were implemented by agreement among just a few powerful
countries, then the system should be resilient against all but the most extreme catastrophes. If this analy-
sis is correct, then termination shock should be much less likely, and therefore much less of a risk, than has
previously been assumed. Much more sophisticated scenario analysis— going beyond simulations purely
of worst-case scenarios—will be needed to allow for more insightful policy conclusions.
1. Introduction
Solar geoengineering (also known as solar radiation management or SRM) is a proposal for slowing the rise
in global temperatures, or even reversing it, by reflecting a small fraction of inbound solar energy back into
space. It has been proposed as a potential method for reducing the climate risks to which the planet is com-
mitted from past greenhouse gas (GHG) emissions. The SRM technique currently receiving the most atten-
tion is stratospheric aerosol injection, which would work by spraying reflective aerosol particles into the
stratosphere to reduce the amount of solar energy reaching the Earth (Rasch et al., 2008); (Crutzen, 2006).
SRM only masks the warming effects of GHGs and is not designed to reduce their concentrations in the
atmosphere. Therefore, if SRM were ever used to mask a high level of warming and its deployment were
terminated suddenly, temperature would rebound toward the levels they would have reached without the
geoengineering (Brovkin et al., 2009; Irvine et al., 2012; Jones et al., 2013; Llanillo et al., 2010; Matthews &
Caldeira, 2007; McCusker et al., 2012, 2014). This effect is referred to as “termination shock” or “termination
effect.” Termination shock could be very damaging for natural and human systems as the rate of warm-
ing would probably be much higher than that otherwise expected under anthropogenic climate change
(Irvine et al., 2012; Llanillo et al., 2010; Matthews & Caldeira, 2007; McCusker et al., 2014), meaning that both
ecosystems and human societies would have less time to adapt to the rapidly changing new conditions
(McCormack et al., 2016; Trisos et al., 2018).
1.1. An Influential Concept
Termination shock has been a very influential concept in both academic and popular commentary, and is
often cited as one of the most serious threats from the development and deployment of SRM (Hamilton,
2013). The idea has also informed a range of policy messages. In numerous academic papers (see, e.g.,
Brovkin et al., 2009; Clark et al., 2016; Pidgeon et al., 2013; Vaughan & Lenton, 2012) and high-profile
commentary pieces (Appell, 2013; Dean, 2010; Klein, 2014; Pierrehumbert, 2015; Plumer, 2014), it is claimed
that once deployed, SRM would need to be maintained for centuries or even millennia to avoid the risk
Key Points::
• We found no reason to dismiss the
threat of termination shock.
Managing this risk should be a key
concern if solar radiation
management (SRM) is ever
considered for use
• But if current projections about
stratospheric aerosols injection
characteristics prove accurate, it
should be easy to build an SRM
system that is resilient and robust
• The motivation to avoid termination
shock would be strong. Where many
parties can maintain SRM, it cannot
be terminated unilaterally
Correspondence to:
A. Parker,
Parker, A., & Irvine, P. J. (2018). The Risk
of Termination Shock From Solar
Geoengineering, Earth’s Future,9999.
Received 31 OCT 2017
Accepted 7 FEB 2018
Accepted article online 13 FEB 2018
© 2018 The Authors.
This is an open access article under
the terms of the Creative Commons
License, which permits use and distri-
bution in any medium, provided the
original work is properly cited, the use
is non-commercial and no modifica-
tions or adaptations are made.
Earth’s Future 10.1002/2017EF000735
of termination shock. It has also been argued that due to the risk of termination shock, SRM should only
be considered an option of last resort for emergency use (Llanillo et al., 2010) or that the cooling from any
deployment of SRM should be limited to a level that would not cause a dangerous temperature rebound
in the event of termination (Kosugi, 2013). Aggressive mitigation would need to be a prerequisite of SRM
deployment in the opinion of McCusker et al. (2014), while Kruger (2015) takes this idea one stage further,
arguing not only that SRM should not be deployed without an “exit plan” in the form of very large-scale
carbon dioxide removal technologies but also that those costs should be seen as an integral component
of the costs of SRM.
Most model studies of SRM have simulated the same general scenario: SRM is deployed to entirely halt
future temperature increases under a scenario of very high future GHG emissions, then at a future date the
deployment is terminated instantly and permanently. For example, the G2, G3, and G4 scenarios of the Geo-
engineering Model Intercomparison Project, which have been used as the basis for many modeling studies,
all follow this approach (Berdahl et al., 2014; Jones et al., 2013; Kravitz et al., 2011). This modeling conven-
tion shows termination shock in its starkest relief, as a large degree of SRM cooling, stopped suddenly and
permanently, produces the clearest climate response. However, this is the most extreme termination sce-
nario and there are many other possibilities. SRM need not be deployed either at a large magnitude or
indefinitely to usefully reduce the risks of climate change (Kosugi, 2013; MacMartin et al., 2014). Further-
more, large-magnitude SRM deployment could be phased out gradually (Irvine et al., 2012), or it could be
interrupted for a time and later restored.
Almost all research papers to date have focused on the impacts of termination shock, ignoring how or why
termination might occur, or how likely it might be (with notable exceptions of Barrett, 2014; Baum et al.,
2013; Goes et al., 2011). Typically, analysis and commentary either proceed on the basis that some undefined
eventuality could cause the sudden and permanent cessation of SRM (Boucher et al., 2013; Plumer, 2014;
Rayner et al., 2013; Specter, 2012; Zürn & Schäfer, 2013) or speculate in general terms that termination could
be caused by events such as terrorist attacks or natural disasters (Appell, 2013; Bellamy & Lezaun, 2017;
Hamilton, 2013; Olson, 2011; Pierrehumbert, 2015).
1.2. Aims and Structure of This Paper
This study aims to address two critical gaps in the literature on termination shock: first, providing some defi-
nitions and boundaries for what would and would not constitute a termination shock and second, exploring
the factors that would affect the likelihood of the occurrence of a termination shock.
The paper is structured as follows. Section 2 defines termination shock and describes the conditions that
must be met for a termination shock to occur. Section 3 plots out the steps by which a range of proposed
driver events (such as terrorism, economic collapse, natural disasters, or the discovery of damaging side
effects) could lead to a termination shock, and considers where appropriate policy responses might increase
the robustness and the resilience of an SRM system against disruptions. In Sections 4 and 5, we draw out
some implications of our analysis.
1.3. Assumptions About Deployment Methods and Costs
For the sake of simplicity, we focus on stratospheric aerosols injection (SAI), delivered by high-flying jets, as
the method of SRM deployment. SAI would involve releasing aerosols, or aerosol precursors, into the strato-
sphere, where they would have an e-folding time of about one year and particles would circulate to form a
global aerosol cloud (Robock et al., 2008). A number of proposals for delivering aerosols to the stratosphere
have been assessed, and high-flying jets have been judged feasible in part as they would require little tech-
nical innovation (McClellan et al., 2012; Moriyama et al., 2016; Robock et al., 2009). In line with the highest
current estimates for a very large-scale deployment, we assume it would cost an initial outlay of about $50
billion for the hardware needed for SAI and $12.5 billion per year for deployment.
2. Defining Termination Shock
We define a termination shock as a rapid and substantial rise in global temperatures following a cessation
of SRM deployment.
There are thus three criteria that need to be met for termination shock to occur: the amount of SRM
cooling would need to be large, it would need to be terminated suddenly, and it would need to stay off for
Earth’s Future 10.1002/2017EF000735
a substantial period. To better determine what would constitute a termination shock, it is therefore
necessary to provide answers to the following questions:
How large must the SRM cooling effect be before termination shock becomes possible?
How slowly would large-scale SRM need to be phased out to avoid causing a rapid and substantial
How long would a disruption to large-scale SRM have to persist to cause a substantial warming?
2.1. How Large Must the Cooling Effect of Solar Geoengineering Be for There to Be the Potential
for a Rapid and Substantial Warming?
If SRM of any scale were to be terminated suddenly, a rapid warming would follow. The magnitude of this
warming would depend on the amount of cooling that was being exerted at the time of termination. Most
of the modeling studies to date have simulated quite dramatic termination scenarios, where SRM is off-
setting the warming effect of many decades of GHG emissions at the time of termination so that a rapid,
substantial rise in global temperature follows. At the other end of the spectrum, it is also clear that a very
small magnitude deployment of SRM poses very little risk as there would not be a substantial rise in temper-
atures if it were terminated. Whether or not solar geoengineering could be deployed at such a scale that it
would have an appreciable effect on global temperatures without carrying a risk of a rapid and substantial
warming if it were to be terminated is therefore an important question (Kosugi, 2013).
There is no widely accepted definition of what would constitute a substantial or dangerous rate of
warming—what therefore would constitute termination “shock.” However, useful benchmarks can be
found in recent climate trends and the temperature response to the representative concentration path-
ways (RCPs), the scenarios of future GHG emissions used by the climate modeling community to simulate
the impacts of climate change. The change in temperature between the last two decades (1997– 2006 and
2007–2016) was approximately 0.1C (Morice et al., 2012). This decadal rate of change is low compared to
other periods in the recent past due to the strong el-Nino event of 1997– 1998 and the so-called global
warming hiatus (1998–2013). The RCP 2.6 scenario represents an optimistic scenario of emissions cuts with
implementation of substantial negative emissions before 2100 that reaches a radiative forcing of 2.6 W m2
by 2100 and keeps temperatures below 2C above the preindustrial in most models. In this “best case”
scenario, the peak rate of warming is around 0.2C per decade. At the other end of the spectrum, in the RCP
8.5 scenario, models project a rise in temperature of more than 4C by 2100 and a peak rate of warming
of around 0.5C per decade (Knutti & Sedlacek, 2013). On the basis of these examples, we suggest that a
substantial rate of additional warming for our purposes is of the order a few tenths of a degree Celsius
per decade. This is in line with the threshold for a termination shock of 0.2C per decade for a termination
suggested by Kosugi (2013).
If a threshold for termination shock is defined in this manner, then it is also possible to define a cooling
threshold below which an instantaneous cessation of SRM would lead to less than this threshold rate of
additional warming. It is possible to estimate this threshold magnitude of cooling forcing by evaluating the
response of global temperatures to an instantaneous change in forcing. Simulations by a range of Earth
system models project that in response to an instantaneous quadrupling of CO2concentrations (an instant
change in radiative forcing of about 7.4 Wm2) around one half of the equilibrium temperature response
occurs within a decade, with around three quarters occurring within 100 years (Caldeira & Myhrvold, 2013).
Therefore, the cooling threshold would be roughly double the threshold decadal rate of warming. Adopt-
ing the suggestion of Kosugi (2013) of a maximum acceptable rate of warming of 0.2C implies a cooling
threshold of 0.4C. Modest deployments of SRM, at or below this threshold, could still have appreciable
benefits, see Kosugi (2013) and MacMartin et al. (2014). Furthermore, it follows that if SRM were exerting
less cooling than this threshold, there would not be a risk of a substantial rapid additional warming. Note
that the warming effects of termination would be in addition to the ongoing climate response to other
anthropogenic forcings and the total rate of warming could therefore be higher.
Therefore, for the purposes of defining termination shock, we suggest that a cooling of order a few tenths
of a degree Celsius, equivalent to roughly a decade of warming, defines the approximate magnitude of SRM
deployment for which the risk of termination shock could become appreciable. For larger deployments of
solar geoengineering that offset many decades of warming there would be a clear risk of a rapid and sub-
stantial global warming, whereas for a smaller deployment that offset less than a tenth of a degree Celsius,
Earth’s Future 10.1002/2017EF000735
there would not. This suggests that if SRM forcing were ramped up slowly, there would be a certain period
of time before the cooling were large enough to pose a potential risk of termination shock.
2.2. How Quickly Could Substantial SRM Forcing Be Phased Out Without Causing a Substantial
and Rapid Warming?
If a larger SRM deployment were phased out sufficiently slowly, the rate of warming could be limited, and
termination shock avoided, even where it exerted a very large cooling effect. As we argued in the previous
section, guidance can be found from the simulated climate response to the RCP emissions scenarios. Limit-
ing the rate of warming to that expected in the extreme RCP 8.5 scenario would imply phasing out SRM at
C per decade, but this is generally regarded as likely to be dangerously rapid. Limiting the rate
of warming to the maximum expected in the optimistic RCP 2.6 scenario, 0.2C per decade, would imply a
phase out of 50 years per degree Celsius of cooling.
2.3. How Long Would a Disruption to Large-Scale SRM Have to Persist to Cause a Substantial
If something disrupted large-forcing SRM deployment activities and they were not resumed, then a rapid
warming would follow. However, if SRM deployment were restarted before a substantial increase in temper-
atures had been realized, then the impacts of the disruption would be much less than for a full termination
shock. The length of this “buffer period” during which forcing can be restored would be determinedby three
factors: the timescale on which the radiative forcing from deployment activities decays, the timescale on
which global temperatures respond to changes in the radiative forcing, and the magnitude of the forcing.
First, instantly halting deployment would not stop the aerosols which had been released up until that point
from scattering light and so exerting a radiative forcing on the climate. From observations of the strato-
spheric aerosol burden following large volcanic eruptions, an e-folding time of roughly one year has been
calculated (Robock et al., 2008; Stenchikov et al., 1998). This means that the cooling effect of stratospheric
aerosol geoengineering would persist for many months after termination. It follows that with an e-folding
time of one year, 98% of the aerosol burdenwould remain after a week, 92% after a month, and that it would
take over eight months for the burden, and hence the forcing, to fall to under half of what it was at the time
of termination. However, this timescale would only be a matter of days for marine cloud brightening and
cirrus cloud thinning, as tropospheric aerosols only have a lifetime of days before they are rained out or fall
to the ground (Latham, 1990; Mitchell & Finnegan, 2009).
Second, it takes some time for surface air temperature to adjust to a change in radiative forcing because it
takes time for the Earth’s atmosphere and its oceans to warm. The process is not well described by a simple
e-folding time, but as we noted before, Caldeira and Myhrvold (2013) reported that around one quarter of
the equilibrium warming response to an instantaneous change in forcing occurs within a year, half within
a decade, and three quarters within 100 years. Cao et al. (2012) extend this analysis to shorter timescales,
finding that global temperature would rise to around 7% of its equilibrium value within a month.
Third, the smaller the magnitude of forcing at the time of termination, the longer it will take to cross a given
threshold of warming. However, the effect of disruptions to very large-scale deployments which would lead
to substantial, rapid warming will still take some time to do so for the reasons given above.
We do not attempt to precisely define the length of disruption required to produce a substantial warming,
but we do provide a rough estimate of the timescale based on the results described above. The timescale
of about one year for the decay of forcing following a cessation of stratospheric aerosol geoengineering
and the delayed response of temperature to a change in radiative forcing means that it would take several
years for the temperature to rise by an appreciable fraction of the eventual response followinga cessation of
stratospheric aerosol geoengineering. This means that if large-scale stratospheric aerosol geoengineering
were redeployed at the original level within a few months of a disruption, then there would not have been
time for an appreciable rise in temperatures to occur. However, for marine cloud brightening and cirrus
cloud thinning, the days-long lifetime of the forcing effect implies the timescale during which they can
be safely restored would be shorter. However, it would still take weeks before a substantial rise in global
temperature could occur due to the thermal inertia of the land and surface ocean. Therefore, there would be,
at minimum, a buffer period of the order of weeks or months, depending on the form and magnitude of SRM
deployed, during which forcing could be restored in order to avert a substantial increase in temperatures.
Earth’s Future 10.1002/2017EF000735
In Section 4, we consider the broader impacts of interruptions to SRM and make suggestions for future
research on this topic.
3. Pathways to Termination
An interruption to the political, economic, or technical capabilities to maintain SRM could result in an inter-
ruption to the SRM deployment. In this section, we plot out the pathways by which different events (referred
to here as drivers) could cause such interruptions and lead to termination, and we identify the points at
which timely policy intervention could avert termination shock.
By sketching out the pathways from driver events to termination, we attempt to provide a foundation for
analysis of this neglected issue. The reader can see the factors that we consider to be influential in deter-
mining whether termination shock occurs and can easily identify points of disagreement, or indeed, more
accurately target their praise.
We have divided the pathways to termination into two classes: (1) where external drivers force termina-
tion even where humanity wants SRM to continue and (2) where termination would be an elective political
decision, taken when continuation of SRM were still possible.
3.1. Forced Termination of SRM
Drivers in this class would entail external events forcing termination even where there was a desire for SRM
to continue. Forced termination might be caused by destruction of the SRM delivery infrastructure, or inter-
ruptions to the economic or political capacity to maintain deployment.
3.1.1. Forced Termination Pathway 1: Destruction of Deployment Infrastructure
It has been suggested that destruction of the SRM deployment infrastructure, for instance by terrorist attack,
could cause termination shock (Bellamy & Lezaun, 2017; Olson, 2011; Pierrehumbert, 2015). By plotting out
the pathway, it becomes clear that a number of conditions would have to be met for destruction of infras-
tructure to lead to termination shock (see Figure 1).
First, an attack must overcome any defensive systems and disable a large fraction of the deployment capa-
bility. Achieving more than 1 W m2of cooling through stratospheric aerosol injection would require fleets
of hundreds of aircrafts, most likely operating out of numerous air fields around the world, so a physical
attack would have to be extremely well planned, coordinated, and supported to be effective. A cyberattack
might conceivably be able to disable an entire system at once, however.
Second, even if defenses failed, a termination shock would still not be automatic. There would still be a
buffer period of months (at least) during which the system could be repaired or a backup system deployed
before the global temperature rise became significant. If no replacement SRM system weredeployed during
this time, termination shock would occur.
It should be relatively simple to greatly increase the robustness of an SRM system against destruction of
deployment infrastructure. Basic defenses of the delivery equipment, such as those that guard nuclear
power plants or military bases, are an obvious precaution. Also, the more geographically distributed the
aerosol delivery system, the harder it would be to disable a large fraction of it.
Even if a terrorist attack were successful in stopping SRM globally, relatively easy and cheap policy choices
would allow operations to restart soon after disruption. If any capable party, anywhere around the world,
Figure 1. The pathway to termination shock from a terrorist attack on the delivery system.
Earth’s Future 10.1002/2017EF000735
kept backup SRM delivery hardware, it could be redeployed to maintain the SRM cooling before tempera-
tures started to rise rapidly.
3.1.2. Forced Termination Pathway 2: Catastrophe, or the Destruction of Economic or Political
Capacity to Maintain SRM
Beyond the destruction of the SRM deployment infrastructure, termination could be forced if a large catas-
trophe (such as a natural disaster, an economic collapse, or a war) were to interrupt the economic or political
ability to maintain an SRM program. Figure 2 outlines this pathway. First, the deployer(s) would have to suf-
fer an event that destroyed their economic and/or political capacity to maintain SRM. Second, there would
need to be no other actors capable of redeploying before temperatures rose rapidly. If neither the original
deployer nor any other party were capable of redeploying an SRM system in the months after a catastrophe,
then termination shock would occur.
The same measures that would increase the resilience of SRM against terrorist attacks would also increase
resilience against disasters that were local or regional in scope. If spare deployment capacity were main-
tained or numerous nations were capable of deployment, then the SRM system would be resilient against
catastrophes that were confined to one country or one region.
Not all catastrophes are regional though, and various global calamities (such as nuclear war, a global pan-
demic, or an economic collapse) have been suggested as possible drivers of termination (Barrett, 2014;
Hamilton, 2013; Olson, 2011). Amidst a true global catastrophe, one that destroyed the capacity of all com-
petent actors to maintain deployment, backup delivery equipment would not help.
While there should be no doubt that a global catastrophe could destroy global capacity to maintain SRM,
it is worth noting quite how destructive such an event would have to be. For example, consider how large
the economic damage would have to be before maintaining SRM became unaffordable. If nations were
to cooperate to maintain SRM— a proposition that we find reasonable given the projected damages of
termination shock—then global gross domestic product (GDP) would still need to drop by over 90% before
maintaining SRM would cost more than 1% of the collective postcatastrophe GDP of the world’s top 20
economies.1Even assuming no international cooperation, the catastrophe would still have to be colossal
to force SRM termination based on economic factors. Even if a disaster wiped off 70% of their GDPs, China
or the United States alone could still deploy SRM for less than 1% of their postcatastrophe GDPs (IMF, 2014).
To put these figures in context, arguably the greatest international calamities of the last century, World War
I and Spanish influenza combined, the Great Depression, and World War II caused GDP in Europe to fall by
13%, 8%, and 21%, respectively (The Maddison-Project, 2013).
Therefore, if spare deployment capacity were maintained, or could be brought online quickly, a catastrophe
would have to be on a scale unprecedented in modern history to force termination shock. Such events are
not inconceivable though, and an asteroid strike, global nuclear war, or catastrophic pandemic could force
an end to SRM even where there was a desire to maintain it. In this instance, few policies could prevent
termination. We reflect on this possibility in Section 4.3.
3.2. Elective Termination of SRM
So far, the paper has only considered factors that could force cessation of SRM where there was a continuing
commitment to maintaining it. It is also possible that termination would be a political choice. We assume
that it would be widely understood how damaging a termination shock would be, such that no party would
bring on termination shock out of ignorance. However, politicians vehemently opposed to geoengineering
could come to power, there could be a revolution, or a coalition of countries might decide that they were
suffering unacceptable harms from SRM and push for it to be stopped.
Figure 3 outlines the pathway from political opposition to SRM through to termination shock. First, actors
opposed to SRM would need to gain sufficient power to force an end to deployment within their own juris-
dictions. It would also need to be the case that they preferred termination shock to other forms of SRM
1Assuming that SRM deployment cost $50 billion in its first year (which is above the highest current cost projection),
and if the global economy did not grow any larger than it is today, global GDP would need to drop from nearly $62 trillion
down to $5 trillion (IMF, 2014).
Earth’s Future 10.1002/2017EF000735
Figure 2. The pathway to termination from a catastrophe.
Figure 3. The pathway to elective termination of SRM.
deployment (such as a different technique, different aerosol spraying locations, or different aerosol type),
or to any form of slower phasing out of deployment. Finally, they would need to have the power to impose
their will for termination on all other parties, all around the world, even those who wanted to maintain SRM.
A number of different policies might be able to reduce the drivers of elective termination. In particular,
decision-making mechanisms that reduced possible grievances over the impacts of SRM, or over the deci-
sion to deploy it in the first place, would reduce the level of antipathyto an ongoing geoengineering deploy-
ment. Making sure that deployment was agreed as widely as possible and was supported by strong support
for adaptation and compensation regimes could help reduce injustices and perceptions of injustice. Slowly
ramping up the SRM cooling, with extensive environmental monitoring before and after deployment, could
reduce the risks of damaging environmental effects being discovered only after the point where termina-
tion shock had become possible. The development of alternative SRM techniques or deployment methods,
which might maintain cooling while avoiding given environmental drawbacks, could allow the SRM system
to be modified to reduce undesired impacts. Finally, stopping SRM need not involve termination shock, as
parties pushing for an end to SRM might be open to a gradual phase out of deployment, reducing the rate
of temperature change and hence the impacts of termination (Irvine et al., 2012).
4. Discussion
Most analysis of termination shock to date has addressed two issues. Modeling studies have mostly sought
to quantify the climate response to a sudden and permanent cessation of large-scale SRM deployment,
while commentaries have discussed the policy implications of this worst-case possibility. Here, we have
addressed a number of key gaps in that analysis. We have looked beyond the extreme cases of termination
shock to create a more nuanced picture of the implications of stopping SRM deployment, and we have
suggested boundaries on what would constitute a termination shock. We have also made a first attempt
to methodically analyze the pathways that could lead to a cessation of SRM deployment, from terrorist
attacks and natural disasters to political decisions, considering also policy options that might be employed
to reduce the likelihood of termination shock.
4.1. The Physical Characteristics of Termination Shock
Clarifying the physical dimensions of what would constitute a termination shock leads us to conclude that
three conditions must be met for termination shock to occur:
1. SRM deployment would have to be exerting a substantial cooling.
2. Deployment would have to be terminated suddenly.
Earth’s Future 10.1002/2017EF000735
3. Disruptions to SRM deployment would have to persist for many months or longer.
From this, we conclude that the common claim “once you start SRM it cannot be stopped” is just not accu-
rate. Even if SRM were used to offset a large amount of GHG warming, it could be stopped without incurring
termination shock if it were phased out slowly. Additionally, it would be possible to suddenly terminate a
low level of SRM cooling (we tentatively suggest an upper limit of few tenths of a degree Celsius), with-
out risking impacts worthy of the name of termination shock (Kosugi, 2013; MacMartin et al., 2014). Some
might question whether using SRM to exert a few tenths of a degree of cooling would have any bene-
fits, but studies investigating the climatic differences between global warming of 1.5Cand2
C—the two
temperature goals cited in the Paris climate agreement— indicate that the extra 0.5C of warming could
mean significantly larger impacts on sea level rise, coral bleaching, agriculture in the tropics, and heat waves
(Schleussner et al., 2016).
Knowing that SRM cooling could be ramped up to a few tenths of a degree without fear of termination
shock also has implications for research. Scientists could explore the environmental impacts of global SRM
deployment knowing that the system could be safely turned off if any unacceptable impacts were discov-
ered. Such research could inform decisions as to whether to deploy SRM at greater intensities, past the point
at which termination shock becomes a concern. Future modeling work could help to quantify the level of
cooling required for a termination of SRM to be reliably detectable against natural variability.
Another important finding is that the years-long residence time of stratospheric aerosols and the substantial
thermal inertia in the climate system imply that it would take months for a disruption to SRM deployment
to result in a substantial change in global temperatures. This is crucial for analyzing the risks of termination
shock, as it means that humanity would have a period of several months in which to resume deployment
of SRM in the event of a disruption. Future modeling work should address the potential impacts of months-
or years-long interruptions to SRM deployment, and to evaluate how long such interruptions would have
to persist to be detectable.
4.2. The Motivation to Avoid Termination Shock
Reflecting on the physical characteristics of termination shock and our analysis of scenarios by which
driver events could lead to termination, we make two broad arguments:
1. There would be a strong motivation to maintain an SRM system and avoid termination shock.
2. The projected technical characteristics of SRM mean that it should be relatively easy to make a system
that would be highly robust and resilient against most proposed drivers of termination.
Regarding the motivation to avoid termination shock, once SRM is deployed to a level at which its termi-
nation would produce large damages, it seems self-evident that there would be a strong and widely held
incentive to maintain the system. The threat of termination shock is already a central concern about SRM
development, and we expect that this concern would only increase if the use of SRM ever appeared to be a
realistic prospect.
We draw parallels to the level of societal effort that goes toward protecting and maintaining critical infras-
tructure. Where damage to infrastructure would present immediate dangers, there is often a high level
of security (as with nuclear power plants or airports). Where prolonged interruption of the infrastructure
service would be dangerous, there is often investment in backup systems. For example, to ensure reliable
electricity supplies, countries maintain spare power stations while hospitals have backup generators. Look-
ing internationally, the European Union, China, and Russia have developed their own independent satellite
navigation systems to parallel the original US global positioning system, even though they are able to use
the American system.
While there would be opposition to SRM use under any deployment scenario, we believe that even where
people or states were initially opposed to the deployment of SRM, this would be unlikely to carry over to
support for sudden termination in the majority of cases. To argue by analogy we believe that few people
who opposed the construction of a nuclear power plant near their homes would want it to be shut down
by removing the cooling rods and allowing a meltdown. Instead, we contend opponents would lambast
the original decision to build the plant while agitating for it to be decommissioned as quickly and safely as
Earth’s Future 10.1002/2017EF000735
possible. Similarly, we believe that regardless of initial antipathy to the use of SRM, most in favour of ending
deployment would prefer to see it phased out carefully rather than terminated suddenly.
Note that we are not arguing that actors must behave with perfect rationality in order to guard against
termination. It is more that they must just avoid wanton irrationality. We think that it would be absurdly
reckless for capable nations to ignore the risk of termination shock and to fail to make simple security and
capacity investments to avoid it. However, there are instances where sufficient precautions are not taken
to avoid damages from failing systems (such as the engineering failures that led to the flooding of New
Orleans caused by Hurricane Katrina (Van Heerden, 2007), or that the Fukushima nuclear plant might be
compromised in the event of a tsunami (Synolakis & Kâno ˘
glu, 2015). It would be useful to research where
and how societies protect against damage from interruption to important infrastructure—and crucially
where they do not.
4.2.1. Would Discovered Damages Overwhelm the Motivation to Maintain SRM?
Some have argued that the motivation to maintain SRM could be trumped by states aggrieved that they
were suffering damages from SRM use (Burns quoted in Kahn, 2018; Robock, 2018). What if a nuclear-armed
nation decided that it wanted SRM deployment stopped, the argument goes, and seemed prepared to act
to get its wish? What if countries experienced damaging droughts and demanded that geoengineering
be terminate immediately? While obviously possible, we argue that there would be strong factors working
against this, meaning it might be much less likely than it superficially appears.
First, it would have to be the case that the discovered impacts of SRM were so damaging that they were
perceived to be more dangerous than termination shock, and that they were not discovered during any
research or ramp-up period. We believe that ceteris paribus these facts work against each other, and that
environmental impacts more damaging than termination shock would likely be discoverable when SRM
were being tested and ramped up. This is a strong reason for any SRM to be studied in depth before being
considered for use, then, if ever deployed, to be ramped up slowly and accompanied by careful monitoring
and evaluation.
Second, termination due to discovered damages would be like any other form of elective termination
(Section 3.2). Those pushing for an end to SRM would need to have the power to force the entire planet
to suffer termination shock in spite of the projected global damages. Therefore, where the capability to
deploy SRM is distributed around the world, SRM cannot be terminated unilaterally— a key finding of this
4.2.2. Implications for Justice and Lock In
It must be noted that the fact that SRM could not be terminated unilaterally is a double-edged sword. While
we believe it means that termination shock is much less likely, it might have significant justice implications.
If a minority of people or regions (even conceivably a majority) were to suffer serious adverse effects of SRM,
they might not be able to force an end to it. This effect could be particularly pronounced if those suffering
were poor and vulnerable, without power or a strong political voice.
In turn, this could have implications for “lock-in.” It has been noted elsewhere that the threat of termination
shock could end up locking humanity into continued use of SRM, even if damaging impacts are discovered
(Burns, 2011; Ott, 2012). Our analysis adds some weight to this idea, as once SRM were being used to the
point where termination shock became possible, there would be an increased incentive to maintain it, and
it could prove very difficult to stop deployment unless the most powerful states agreed.
4.3. The Capacity to Avoid Termination Shock
Where societal systems are affordable, distributed, and flexible, and where interruptions take considerable
time to result in harms, they should be quite robust and resilient against disruptions caused by external
shocks. We think that this is evident from systems like global food production or electricity generation.
Prolonged termination of either would cause significant damage, but there is little societal worry about
this, in part because it is reasonable to surmise that persistent termination would not happen in all but the
most calamitous of disasters. Our analysis in Section 3.1 leads us to believe that it would be relatively
simple to build an SRM system that fit these criteria:
Geographically distributed, with deployment from many sites.
Earth’s Future 10.1002/2017EF000735
Affordable enough for multiple actors to maintain independent systems or backup deployment
Slow to lead to damages following disruption, with months between termination and the onset of the
effects of termination shock.
Therefore, if a few nations separately or collectively invested in excess deployment capacity, and the capa-
bility to redeploy SRM were distributed, then the system should be resilient in the face of anything but a
sustained global calamity, such as a global nuclear war or an unprecedented pandemic.
However, major global catastrophes may occur, which could conceivably cripple the global capacity to
maintain SRM. Some have questioned how much additional suffering would be caused by SRM termina-
tion shock in the wake of a global nuclear war, for instance (Barrett, 2014). However, Baum et al. (2013)
explore this possibility in compelling depth, and argues that a termination shock could compound the risks
of an initial catastrophe, conceivably turning a disaster into an extinction-level event if a rapidly warming
climate made it even harder for humans to survive. While noting the potential for termination shock to
cause a “double catastrophe,” we suggest that SRM is not unique in this regard. Global society is reliant on a
number of other advanced sociotechnical systems (such as farming or healthcare) whose disruption would
compound human suffering following a global calamity. Future research on such “double catastrophes” is
warranted in the context of ongoing research into existential risks (Bostrom, 2013). It is important to under-
stand the potential impacts of such high-impact low-probability events, and consider how the risks might
be managed, or whether they could be great enough to preclude SRM use beyond a certain magnitude.
5. Conclusion
This study has made a first attempt to rigorously analyze the concept of SRM termination shock. It has char-
acterized the physical dimensions and has explored the potential pathways by which a range of drivers
could lead to termination, as well as the potential policy responses that might reduce its risk. We havefound,
as others have before, that starting SRM does not mean that it cannot be stopped, as it should be possible to
exert a few tenths of a degree of geoengineered cooling before reaching a point where termination shock
even became a possibility. We believe that beyond that point, there would a strong motivation to maintain
an SRM system, and we note that if the assumptions in this paper prove reasonable, it should be feasible
to build an SRM system that would be robust and resilient against all but the most destructive of global
calamities, and therefore secure against termination shock, except in extremis.
Protecting SRM infrastructure from attack should not be any harder than protecting existing critical infras-
tructure such as power plants or airports, making it hard for an act either of God or man to destroy enough
of the infrastructure to severely disrupt deployment. There would be a buffer period of (at least) several
months after SRM were terminated before global temperatures started to rise dangerously, giving time for
redeployment. Furthermore, if the technology and knowledge required to implement SRM were widely
available, numerous states could maintain SRM cooling if the original deployer proved unable or unwilling
to continue. This means that a distributed deployment system would be resilient against local or regional
catastrophes, be they economic, political, or physical.We have outlined how vast a global catastrophe would
need to be before it forced the termination of SRM on economic grounds. While some have argued that
political pressure could easily force termination shock, we disagree with this simplistic characterization.
In a world where multiple parties were capable of deployment, SRM could not be terminated unilaterally.
Those wishing to stop SRM cooling would need to be able to force their preference— and the damages of
termination shock—on all other actors around the world.
We therefore conclude that termination shock is much less likely than previous work seems to assume,
because we think we have demonstrated that it should be much easier to avoid than has been previously
been recognized. As such, this paper challenges the common and unhelpful framing in which evaluation
of termination shock had become mired, where only the most dramatic termination scenarios are modeled
and discussed, and where the risks of termination shock are evaluated by considering only the magnitude
of impacts without exploration of their likelihood. We believe that this warrants a change to the way in
which termination shock scenarios are both modeled and discussed in popular and academic literature. It
is not possible to reach useful policy conclusions based on analysis of the worst-case scenarios, such as those
modeled in the G2 and G3 scenarios from the Geoengineering Model Intercomparison Project. Similarly, it is
Earth’s Future 10.1002/2017EF000735
not justifiable to draw insights about the risk of termination shock by reporting the magnitude of the worst
possible impacts, while leaving aside consideration of the likelihood of events that could cause or prevent
termination. Risk analysis must address both likelihood and impacts, and a range of more sophisticated
SRM deployment and termination scenarios is needed to build up an evidence base that allows for more
informed policy messages.
This paper has found no reasons to dismiss the threat of termination shock, however, and if SRM is ever
developed to the point where deployment appears a realistic possibility, management of the risk of termi-
nation shock should be a central concern. We have outlined ways that an SRM system could be made robust
and resilient, but this does not mean that a robust and resilient system would be implemented in real life.
Also, our analysis is founded on a set of assumptions that we believe reflect the current understanding of
the costs and characteristics of an SRM system, but different assumptions might produce different conclu-
sions. For instance, if SRM proves much costlier than currently projected, or much more difficult to realize
technically, then fewer nations would be able tobuild SRM systems and the capacity to maintain SRM would
be less distributed, and therefore less resilient against external threats.
Our final conclusion is the most obvious and important. The best way to avoid termination shock would
be to avoid a situation where a large amount of SRM would be needed to reduce committed climate risks.
Strong action on mitigation would reduce the amount of SRM necessary to maintain a stable global temper-
ature. The development of safe and scalable carbon dioxide removal techniques could reduce the cooling
needed from SRM after deployment, and strong adaptation investment would reduce the suffering from
the residual climate impacts to which Earth is already committed.
Appell, D. (2013). Strange bedfellows? Climate change denial and support for geoengineering. Retrieved from http://www climate-change- denial-and- support-for-geoengineering/
Barrett, S. (2014). Solar geoengineering’s brave new world: Thoughts on the governance of an unprecedented technology. Review of
Environmental Economics and Policy,8(2), 249– 269.
Baum, S. D., Maher, T. M., & Haqq-Misra, J. (2013). Double catastrophe: Intermittent stratospheric geoengineering induced by societal
collapse. Environment Systems & Decisions,33(1), 168– 180. y
Bellamy, R., & Lezaun, J. (2017). Crafting a public for geoengineering. Public Understanding of Science,26(4), 402– 417.
Berdahl, M., Robock, A., Ji, D., Moore, J. C., Jones, A., Kravitz, B., & Watanabe, S. (2014).Arc tic cryosphere responsein the G eoengineering
Model Intercomparison Project G3 and G4 scenarios. Journal of Geophysical Research: Atmospheres.
Bostrom, N. (2013). Existential risk prevention as global priority. Global Policy,4(1), 15– 31.
Boucher, O., Randall, D., Artaxo, P., Bretherton, C., Feingold, G., Forster, P., et al. (2013). Clouds and aerosols. In T. F. Stocker, D. Qin, G.-K.
Plattner, M. Tignor, S. K. Allen, J. Boschung, et al. (Eds.), Climate change 2013: The physical science basis. Contributionof Working Group I
to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change (pp. 571– 657). Cambridge, England; New York, NY:
Cambridge University Press.
Brovkin, V., Petoukhov, V., Claussen, M., Bauer, E., Archer, D., & Jaeger, C. (2009). Geoengineering climate by stratospheric sulfur injections:
Earth system vulnerability to technological failure. Climatic Change,92, 243– 259. 1
Burns, W. C. (2011). Climate geoengineering: Solar radiation management and its implications for intergenerational equity. Stanford
Journal of Law, Science & Policy,4, 39– 55.
Caldeira, K., & Myhrvold, N. P. (2013). Projections of the pace of warming following an abrupt increase in atmospheric carbon dioxide
concentration. Environmental Research Letters,8(3), 034039.
Cao, L., Bala, G., & Caldeira, K. (2012). Climate response to changes in atmospheric carbon dioxide and solar irradiance on the time scale of
days to weeks. Environmental Research Letters,7(3), 034015.
Clark, P. U., Shakun, J. D., Marcott, S. A., Mix, A. C., Eby, M., Kulp, S., et al. (2016). Consequences of twenty-first-century policy for
multi-millennial climate and sea-level change. Nature Climate Change,6(4), 360– 369.
Crutzen, P. J. (2006). Albedo enhancement by stratospheric sulfur injections: A contribution to resolve a policy dilemma? Climatic Change,
77(3), 211 –220. 9101-y
Dean, C. (2010, June 28). When the day after tomorrow has come. New York Times,p.D2.
Goes, M., Tuana, N., & Keller, K. (2011). The economics (or lack thereof) of aerosol geoengineering. Climatic Change,109, 719 –744. https:// 9961-z
Hamilton, C. (2013). Earthmasters: The dawn of the age of climate engineering. New Haven, CT: Yale University Press.
IMF. (2014). World Economic Outlook. Author: Retrieved from
Irvine, P.J., Sriver, R. L., & Keller, K. (2012). Tension between reducing sea-level rise and global warming through solar-radiation
management. Nature Climate Change,2, 97– 100. Retrieved from
Jones, A., Haywood, J. M., Alterskjær, K., Boucher, O., Cole, J. N. S., Curry, C. L., et al. (2013). The impact of abrupt suspension of solar
radiation management (termination effect) in experiment G2 of the Geoengineering Model Intercomparison Project (GeoMIP). Journal
of Geophysical Research: Atmospheres,118(17), 9743– 9752.
Kahn, B. (2018). Geoengineering holds promise; solutions not ready. Retrieved from
solutions-not- ready-18649
The authors would like to acknowl-
edge the helpful comments and discus-
sions with David Keith, Jesse Reynolds,
Sebastian Eastham, John Shepherd,
Andreas Oschlies, Oliver Morton, Klaus
Keller, and their former colleagues at
the Institute for Advanced Sustainabil-
ity Studies, Potsdam, Germany. No new
data or model outputs were generated
as part of this study. This paper was
written while Andy Parker was at the
Institute for Advanced Sustainability
Studies, Potsdam, Germany. He is now
an Honorary Senior Research Fellow in
the School of Earth Sciences at Bristol
University and the project director of
the SRM Governance Initiative.
Earth’s Future 10.1002/2017EF000735
Klein, N. (2014). This changes everything: Capitalism vs. the climate. New York, NY: Simon & Schuster.
Knutti, R., & Sedlacek, J. (2013). Robustness and uncertainties in the new CMIP5 climate model projections. Nature Climate Change,3(4),
369– 373.
Kosugi, T. (2013). Fail-safe solar radiation management geoengineering. Mitigation and Adaptation Strategies for Global Change,18, 1– 26. 9414-2
Kravitz, B., Robock, A., Boucher, O., Schmidt, H., Taylor, K. E., Stenchikov, G., & Schulz, M. (2011). The Geoengineering Model
Intercomparison Project (GeoMIP). Atmospheric Science Letters,12(2), 162 –167.
Kruger, T. (2015). Dimensions of geoengineering: An analysis of the royal society’s ‘blob’diagram (Clim. Geoeng. Governance Working Paper
Series No 26).
Latham, J. (1990). Control of global warming. Nature,347(6291), 339– 340.
Llanillo, P., Jones, P. D., & von Glasow, R. (2010). Theinfluence of stratospheric sulphate aerosol deployment on the surface air
temperature and the risk of an abrupt global warming. Atmosphere,1(1), 62– 84.
MacMartin, D. G., Caldeira, K., & Keith, D. W. (2014). Solar geoengineering to limit the rate of temperature change. Philosophical
Transactions of the RoyalS ociety A: Mathematical,Physical and Engineering Sciences,372(2031).
Matthews, H. D., & Caldeira, K. (2007). Transient climate-carbon simulations of planetary geoengineering. Proceedings of the National
Academy of Sciences of the United States of America,104(24), 9949– 9954.
McClellan, J., Keith, D. W., & Apt, J. (2012). Cost analysis of stratospheric albedo modification delivery systems. Environmental Research
Letters,7(3), 34019.
McCormack, C. G., Born, W., Irvine, P. J., Achterberg, E. P., Amano, T., Ardron, J., et al. (2016). Key impacts of climate engineering on
biodiversity and ecosystems, with priorities for future research. Journal of IntegrativeEnvironmental Sciences, 1 –26.
McCusker, K. E., Armour, K. C., Bitz, C. M., & Battisti, D. S. (2014). Rapid and extensive warming following cessation of solar radiation
management. Environmental Research Letters,9(2), 24005.
McCusker, K. E., Battisti, D. S., & Bitz, C. M. (2012). The climate response to stratospheric sulfate injections and implications for addressing
climate emergencies. Journal of Climate,25(9), 3096– 3116. 00183.1
Mitchell, D. L., & Finnegan, W. (2009). Modification of cirrus clouds to reduce global warming. Environmental Research Letters,4(4). https://
Morice, C. P., Kennedy, J. J., Rayner, N. A., & Jones, P. D. (2012). Quantifying uncertainties in global and regional temperature change using
an ensemble of observational estimates: The HadCRUT4 data set. Journal of Geophysical Research: Atmospheres,117(D8), D08101.
Moriyama, R., Sugiyama, M., Kurosawa, A., Masuda, K., Tsuzuki, K., & Ishimoto, Y. (2016). The cost of stratospheric climate engineering
revisited. Mitigation and Adaptation Strategies for Global Change, 1 –22.
Olson, R. L. (2011). Geoengineering for decision makers. Washington, DC: Science and Technology Innovation Program, Woodrow Wilson
International Center for Scholars.
Ott, K. (2012). Might solar radiation management constitute a dilemma? In C. Preston (Ed.), Engineering the climate. The ethics of solar
radiation management (pp. 33– 42). Lanham, MD: Rowman and Littlefield. Retrieved from
Pidgeon, N., Parkhill, K., Corner, A., & Vaughan, N. (2013). Deliberating stratospheric aerosols for climate geoengineering and the SPICE
project. Nature Climate Change,3(5), 451– 457.
Pierrehumbert, R. T. (2015, February). Climate hacking is barking mad. Slate.
Plumer, B. (2014). One problem with geoengineering: Once you start, you can’t really stop. Retrieved from https://www.washingtonpost
.com/news/wonk/wp/2014/01/02/one-problem- with-geoengineering-once -you-start-you-cant- ever-stop/
Rasch, P. J., Tilmes, S., Turco, R. P., Robock, A., Oman, L., Chen, C.-C., et al. (2008). An overview of geoengineering of climate using
stratospheric sulphate aerosols. Philosophical Transactions. Series A, Mathematical, Physical, and Engineering Sciences,366(1882),
4007– 4037.
Rayner, S., Heyward, C., Kruger, T., Pidgeon, N., Redgwell, C., & Savulescu, J. (2013). The Oxford principles. Climatic Change,121, 499– 512. 0675-2
Robock, A. (2018). Climate engineering, once started, would have severe impacts if stopped. Retrieved from
pub_releases/2018-01/ru- ceo011618.php
Robock, A., Marquardt, A., Kravitz, B., & Stenchikov, G. (2009). Benefits, risks, and costs of stratospheric geoengineering. Geophysical
Research Letters,36(L19703), 9.
Robock, A., Oman, L., & Stenchikov, G. L. (2008). Regional climate responses to geoengineering with tropical and Arctic SO2injections.
Journal of Geophysical Research: Atmospheres,113(D16), D16101.
Schleussner, C.-F., Lissner, T. K., Fischer, E. M., Wohland, J., Perrette, M., Golly, A., et al. (2016). Differential climate impacts for
policy-relevant limits to global warming: The case of 1.5Cand2
C. Earth System Dynamics,7(2), 327 –351.
esd-7- 327-2016
Specter, M. (2012, May). The climate fixers. The New Yorker, pp. 96–103.
Stenchikov, G. L., Kirchner, I., Robock, A., Graf, H.-F., Antuña, J. C., Grainger, R. G., et al. (1998). Radiative forcing from the 1991 Mount
Pinatubo volcanic eruption. Journal of Geophysical Research: Atmospheres,103(D12), 13837– 13857.
Synolakis, C., & Kâno ˘
glu, U. (2015). The Fukushima accident was preventable. Philosophical Transactions of the Royal Society A,373(2053).
The Maddison-Project. (2013). The Maddison Project Database, 2013 version. Retrieved from
Trisos, C. H., Amatulli, G., Gurevitch, J., Robock, A., Xia, L., & Zambri, B. (2018). Potentially dangerous consequences for biodiversity of solar
geoengineering implementation and termination. Nature Ecology and Evolution,1. 0
Van Heerden, I. L. (2007). The failure of the New Orleans levee system following Hurricane Katrina and the pathway forward. Public
Administration Review,67, 24 –35.
Vaughan, N. E., & Lenton, T. M. (2012). Interactions between reducing CO2emissions, CO2removal and solar radiation management.
Philosophical Transactions of the RoyalSociet y A: Mathematical,Physical and Engineering Sciences,370, 4343 –4364.
Zürn, M., & Schäfer, S. (2013). The paradox of climate engineering. Global Policy,4(3), 266– 277.
... Ultimately, the level of emission mitigation and CDR deployed will dictate how much SRM may be strategically desirable (Shepherd, 2009). However, SRM applications also have their limits and may not be used unlimited, for example when it comes to increasing impacts with injection amount (e.g., and increasing risks with termination (Parker & Irvine, 2018). Nevertheless, the near-term upper bound costs projection for moderate injection amounts has been suggested as between $24.9B and $68.5B (USD) over the first fifteen years of deployment, with higher-end estimates contingent on the extent of deployment (Smith & Wagner, 2018). ...
... Abrupt termination of SRM is dangerous, as the rate of temperature rise would be much faster than from current CO2 emissions, likely causing massive ecological damage as the biosphere would not adapt to such rapid changes Trisos et al., 2018). Accordingly, integrating SRM into a climate policy portfolio would require orderly start and exit strategies (ramping aerosol loading up and down), which do not expose the climate to an otherwise avoidable risk of sudden cooling or warming (Parker & Irvine, 2018). These constraints for integrating SRM into a broader climate policy portfolio make scenario modeling helpful in examining possible strategies for combining SRM with mitigation and CDR (e.g., Global management scenarios). ...
... In a real-world scenario -where termination shock is known to have adverse effects, and the technology to reverse termination is available -it seems implausible that no attempts by other parties to restart the SRM program would be made (Parker & Irvine, 2018;Halstead 2018). However, it is feasible that unexpected events can temporarily disrupt future SRM applications. ...
Full-text available
Solar geoengineering, also known as Solar Radiation Modification (SRM), has been proposed to alter Earth’s radiative balance to reduce the effects of anthropogenic climate change. SRM has been identified as a research priority, as it has been shown to effectively reduce surface temperatures, while substantial uncertainties remain around side effects and impacts. Global modeling studies of SRM have often relied on idealized scenarios to understand the physical processes of interventions and their widespread impacts. These extreme or idealized scenarios are not directly policy-relevant and are often physically implausible (such as imposing global solar reduction to counter the warming of an instantaneous quadrupling of CO2). The climatic and ecological impacts of politically relevant and potentially plausible SRM approaches have rarely been modeled and assessed. Nevertheless, commentators and policymakers often falsely assume that idealized or extreme scenarios are proposed solutions to climate change. This paper proposes 18 scenarios that appear to be broadly plausible from political and Earth System perspectives and encompass futures that could be both warnings or perhaps desirable. We place these scenarios into four groups following broader strategic contexts: (1) Global Management; (2) Regional Emergencies; (3) Coordinated Regional Interventions; and (4) Reactive Global Interventions. For each scenario, relevant model experiments are proposed. Some may be performed with existing setups of global climate models, while others require further specification. Developing and performing these model experiments – and assessing likely resulting impacts on society and ecosystems – would be essential to inform public debate and policymakers on the real-world issues surrounding SRM.
... A large body of research has evaluated social and political risks associated with SAI without GHG mitigation. These include geopolitical risks like weaponization of the technology and use for terrorism, "moral hazard" of overreliance on SRM to mitigate harm resulting from climate change, inequitable distribution of climate risk, and risks due to faulty implementation including "termination shock" if the use of SAI ceases abruptly (McCusker et al., 2014;Faran and Olsson, 2018;Flegal and Gupta, 2018;Parker and Irvine, 2018;Grieger et al., 2019). A smaller body of literature focuses on possible implications for human health and related ecological impacts (Irvine et al., 2017;Zarnetske et al., 2021). ...
... The pathways connecting SAI and biodiversity are highly dependent on impacts to global temperature, precipitation, air quality, vegetative prevalence, and soil productivity. Current literature suggests that many risks to biodiversity are linked to termination shock-a fast, drastic rise in global temperature resulting from the sudden stoppage of SAI that occurs as global temperatures quickly rebound to what would have occurred in the absence of SAI (Parker and Irvine, 2018;Trisos et al., 2018a). Termination shock is not a feature of SAI deployment but rather a potential risk that may occur if once deployed SAI is suddenly stopped. ...
Full-text available
Solar radiation management (SRM) is a climate engineering strategy to reduce temperature increases due to global climate change. The most well-researched SRM methodology is stratospheric aerosol injection (SAI), which involves increasing the concentration of aerosol particles in the stratosphere to reduce the amount of solar radiation reaching Earth’s surface. The most considered and heavily researched aerosol for SAI is sulfate. SAI has been extensively modeled using various climate scenarios and investigated using data from previous volcanic eruptions, which provide an analog of the climate effects of SAI. Prior research has determined that SAI will not only decrease global temperatures but is likely to have direct impacts on ecosystem and public health. This review seeks to investigate the various ways by which SAI may impact global public health outcomes related to hydrologic cycling, atmospheric chemical cycling, frequency of natural disasters, food system disruptions, and ecological health through the pathways of water, air, soil, and biota. SAI has the potential to decrease negative health outcomes associated with rising temperatures but may have a myriad of impacts on global environmental systems. Anthropogenically altering the global climate, through both the release of greenhouse gases or through climatic engineering, has unknown consequences, many of which will likely impact global health and quality of life. A more holistic approach is necessary to understand the relative benefits and harms in using SAI as compared to the implication of global climate change.
... Solar geoengineering generally comes with a host of potential side effects and governance challenges. Moreover, unlike CO 2 removal, it cannot simply reverse the effects of climate change (average temperatures may return to pre-industrial levels, but location-specific climates still change), and also comes with the risk of termination shock (fast, catastrophic warming if humanity undertakes solar geoengineering but stops suddenly) [615]. Because of these and other issues, it is not within the scope of this article to evaluate or recommend any particular technique. ...
Full-text available
Climate change is one of the greatest challenges facing humanity, and we, as machine learning (ML) experts, may wonder how we can help. Here we describe how ML can be a powerful tool in reducing greenhouse gas emissions and helping society adapt to a changing climate. From smart grids to disaster management, we identify high impact problems where existing gaps can be filled by ML, in collaboration with other fields. Our recommendations encompass exciting research questions as well as promising business opportunities. We call on the ML community to join the global effort against climate change.
... Also, G1 is easy to perform in all climate models, providing a low barrier to participation in GeoMIP, which is important for community-building and developing from SO 2 oxidation to heterogeneous chemistry processes crucial for ozone chemistry was also highlighted. The termination of the implementation in 2070 has also produced interesting research over the impacts of a potential "termination effect," were geoengineering to be stopped abruptly (Robock et al., 2008;Parker and Irvine, 2018;Trisos et al., 2018). ...
Full-text available
The Geoengineering Model Intercomparison Project (GeoMIP) is a coordinating framework, started in 2010, that includes a series of standardized climate model experiments aimed at understanding the physical processes and projected impacts of solar geoengineering. Numerous experiments have been conducted, and numerous more have been proposed as "testbed'' experiments, spanning a variety of geoengineering techniques aimed at modifying the planetary radiation budget: stratospheric aerosol injection, marine cloud brightening, surface albedo modification, cirrus cloud thinning and sunshade mirrors. To date, more than one hundred studies have been published that used results from GeoMIP simulations. Here we provide a critical assessment of GeoMIP and its experiments. We discuss its successes and missed opportunities, for instance in terms of which experiments elicited more interest from the scientific community and which didn't, and the potential reasons why that happened. We also discuss the knowledge that GeoMIP has contributed to the field of geoengineering research and climate science as a whole: what have we learned in terms of inter-model differences, robustness of the projected outcomes for specific geoengineering methods and future areas of models' development that would be necessary in the future. We also offer multiple examples of cases where GeoMIP experiments were fundamental for international assessments of climate change. Finally, we provide a series of recommendations, regarding both future experiments and more general activities, with the goal of continuously deepening our understanding of the effects of potential geoengineering approaches, as well as reducing uncertainties in climate outcomes, important for assessing wider impacts on societies and ecosystems. In doing so, we refine the purpose of GeoMIP and outline a series of criteria whereby GeoMIP can best serve its participants, stakeholders, and the broader science community.
... Hence, local changes in climate-such as continued warming or the occurrence of extreme events-may cause climate interventions such as SAI to be perceived as a failure. Given the potential for SAI to abruptly cease and the likelihood of rapid climate change following such cessation (e.g., 19,43), the perception of failure carries particular risks. ...
Full-text available
As anthropogenic activities warm the Earth, the fundamental solution of reducing greenhouse gas emissions remains elusive. Given this mitigation gap, global warming may lead to intolerable climate changes as adaptive capacity is exceeded. Thus, there is emerging interest in solar radiation modification, which is the process of deliberately increasing Earth’s albedo to cool the planet. Stratospheric aerosol injection (SAI)—the theoretical deployment of particles in the stratosphere to enhance reflection of incoming solar radiation—is one strategy to slow, pause, or reverse global warming. If SAI is ever pursued, it will likely be for a specific aim, such as affording time to implement mitigation strategies, lessening extremes, or reducing the odds of reaching a biogeophysical tipping point. Using an ensemble climate model experiment that simulates the deployment of SAI in the context of an intermediate greenhouse gas trajectory, we quantified the probability that internal climate variability masks the effectiveness of SAI deployment on regional temperatures. We found that while global temperature was stabilized, substantial land areas continued to experience warming. For example, in the SAI scenario we explored, up to 55% of the global population experienced rising temperatures over the decade following SAI deployment and large areas exhibited high probability of extremely hot years. These conditions could cause SAI to be perceived as a failure. Countries with the largest economies experienced some of the largest probabilities of this perceived failure. The potential for perceived failure could therefore have major implications for policy decisions in the years immediately following SAI deployment.
Full-text available
Esta traducción del “Capítulo 14: Cooperación internacional de la contribución del Grupo de Trabajo III al Sexto Informe de Evaluación del IPCC” no es una traducción oficial del IPCC. Ha sido realizada por profesores y alumnos de la Maestría en Derecho y Economía del Cambio Climático de FLACSO Argentina con el objetivo de reflejar de la manera más precisa el lenguaje utilizado en el texto original. Citar como: Patt, A., L. Rajamani, P. Bhandari, A. Ivanova Boncheva, A. Caparrós, K. Djemouai, I. Kubota, J. Peel, A.P. Sari, D.F. Sprinz, J. Wettestad, 2022: Cooperación internacional. En IPCC, 2022: Cambio Climático 2022: Mitigación del Cambio Climático. Contribución del Grupo de Trabajo III al Sexto Informe de Evaluación del Grupo Intergubernamental de Expertos sobre el Cambio Climático [P.R. Shukla, J. Skea, R. Slade, A. Al Khourdajie, R. van Diemen, D. McCollum, M. Pathak, S. Some, P. Vyas, R. Fradera, M. Belkacemi, A. Hasija, G. Lisboa, S. Luz, J. Malley, (eds.)]. Cambridge University Press, Cambridge, Reino Unido y Nueva York, NY, Estados Unidos. (Traducido por FLACSO Argentina) (2022).
Full-text available
The teleconnection between the Quasi-Biennial Oscillation (QBO) and the Arctic stratospheric polar vortex, or the Holton-Tan (HT) relationship may change in a warmer climate or one with stratospheric aerosol intervention (SAI) as compared to present day climate (PDC). Our results from an Earth system model indicate that, under both global warming (based on RCP8.5 emission scenario) and SAI scenarios, the HT relationship weakens, although it is closer to PDC under SAI than under the RCP8.5 scenario. Such weakening of the HT relationship is more pronounced in early winter (Nov-Dec) 5 compared to the mid-late winter period (Jan-Feb). While the high-latitude responses of temperature to the QBO anomalies are statistically significant under PDC, the responses are not statistically significant in the RCP8.5 and SAI scenarios. While the weakening of the HT relationship under RCP8.5 scenario is likely due to the weaker QBO wind amplitudes at the equator, another physical mechanism must be responsible for the weaker HT relationship under SAI scenario, since the amplitude of the QBO wind is comparable to the PDC. The strength of the polar vortex does not change under the RCP8.5 scenario compared to 10 PDC, but it becomes stronger under SAI; we attribute the weakening of the HT relationship under SAI to such stronger polar vortex. In general, the changes in the HT relationship cannot be solely explained by changes to the critical line; the changes in the residual circulation (particularly due to the gravity wave contributions) are important too in explaining the changes in the HT relationship under RCP8.5 and SAI scenarios.
As the effects of climate change continue to worsen, nations will have the opportunity to develop and deploy climate manipulation techniques known as geoengineering to forestall the worst effects. Indeed, some have argued that the nations of the world cannot meet their Paris Agreement goals without them. However, these technologies can be global in their effects, ecologically uncertain, and potentially prejudicial to non-deploying nations. Canada is well suited to lead the formation of an anticipatory governance regime due to its technological knowledge, the proximity of the Arctic as a potential testing ground, and its role as an internationally respected middle power. By stepping forward to lead the effort, Canada can ensure its own security and environmental interests as well as the stability of the rules-based international order.
Full-text available
The risk of global catastrophe from natural sources may be significantly larger than previous analyses have found. In the study of global catastrophic risk (GCR), one line of thinking posits that deep human history renders natural GCRs insignificant. Essentially, the fact that natural hazards did not cause human extinction at any previous time makes it unlikely that they would do so now. This paper finds flaws in this argument, refines the theory of natural GCR, analyzes the space of natural GCRs, and presents implications for decision-making and research. The paper analyzes natural climate change, natural pandemics, near-Earth objects (asteroids, comets, and meteors), space weather (coronal mass ejections, solar flares, and solar particle events), stellar explosions (gamma-ray bursts and supernovae), and volcanic eruptions. Almost all natural GCR scenarios involve important interactions between the natural hazard and human civilization. Several natural GCR scenarios may have high ongoing probability. Deep human history provides little information about the resilience of modern global civilization to natural global catastrophes. The natural GCRs should not be dismissed on grounds of deep human history. Work on natural GCRs should account for their important human dimensions. A case can even be made for abandoning the distinction between natural and anthropogenic GCR.
This chapter describes the principles of two prominent proposals to deliberately modify Earth's climate using aerosol: injection of particles into the stratosphere to mimic the effects of volcanic eruptions and injection of sea spray particles into shallow marine clouds to increase their reflection of solar radiation. The chapter describes the physical mechanisms and the analogs in the natural or human-perturbed atmosphere that can be used to test our understanding. It then describes the climatic impacts of climate engineering based on model simulations, including the inadvertent effects on regional climate.
Full-text available
Solar geoengineering is receiving increased policy attention as a potential tool to offset climate warming. While climate responses to geoengineering have been studied in detail, the potential biodiversity consequences are largely unknown. To avoid extinction, species must either adapt or move to track shifting climates. Here, we assess the effects of the rapid implementation, continuation and sudden termination of geoengineering on climate velocities-the speeds and directions that species would need to move to track changes in climate. Compared to a moderate climate change scenario (RCP4.5), rapid geoengineering implementation reduces temperature velocities towards zero in terrestrial biodiversity hotspots. In contrast, sudden termination increases both ocean and land temperature velocities to unprecedented speeds (global medians >10 km yr-1) that are more than double the temperature velocities for recent and future climate change in global biodiversity hotspots. Furthermore, as climate velocities more than double in speed, rapid climate fragmentation occurs in biomes such as temperate grasslands and forests where temperature and precipitation velocity vectors diverge spatially by >90°. Rapid geoengineering termination would significantly increase the threats to biodiversity from climate change.
Full-text available
Stratospheric aerosol injection (SAI) has been receiving increasing attention as a possible option for climate engineering. Its direct cost is perceived to be low, which has implications for international governance of this emerging technology. Here, we critically synthesize previous estimates of the underlying parameters and examine the total costs of SAI. It is evident that there have been inconsistencies in some assumptions and the application of overly optimistic parameter values in previous studies, which have led to an overall underestimation of the cost of aircraft-based SAI with sulfate aerosols. The annual cost of SAI to achieve cooling of 2 W/m² could reach US$10 billion with newly designed aircraft, which contrasts with the oft-quoted estimate of “a few billion dollars.” If existing aircraft were used, the cost would be expected to increase further. An SAI operation would be a large-scale engineering undertaking, possibly requiring a fleet of approximately 1,000 aircraft, because of the required high altitude of the injection. Therefore, because of its significance, a more thorough investigation of the engineering aspects of SAI and the associated uncertainties is warranted.
Full-text available
Robust appraisals of climate impacts at different levels of global-mean temperature increase are vital to guide assessments of dangerous anthropogenic interference with the climate system. The 2015 Paris Agreement includes a two-headed temperature goal: "holding the increase in the global average temperature to well below 2 °C above pre-industrial levels and pursuing efforts to limit the temperature increase to 1.5 °C". Despite the prominence of these two temperature limits, a comprehensive overview of the differences in climate impacts at these levels is still missing. Here we provide an assessment of key impacts of climate change at warming levels of 1.5 °C and 2 °C, including extreme weather events, water availability, agricultural yields, sea-level rise and risk of coral reef loss. Our results reveal substantial differences in impacts between a 1.5 °C and 2 °C warming that are highly relevant for the assessment of dangerous anthropogenic interference with the climate system. For heat-related extremes, the additional 0.5 °C increase in global-mean temperature marks the difference between events at the upper limit of present-day natural variability and a new climate regime, particularly in tropical regions. Similarly, this warming difference is likely to be decisive for the future of tropical coral reefs. In a scenario with an end-of-century warming of 2 °C, virtually all tropical coral reefs are projected to be at risk of severe degradation due to temperature-induced bleaching from 2050 onwards. This fraction is reduced to about 90 % in 2050 and projected to decline to 70 % by 2100 for a 1.5 °C scenario. Analyses of precipitation-related impacts reveal distinct regional differences and hot-spots of change emerge. Regional reduction in median water availability for the Mediterranean is found to nearly double from 9 % to 17 % between 1.5 °C and 2 °C, and the projected lengthening of regional dry spells increases from 7 to 11 %. Projections for agricultural yields differ between crop types as well as world regions. While some (in particular high-latitude) regions may benefit, tropical regions like West Africa, South-East Asia, as well as Central and northern South America are projected to face substantial local yield reductions, particularly for wheat and maize. Best estimate sea-level rise projections based on two illustrative scenarios indicate a 50 cm rise by 2100 relative to year 2000-levels for a 2 °C scenario, and about 10 cm lower levels for a 1.5 °C scenario. In a 1.5 °C scenario, the rate of sea-level rise in 2100 would be reduced by about 30 % compared to a 2 °C scenario. Our findings highlight the importance of regional differentiation to assess both future climate risks and different vulnerabilities to incremental increases in global-mean temperature. The article provides a consistent and comprehensive assessment of existing projections and a good basis for future work on refining our understanding of the difference between impacts at 1.5 °C and 2 °C warming.
Full-text available
Climate change has significant implications for biodiversity and ecosystems. With slow progress towards reducing greenhouse gas emissions, climate engineering (or ‘geoengineering’) is receiving increasing attention for its potential to limit anthropogenic climate change and its damaging effects. Proposed techniques, such as ocean fertilization for carbon dioxide removal or stratospheric sulfate injections to reduce incoming solar radiation, would significantly alter atmospheric, terrestrial and marine environments, yet potential side-effects of their implementation for ecosystems and biodiversity have received little attention. A literature review was carried out to identify details of the potential ecological effects of climate engineering techniques. A group of biodiversity and environmental change researchers then employed a modified Delphi expert consultation technique to evaluate this evidence and prioritize the effects based on the relative importance of, and scientific understanding about, their biodiversity and ecosystem consequences. The key issues and knowledge gaps are used to shape a discussion of the biodiversity and ecosystem implications of climate engineering, including novel climatic conditions, alterations to marine systems and substantial terrestrial habitat change. This review highlights several current research priorities in which the climate engineering context is crucial to consider, as well as identifying some novel topics for ecological investigation.
Full-text available
Most of the policy debate surrounding the actions needed to mitigate and adapt to anthropogenic climate change has been framed by observations of the past 150 years as well as climate and sea-level projections for the twenty-first century. The focus on this 250-year window, however, obscures some of the most profound problems associated with climate change. Here, we argue that the twentieth and twenty-first centuries, a period during which the overwhelming majority of human-caused carbon emissions are likely to occur, need to be placed into a long-term context that includes the past 20 millennia, when the last Ice Age ended and human civilization developed, and the next ten millennia, over which time the projected impacts of anthropogenic climate change will grow and persist. This long-term perspective illustrates that policy decisions made in the next few years to decades will have profound impacts on global climate, ecosystems and human societies-not just for this century, but for the next ten millennia and beyond.
Full-text available
Robust appraisals of climate impacts at different levels of global-mean temperature increase are vital to guide assessments of dangerous anthropogenic interference with the climate system. Currently, two such levels are discussed in the context of the international climate negotiations as long-term global temperature goals: a below 2 °C and a 1.5 °C limit in global-mean temperature rise above pre-industrial levels. Despite the prominence of these two temperature limits, a comprehensive assessment of the differences in climate impacts at these levels is still missing. Here we provide an assessment of key impacts of climate change at warming levels of 1.5 °C and 2 °C, including extreme weather events, water availability, agricultural yields, sea-level rise and risk of coral reef loss. Our results reveal substantial differences in impacts between 1.5 °C and 2 °C. For heat-related extremes, the additional 0.5 °C increase in global-mean temperature marks the difference between events at the upper limit of present-day natural variability and a new climate regime, particularly in tropical regions. Similarly, this warming difference is likely to be decisive for the future of tropical coral reefs. In a scenario with an end-of-century warming of 2 °C, virtually all tropical coral reefs are projected to be at risk of severe degradation due to temperature induced bleaching from 2050 onwards. This fraction is reduced to about 90 % in 2050 and projected to decline to 70 % by 2100 for a 1.5 °C scenario. Analyses of precipitation-related impacts reveal distinct regional differences and several hot-spots of change emerge. Regional reduction in median water availability for the Mediterranean is found to nearly double from 9 to 17 % between 1.5 °C and 2 °C, and the projected lengthening of regional dry spells increases from 7 % longer to 11 %. Projections for agricultural yields differ between crop types as well as world regions. While some (in particular high-latitude) regions may benefit, tropical regions like West Africa, South-East Asia, as well as Central and Northern South America are projected to face local yield reductions, particularly for wheat and maize. Best estimate sea-level rise projections based on two illustrative scenarios indicate a 50 cm rise by 2100 relative to year 2000-levels under a 2 °C warming, which is about 10 cm lower for a 1.5 °C scenario. Our findings highlight the importance of regional differentiation to assess future climate risks as well as different vulnerabilities to incremental increases in global-mean temperature. The article provides a consistent and comprehensive assessment of existing projections and a solid foundation for future work on refining our understanding of warming-level dependent climate impacts.
This book goes to the heart of the unfolding reality of the twenty-first century: international efforts to reduce greenhouse gas emissions have all failed, and before the end of the century Earth is projected to be warmer than it has been for 15 million years. The question "can the crisis be avoided?" has been superseded by a more frightening one, "what can be done to prevent the devastation of the living world?" And the disturbing answer, now under wide discussion both within and outside the scientific community, is to seize control of the very climate of the Earth itself. Clive Hamilton begins by exploring the range of technologies now being developed in the field of geoengineering--the intentional, enduring, large-scale manipulation of Earth's climate system. He lays out the arguments for and against climate engineering, and reveals the extent of vested interests linking researchers, venture capitalists, and corporations. He then examines what it means for human beings to be making plans to control the planet's atmosphere, probes the uneasiness we feel with the notion of exercising technological mastery over nature, and challenges the ways we think about ourselves and our place in the natural world.