Climate Engineering – Avoiding Pandora's Box through Research and Governance

Full text

FNI Climate Policy Perspectives 5 May 2012
Climate Engineering
Avoiding Pandora’s Box through Research and
Governance
Matthias Honegger, Axel Michaelowa and
Sonja Butzengeiger-Geyer

FNI Climate Policy Perspectives 5 May 2012
Climate Engineering
Avoiding Pandora’s Box through Research and
Governance
Matthias Honegger
Climate Policy Intern
Perspectives GmbH
honegger@
perspectives.cc
Axel Michaelowa
Senior Founding Partner
Perspectives GmbH
michaelowa@
perspectives.cc
Sonja Butzengeiger-Geyer
Managing Director
Perspectives GmbH
butzengeiger@
perspectives.cc
 The gap between the emissions reductions required by the 2°C target and those actually
undertaken is growing. Thus, climate engineering as an alternative proposition to mitigate
climate change is expected to become increasingly relevant and likely to enter the mainstream
discourse on climate mitigation within a decade.
 The term ‘climate change mitigation’ should be broadened to include all measures that limit
the extent of climatic changes. Carbon dioxide removal is akin to classical biological and
geological sequestration. Solar radiation management could mask – and thus in the broader
sense mitigate – climate change as long as the intervention is continued.
 Given the high perceived attractiveness of solar radiation management due to costs two
orders of magnitude less than those of equivalent emissions reductions, the level of risks
must be established with a high degree of certainty, and accompanying measures need be in
place, before the option can be seriously contemplated.
 In order to prevent potentially catastrophic unilateral deployment or military use, international
governance is required to coordinate research in all disciplines concerned with climate
engineering in the long term and make its results publicly accessible. A Special Report on
Climate Engineering by the IPCC would provide an ideal basis for such international norm
building concerning research, development and deployment.
 In order to avoid capture by interest groups and prevent premature irreversible decisions,
climate engineering governance and monitoring under the UNFCCC could be based partly on
approaches used in nuclear weapons control and terrorism prevention.
The Fridtjof Nansen Institute (FNI) is an independent, non-profit institution engaged in research on international
environmental, energy and resource management politics. Perspectives is an independent service enterprise that works in
consultation with the private sector as well as governments and NGOs in realizing and enhancing instruments in the
international greenhouse gas market. FNI exercises quality control and editing of the papers, but the views expressed are
the sole responsibility of the authors.

F N I C L I M A T E P O L I C Y P E R S P E C T I V E S 5
3
Climate Engineering
Avoiding Pandora’s Box through Research and
Governance
Climate engineering
1
(CE) has been defined as
the deliberate large-scale manipulation of the
planetary environment to counteract anthro-
pogenic climate change. Proposals for modi-
fying weather or climate have been offered for
more than a century. Only after 1965, how-
ever, have these ideas focused on trying to
cool the climate and thus counteract warming
caused by greenhouse gases (GHGs). In 2006,
Nobel Prize Laureate Paul Crutzen published a
discussion paper on stratospheric sulphur
injection. This gave rise to a growing amount
of attention, among the public as well as the
scientific community, to the entire field, which
had long been sidelined. Many of the relatively
recent ideas have entered the emerging de-
bate since 2006; some have received further
attention in the form of related small-scale
scientific experiments (e.g. the effect of iron
on algal growth and concept studies of the
technical feasibility for stratospheric sulfur
particles).
The path taken by the international climate
policy regime is appearing less and less suf-
ficient, given the growing gap between actual
emissions reductions and requirements for
achieving the 2°C target agreed internatio-
nally. As we are currently headed towards an
average temperature increase of 3-4°C, a
climate emergency – a situation of sudden
heavy impacts by extreme events or an
acceleration of temperature increase due to
e.g. methane releases in the Arctic – cannot
be ruled out. Such an event would create
pressure for rapid responses.
Public awareness of climate engineering is still
very limited, and public opinion is likely to be
shaped by the initial framing of the issue in
popular media in the coming years. As the 5th
IPCC Assessment Report to be published in
1
The term geoengineering (or geo-engineering) has been
used in the literature and is still being used for large-scale
engineering projects, only some of which are related to
climate change. The term CE is thus more accurate to
solely describe technologies that intentionally affect
climate and is used here for climate related geo-
engineering.
2014 will take the issue up for the first time,
CE will soon be elevated from a science-
fiction-like curiosity into mainstream climate
policy discourse. It is crucial that this discus-
sion be based on the best information avail-
able. We want to provide some pathways and
signposts for the years to come. This paper
presents the types of CE under discussion,
with tentative cost estimates as well as the
risks and uncertainties attributed to these
options. Subsequently we discuss the argu-
ments for and against research and develop-
ment of CE, showing that well-governed CE
research is necessary and that there are valid
arguments for making CE part of the dis-
course on climate change mitigation. We then
develop a set of fundamental guidelines for
governance of CE research and monitoring to
avoid potentially dangerous developments.
Finally we sketch the next steps for CE and
international cooperation on climate change
mitigation.
Types of climate engineering: Costs
and risks
CE comes in two main forms, with very differ-
rent characteristics: carbon dioxide removal
(CDR), and solar radiation management
(SRM)
2
. CDR aims to reduce the concentration
of CO2 in the atmosphere and is thus closer to
the conventional mitigation approaches as
illustrated in figure 1. The following sub-types
of CDR have been proposed, arranged by the
estimated cost-risk trade-off based on the
available literature.
costs
risks
Direct air capture through artificial means
Chemical weathering of rocks
Biomass energy with carbon sequestration
Increasing ocean alkalinity with lime
Ocean fertilization (Fe, N, P)
2
Shepherd, W.; Cox, P.; Haigh, J.; Keith, D.; Launder, B.;
Mace, G.; Mackerron, G.; Pyle, J.; Rayner, S.; Redgewell, C.;
Watson, A. (2009) Geoengineering the climate: science,
governance and uncertainty, Royal Society, London.

F N I C L I M A T E P O L I C Y P E R S P E C T I V E S 5
4
The options afforestation, reforestation and
soil sequestration have relatively low risks and
costs, but face implementation barriers for
large-scale applications in terms of e.g. land-
use conflicts. Proving the additionality of such
projects presents a major challenge in itself.
As to SRM, it either reduces incoming short-
wave solar radiation or increases outgoing
long-wave thermal radiation
3
.
costs
risks
Increased albedo of land (urban
areas, arable land, desert)
Marine cloud whitening
Stratospheric aerosols
Space reflectors cannot yet be properly asses-
sed due to the highly tentative characteristics
of the technology proposed.
The mitigation costs of both cloud whitening
and stratospheric aerosol seeding are current-
ly estimated to be two orders of magnitude
lower than those of GHG emission reduction,
while land albedo changes are comparable in
cost with emission reduction measures.
Figure 1: Taxonomy of the broadened
climate change mitigation term including
carbon dioxide removal and solar
radiation management. Source: authors.
From the rough cost estimates available and a
first look at risks, the general trade-off be-
tween costs and risks of the different ap-
proaches can be seen – the more costly CDR
options generally have lower uncertainties and
their risks are not as systemic as those of the
seemingly cheap SRM. The nature of the risks
of CDR and SRM technologies are described in
box 1 and 2 respectively.
3
Rickels et al. (2011) use the term ‘Radiation
Management’. Due to the semantic problem of ‘radiation
management’ being linked to radiation related to nuclear
fission, we retain the term SRM for both incoming and
outgoing radiation.
Box 1: Risks of CDR
Carbon Dioxide Removal methods can
be expected to have lower risks than
SRM methods because they return the
climate system closer to its natural state
and their radiative forcing effect is ho-
mogenous. Some risks are attributed to
measures that require massive mining
and transportation. Even greater risks
can be attributed to approaches that
affect the oceanic food-web, as second-
ary effects within oceanic ecosystems
can result in a net increase rather than
net absorption of greenhouse gases and
possibly affect biodiversity adversely.
The reason why the non-surface related SRM
options generate great systemic risk is that
they directly and unequally affect temperature
gradients in the atmosphere, which determine
the behavior of this complex system in ways
that possibly cannot be foreseen. Given our
incomplete understanding of the atmosphere
(e.g. the considerable uncertainties as to the
role of aerosols in the climate system) and its
many feedback mechanisms, applying short-
term cost-benefit calculations better suited for
other technological investments would be
inappropriate and dangerous in the case of
SRM.
Box 2: Risks of SRM
Solar Radiation Management methods
present a high-leverage interference in
the climate system with largely unknown
consequences. Atmospheric and stratos-
pheric SRM might succeed in masking
the average temperature increase, but
temperatures and rainfall will be affec-
ted unevenly creating significant chang-
es in atmospheric circulation, with po-
tentially severe impacts on ecosystems
and agriculture. SRM needs to be sus-
tained through continuous action; if it
stops the climate will abruptly jump to
an ‘unmasked’ state of warming or even
overshoot. The brightening of land sur-
faces presents a lower risk. While miti-
gating climate change, SRM does not re-
duce acidification of oceans due to
further increases of atmospheric CO2
concentrations.

F N I C L I M A T E P O L I C Y P E R S P E C T I V E S 5
5
Given the probabilistic nature of climate
change damages and benefits, the concept of
systemic risk is crucial in policy decisions
concerning mitigation, especially when CE is
included. Non-systemic risk can be quantified
in an insurance policy, but quantifying CE
risks presents a seemingly insurmountable
challenge. Insurance has trouble addressing
large, unknown risks with low return periods
so quantifying the probability and damage of
a governance failure leading to climate col-
lapse might simply be impossible. This issue,
however, applies both worlds: The one with
possibly insufficient emission reductions, as
well as the one, where risky CE measures
additionally play a role in mitigating climate
change. Because climatic developments are
irreversible to a large degree, only long-term
cost-benefit calculations can be adequate. In
view of the systemic risks and irreversibility of
climate changes, rather than aggregating risk
and cost-benefit into one single policy variable
– as previously done in economic analyses of
CE technologies – we propose a different ap-
proach. To balance various mitigation options
we suggest dealing with risk and cost-benefit
separately: Balancing the risks of insufficiently
mitigated climate change against the risks of
deploying specific CE technologies should
come before any aggregated cost-benefit
analysis is attempted – since the latter can to
date only be based on sketchy quantifications
of the risks. If at a later stage research allows
us to develop more sophisticated analyses of
the risks associated with CE approaches, long-
term cost-benefit assessments of different
mitigation options can be based on these risk
quantifications. Significant difficulties in
quantifying risks of governance failures as
well as uncertainties due to the systemic
nature of atmospheric changes are however
likely to remain.
A research moratorium?
In view of the unprecedented risks of key SRM
options, some analysts
4
have queried whether
CE should even be researched. To some
degree, the topic has been effectively taboo
within the scientific community not least due
to its implications on climate policy.
4
The entire map of arguments has been presented by
Rickels, W.; Klepper, G.; Dovern, J.; Betz, G.; Brachatzek,
N.; Cacean, S.; Güssow, K.; Heintzenberg, J.; Hiller, S.;
Hoose, C.; Leisner, T.; Oschlies, A.; Platt, U.; Proelß, A.;
Renn, O.; Schäfer, S.; Zürn M. (2011):. Large-Scale
Intentional Interventions into the Climate System?
Assessing the Climate Engineering Debate, Kiel
The supporters of research argue that CE
could be developed as an insurance against
dangerous climate change, or see it as an
important array of options for improving the
cost-effectiveness of mitigation or even the
silver bullet to avoid the necessity of inter-
national agreement. Even those who focus on
the dangers of CE acknowledge that a mora-
torium could play into the hands of rogue
states and actors who would not hesitate to
use their CE technology. Yet others hope that
research and public attention might ‘unmask’
the flaws of CE, and thus contribute to greater
willingness to reduce emissions.
Critics of research especially fear that SRM will
lead to a ‘sword of Damocles’ poised over our
heads for centuries: An unplanned terminat-
ion of the SRM activities could bring a sudden
jump in temperatures with devastating effects
if no substitute technology could step into the
breach. Moreover, unplanned side effects of
SRM on atmospheric circulation and precipi-
tation could make the cure worse than the
disease, at least for some regions in develop-
ing countries.
Some aspects call for caution, but might not
be seen as a black-or-white issue concerning
CE research. Moral hazard could be triggered,
reducing the willingness to engage in immedi-
ate and strong emissions reductions. Simi-
larly, stepping up CE research might restrain
research on emissions-reduction options and
climate science, due to limited resources and
human capacity. As yet, few CE approaches
seem able to deliver a low-risk, low-cost
solution that could safely replace emissions
reductions; therefore, any decrease in emis-
sions reduction or research efforts should be
avoided. Other concerns center on the risks of
testing CE. Even if the scale of tests were
increased only gradually, some effects might
become apparent only in large-scale tests. The
risks of such tests could be equal to those of
actual deployment. Without coordinated
research, such tests might be carried out
unilaterally and without proper monitoring.
The possibility of military or terrorist use of
CE technologies would add a further threat,
on the scale of nuclear weapons. Arguably the
best policy for controlling and preventing
such secretive engagement is transparency
and public awareness of the issue, leading to
norms for acceptable use of the technologies.
This should help to deter rogue actors from
embarking on clandestine R&D, as coordi-
nated research efforts could stay ahead of
individual development efforts – making these
useless. It would also prevent research from

F N I C L I M A T E P O L I C Y P E R S P E C T I V E S 5
6
taking on a life of its own, as the vested inte-
rests would be closely watched. This should
make it possible to stop development in the
case of new risks becoming apparent. An ass-
essment of how to shield research from parti-
cular interests could be based on studies of
more established technologies, like nuclear
power or fossil fuel industries.
An argument that clearly cuts both ways is not
to impose any irreversible changes on coming
generations. This could mean not to impose
an ongoing CE regime on future generations,
but could just as well be seen as an obligation
to explore any chance of reducing the severity
of climate change for coming generations.
The widening gap between actual emission
reductions and the ones required for the 2°C
limit and the resulting possibility of a dra-
matic increase of extreme events could result
in political pressure for a quick fix. In view of
these developments more and more research-
ers favor researching CE – not least to avoid
being surprised by unilateral deployment. Due
to the described risks emission reduction
efforts need to be kept up and the issues of
CE governance have to be addressed from the
beginning. The mitigation discourse should
therefore be cautiously expanded to include
CE, with a specific focus on adequate framing
of the issue in the media to allow a balanced
opinion to develop in the general public.
Research design
Credible economic cost-benefit estimates as
well as risk assessments require a strong
foundation of research on the physical effects
of different technologies. The CE research
challenge will be – even more so than with
previous climate research – to link efforts
from the human and the environmental side:
atmospheric modeling, atmospheric chem-
istry, oceanography, plant biology and eco-
logy are among the disciplines on the environ-
mental side. Political science, ethics, history,
sociology, psychology, media sciences, agri-
cultural science, forest science, economics,
national security studies, engineering and
more are concerned with the human side.
Results from all such disciplines will be
needed to advance understanding in other
disciplines: this calls for an internationally
coordinated and balanced effort.
Extensive atmospheric modeling of primary
effects caused by specific CE interferences will
be required for a comprehensive assessment
of the spatial and temporal distribution of the
physical effects. Based on such modeling
results the impacts on terrestrial and aquatic
ecosystems should be assessed as well as the
impacts on agriculture, forestry and other
sectors of the economy as well as human
health and infrastructure. From such assess-
ments, economic estimates could quantify the
damages and benefits caused by any CE
scheme in comparison to likely climate
change impacts without the scheme.
Aside from such an effect-based cost-benefit
approach, special focus should be placed on
the range of possible damages – the risks.
From a political science perspective, the pos-
sibility of societal failure in maintaining a SRM
scheme needs to be addressed. Such an un-
planned termination, due to an act of terror,
international conflicts and governance failure
or a major economic breakdown, would cause
the nightmare scenario of abrupt warming by
several degrees. It is crucial to determine how
the termination risk could be reduced, e.g. by
building spare installations and thus introduc-
ing redundancies in the scheme. Concerning
moral hazard the framing of CE in public
media should be closely monitored through a
dedicated media science research project: It is
vital to avoid misrepresentation of SRM as
equal to emission reduction or CDR, as this
could adversely affect efforts aimed at emis-
sions reduction. The independence of public
research and the absence of biases introduced
by specific interest groups should be moni-
tored, including the possibility of halting
research if a technology should represent
unacceptable threats. A preliminary assess-
ment of those risks could be done in a com-
parative way looking at the lobbying power of
more established technologies, such as nuc-
lear power or fossil fuel industries.
Monitoring for CE development and potential
deployment will be necessary at some point.
Such monitoring will require in-depth know-
ledge of the technologies – a further argu-
ment for solid research.
In the end, the various engineering challenges
must be addressed; current estimates of
costs, infrastructure and material require-
ments seem inadequate. Technologies with
termination risk require a special focus on
reliability and long-term effectiveness.
In view of the many challenges regarding CE
research and testing, a panel of experts
5
has
5
US Government Accountability Office (2010): Climate
Change - A Coordinated Strategy Could Focus Federal

F N I C L I M A T E P O L I C Y P E R S P E C T I V E S 5
7
estimated the time required for developing
and evaluating a CE scheme to be at least two
decades. This may even be optimistic, as co-
ordinated international research efforts, es-
pecially in the atmospheric sciences, have
tended to take longer than expected. Major
politically motivated disputes fuelling the
already limited public acceptance of such
technologies might possibly block CE research
– an early instance is the UK engineering
research project SPICE.
6
Governance: Next steps
Climate engineering is not well represented in
existing international norms: Whereas the
1996 London Protocol of the Convention on
the Prevention of Marine Pollution by Dump-
ing of Wastes and Other Matter prohibits CO2
storage in the ocean, no other international
treaty addresses CE directly. However, the
1976 Convention on the Prohibition of Military
or Any Other Hostile Use of Environmental
Modification Techniques (ENMOD) might
provide a starting point for regulation of CE
interventions. In 2010 the Conference of the
Parties to the Convention on Biological Div-
ersity provided the non-binding guidance that
no ‘[…] climate-related geoengineering active-
ties that may affect biodiversity take place,
until there is an adequate scientific basis on
which to justify such activities and appropriate
consideration of the associated risks for the
environment and biodiversity.’ Codes of
conduct have been suggested concerning
research the Oxford Principles of 2010
include the following points:
 CE needs regulation as a public good.
 Public participation is a must in CE
decision-making.
 Research plans should be fully disclosed
and all research results made publicly
available.
 Impacts are to be independently
assessed.
 Robust governance needs to be in place
before CE deployment.
Geoengineering Research and Inform Governance Efforts,
Washington
6
The SPICE project – a collaboration of the Universities of
Bristol, Cambridge, Edinburgh and Oxford as well as the
Marshall Aerospace institute – was intended in September
2011 as a small-scale field test of a hose supported by a
helium balloon. It was meant to offer insights into the
technological requirements for atmospheric distribution
of aerosol particles.. Due to a number of protests the test
was put on hold.
Rather than governance through an outright
ban or continued neglect of the topic,
7
we
recommend adequate decisions under the
UNFCCC or a dedicated international treaty on
CE, as a protocol to the UNFCCC. Further,
developing sound norms requires a solid basis
of information that can be delivered only
through a coordinated research effort, in
accordance with a set of principles like those
mentioned above. Ideally, the IPCC would
write a Special Report on CE as soon as the
bulk of work on the 5th Assessment Report has
been completed in 2013. This would trigger
more academic research, as the taboo within
the scientific community would finally be
broken. In addition a Special Report could
provide a shared terminology and a frame-
work of criteria to assess the various aspects
of CE technologies. Both are necessary in
order to develop adequate norms.
As mentioned, a general ban would be coun-
terproductive or even dangerous, as it would
continue to put off a major share of research
activities, without deterring other entities
from clandestinely engaging in CE activities.
Such entities – be it single nations or even
companies – might be acting in their best
interests. For example a small island state like
Tuvalu might want to save its territorial inte-
grity. Unilateral deployment would be highly
problematic from the perspective of national
security and international relations, as high-
risk CE might benefit a few but create grave
threats to other nations, perhaps provoking
the latter to retaliation. The international
community will thus have to develop mecha-
nisms to monitor for unilateral development
or deployment, similar to the monitoring of
nuclear technology or terrorist activities. Such
monitoring could be based on an international
regime for CE that would require parties to
control activities within their jurisdiction, and
would clarify jurisdiction over activities
outside of national territory, as on the high
seas or in outer space.
The military will have an interest in turning CE
into a classified matter of national security.
Most armies could – if they were to engage –
offer significant means to develop CE techno-
logies. In fact, the US agency DARPA already
assessed CE in a non-classified meeting in
2009 involving leading CE researchers. Cap-
ture of CE by the military or other interest
groups could be the worst start imaginable,
7
Bodansky, D. (2011): Governing Climate Engineering:
Scenarios for Analysis”, Harvard Project on Climate
Agreements Discussion Paper 2011-47, Cambridge Mass.

F N I C L I M A T E P O L I C Y P E R S P E C T I V E S 5
8
‘tainting’ the issue before the public could
form any kind of unbiased opinion. Transpar-
ency and clear disclosure of conflicts of inte-
rest – similar to current practice in medical
research – are crucial, to prevent such loss of
credibility and the potential for a new type of
arms race.
Dedicated public participation in developing
norms on the use of CE technology is a neces-
sary condition to create the long-term support
and stability essential to successful gover-
nance of such an intergenerational effort.
As a key element of governance, research on
the design of policy instruments to include CE
in ongoing mitigation schemes will need
special attention. For example, integrating CE
into carbon markets will require the definition
of a common trading unit for emissions reduc-
tions, CDR and SRM activities. This unit might
be based on the contribution of the mitigation
option to the reduction of radiative forcing
and it could for practical reasons represent
the equivalent forcing of 1 metric ton of CO2
(roughly 5x10-13W/m2). The design of such a
novel ‘climate currency’ will have to account
for the various risk components and the side-
effects, as well as temporal aspects in attribut-
ing monetary value of the effects of each
specific technology. Given our interest in de-
signing efficient mitigation policy instru-
ments, we will look into design options both
for market and non-market policy instruments
addressing CE in detail in forthcoming public-
cations.
Conclusion
Despite two decades of international climate
policy, emissions of greenhouse gases have
continued to rise. Emission reduction efforts
are increasingly seen as inadequate to stay
below the 2°C threshold agreed internation-
ally, but countries shy away from shouldering
the burden of emissions abatement. There-
fore, in the next decades an increase in mete-
orological extreme events is increasingly likely
to trigger public pressure to find quick solut-
ions to halt climate change. Climate engineer-
ing, especially the apparently cheap and high-
leverage Solar Radiation Management pro-
posals, will be attractive in this context. But
SRM could turn into a Pandora’s Box if not
managed carefully. A sudden political demand
for implementing CE could end in disaster if
pressure leads to premature deployment. It is
vital to establish a solid understanding of CE
with all its indirect effects as well as signifi-
cant acceptance and thus legitimacy. Since for
many CE options, the risks seem negatively
correlated to costs, a global research coordi-
nation effort is needed that is fully transpar-
ent and avoids biases introduced by interest
groups. The IPCC would be the right forum to
harness this research. Research should go
hand in hand with the development of new
norms and international approaches in moni-
toring, similar to the case of nuclear weapons
or terrorism. It is time for climate engineering
to enter the discourse on climate change miti-
gation – in a research-led, transparent and
conscientiously governed manner.
About the author(s)
Matthias Honegger holds a degree in environmen-
tal science from ETH Zurich. His studies in Switzer-
land, Norway, the UK and the US join economic and
natural science outlooks on climate change. At
Perspectives his work deals with the various
implications of climate engineering proposals.
Axel Michaelowa is senior founding partner of Per-
spectives and researcher at the University of
Zurich. He has 18 years of experience in climate
policy.
Sonja Butzengeiger-Geyer has worked on climate
policies since 1999. She started her professional
career at the Hamburg Institute of International
Economics, and worked for the German Environ-
ment Ministry for one year. In 2003, she cofounded
Perspectives GmbH and is one of the Managing
Directors of the company.
© Fridtjof Nansen Institute, P.O.Box 326, 1326 Lysaker, Norway / www.fni.no and
Perspectives, Baumeisterstrasse 2, 20099 Hamburg, Germany / www.perspectives.cc
ISSN 1893-1510 / ISBN 978-82-7613-650-0
Series Editor: Anna Korppoo (anna.korppoo@fni.no)