The case for Geoengineering Research MIT 30 October 2009 Cambridge, - - PowerPoint PPT Presentation

1 1

The case for Geoengineering Research

David Keith keith@ucalgary.ca

• www.ucalgary.ca/~keith

Director, Energy and Environmental Systems Group Institute for Sustainable Energy, Environment and Economy University of Calgary

MIT

30 October 2009

Cambridge, MA

2

We can & must stop emissions

Example: Electricity accounts for 40% of global emissions We can make near‐zero emissions power today. With strong cost‐effective policy we could decarbonize electricity production within a quarter century. Technologies that are ready to go at reasonable cost now:

• Wind power (with backup & long distance transmission)
• Nuclear power
• Coal with CO2

capture and storage

• Central‐station solar thermal (with long distance transmission)

Cost: a few % of GDP. In the rich countries that is:

• Similar to the cost of all other environmental controls
• A bit bigger that current military spending
• Many times smaller than the cost of health care
• A good year’s GDP growth

3

Geoengineering Adaptation Mitigation Human actions that change climate Climate System Climate impact

• n human welfare

4

Solar radiation management ≠ Carbon cycle engineering

Carbon cycle engineering (CDR)

• Biomass + CCS
• Direct capture of CO2

from air

• Adding Fe to oceans
• Adding macro‐nutrients to oceans
• Adding alkalinity (Mg) to oceans
• Bio‐char
• Adding alkalinity to soils

Slow and expensive, but it gets the carbon out

Solar radiation management (SRM)

• Sulfates in the stratosphere
• Sea salt aerosols in low clouds
• Altering plant albedo
• Engineered particles in mesosphere

Fast, cheap, imperfect and uncertain; and it does very little to manage the carbon in the air

5

Cheap, Fast and Imperfect

Cheap Aerosol mass flux needed to create 4 Wm‐2 radiative forcing:

• 3×104

kg sec‐1

• f ≈1 μm aerosol into marine stratus clouds.
• 150 kg sec‐1
• f S ≈0.2 μm sulfate aerosol into the stratosphere.

Comparison

• 2.6×106

kg sec‐1 CO2 mass flux required to remove 2000 Gt‐CO2 (4 Wm‐2)† in 25 years. Fast Aerosols radiative forcing acts on timescales of days (troposphere) to years (stratosphere). Comparison

• Cutting emissions to zero stops concentration increase CO2

but does (almost) nothing for the CO2 already emitted.

• Economic inertia + carbon cycle inertia century long response times

6

Cheap, Fast and Imperfect

Imperfect Reducing insolation ≠ Decreasing CO2 Reducing insolation less energy input to the surface less flux from surface to mid troposphere weaker hydrological cycle Increasing CO2 cooler troposphere & warmer stratosphere Decreasing insolation cooler troposphere & stratosphere Decreasing insolation no direct reduction in geochemical impacts of CO2.

7

Timescales and inertia: carbon and sunlight

8

MIT, Center for Global Change Studies (2009)

Business as usual Strong climate policy

9

Morgan & Keith (1995) Forest et al (2002)

1 2 3 4 0.0 0.2 0.4 0.6 0.8 1.0 DV ER JO TP vanderveen clarkeg vanderwaal bindschadler greve truffer bamber calov ritz

Given 4 C global @ 2100 Greenland only Marshal & Keith

10

Inertia: CO2 lasts longer than nuclear waste

Solomon et al (2009) Irreversible climate change due to carbon dioxide emissions, PNAS, 106, 1704–1709.

If we instantly stop emissions of CO2 temperatures remain high for millennia

• 10,000 years after emissions stop, CO2

concentrations and warming will be about 1/2 to 1/3 the level they were on the day the coal plants shut down. The fission products in nuclear waste are gone in about 1000 years.

• 10,000 years after discharge nuclear waste is 3000

times less radioactive than it was one year after it left the reactor.

3000 ×

MIT Study on the Future of Nuclear Power (2003)

11

Inertia

Turn wheel (e.g., enact policy) Low emissions infrastructure is built at some rate after a time delay Emission reductions grow as the integral of the infrastructure build rate. Concentration reductions grow as the integral of emissions reductions. Reduction in temperature (from BAU) responds more slowly that reduction in concentrations due to ocean thermal inertia Climate reacts

Uncertainty + Inertia = Danger

12

Physical‐science research for solar radiation management

13

Stratospheric scatterers

Of order 2‐4 Mt‐S per year offsets the radiative forcing of 2×CO2 (~5% of current global S emissions, assuming fine‐mode particles) ~3 gram sulfur in the stratosphere roughly

• ffsets 1 ton carbon in the

atmosphere (S:C ~ 1:200,000) 10 $/kg 10’s of $bn per year ≈ 0 Cost not the deciding issue. Risk to risk. Lofting methods:

• Aircraft
• Naval guns
• Tethered balloon with a hose

Scattering design goals:

• Lower mass
• Spectral selectivity
• Altitude selectivity
• Direct: diffuse selectivity
• Latitude selectivity

Alternative scattering systems Oxides

• H2

SO4

• r Al2

O3 Metallic particles (10‐103 × lower mass)

• Disks, micro‐balloons or gratings

Resonant (104‐106 × lower mass ??)

• Encapsulated organic dyes

Self‐lofting particles

• Photophoresis.

14

Physical‐science research needs

• 0. Assessing climate response to solar radiation management.
• 1. Assessing the utility of sub‐scale experiments
• How small a perturbation (sulfate loading) can be detected over a given
• bservational period?
• How does SNR vary with modulation frequency and location?
• What can we say about the ability to detect side‐effects and problems?
• 2. Getting ready for the next big volcano
• What instrumentation and research planning is needed to take best

advantage of an eruption that puts Mt of S in the stratosphere?

• 3. Understanding and engineering the plume‐scale dispersal of sulfates.
• 4. Engineering system studies of stratospheric delivery systems.
• 5. Engineering dispersal systems and operational strategies for sea‐salt aerosols.
• 6. Non‐sulfate aerosols or engineered particles in the upper atmosphere.

15

Aerosol dynamics: AER‐ETH results

2D AER sulfate model (see below) First look at sulfate geoengineering with a model that does prognostic aerosol dynamics. Figure credits: Debra Weisenstein (AER) and Thomas Peter (ETH)

16

Sulfur questions

What are we injecting: H2 S? SO2 , S? H2 SO4 ? If H2 S then local dispersal does not matter since H2 S must be oxidized before aerosol dynamics start.

• H2

S injections end up being deposited on the large end of the particle size distribution very inefficient.

• Generalize to any gas phase injection?

If we are injecting S or H2 SO4 then result will depend on dynamic competition between nucleation and turbulent mixing in aircraft wake. Essentially no work as been done to look at engineering of injection systems.

17

Measuring the effectiveness of SRM

Most model experiments to date compare pre‐industrial with 2×CO2 plus SRM sufficient to bring surface temps back to pre‐industrial means. It has become common to talk about the 2×CO2 plus SRM

‐

pre‐industrial differences, particularly precipitation differences as the “impacts”

• f SRM.

There are several problems with this: 1. This would make sense if it were 1800 and we were analyzing the choice to burn a few 1000 Gt‐C and compensate its worst effects with SRM. 2. SRM reduces precip more strongly than it reduces precip‐evap. Quantities like soil moisture are more relevant to assessing impacts. 3. Analysis of CO2 ‐driven climate impacts has identified variability as a crucial driver, and one might expect precip variability to be reduced in high‐CO2 low‐insolation world. Clarity about language: Impacts

• r side‐effects

= ozone chemistry, rain out of particles into troposphere. Imperfection

• r ineffectiveness

= the fact that SRM climate change not same as CO2

18

Temperature and Precipitation Changes Under Variable SRM

Precip and Temp changes scaled to standard deviation of pre-industrial climate

Ricke, Morgan and Allen, submitted to Nature

19

Research strategy: How to balance the research portfolio?

• Carbon capture via bio‐char burial
• Climate model studies of the impacts/effectiveness of SRM.
• Economics, risk assessment, public policy analysis of SRM.
• Engineering design of hardware for sea salt aerosols.
• Engineering studies of stratospheric scattering systems: production,

dispersal, and delivery of sulfate (or other) aerosols.

• Micro‐scale field testing of sea salt aerosols.
• Micro‐scale field testing of sulfate dispersal systems.

Balancing critics & engineers

“It’s vital to have this “red team”

• f skeptics questioning geo‐engineering. But we need

more emphasis on a “blue team” to figure out what geo‐engineering approaches might work” ‐‐Homer‐Dixon & Keith. NYT Op‐Ed, 19 September 2008 Ease of Funding: Easy Difficult

20

Credit: Novim

21

Valuing information

22

A simple representation of imperfect temperature compensation

( )

2

ˆ ˆ

G

CO

T RF a T gT λ λ = − + r

2

0 ˆ

CO

RF T λ

T r

ˆ

G

gT λ

2

ˆ

CO

aT λ

2

ˆ

CO

T

ˆ

G

T

ϕ

( )

( )

,

T

D a g K T T = ⋅ r r

( )

[ ] [ ]

2 2 2

, cos

T

Additive Factors Multiplicative Factor

D a g K RF a g RF a g λ ϕ ⎛ ⎞ ⎜ ⎟ = − + + − ⎜ ⎟ ⎜ ⎟ ⎝ ⎠ 64 4 744 8 644 744 8

22

Climate Policy under Uncertainty: A Case for Geoengineering Juan Moreno-Cruz and David Keith, submitted to Environmental Research Letters

23

Value of information

Total Costs, $ Trillions

σ

σ σ

Radians Full uncertainty Learn after abatement Learn before abatement Optimal decision under uncertainty

• Two period model
• Climate sensitivity uncertainty
• Geoengineering uncertainty

24

25

Implications

26

Cheap, Fast and Imperfect

Cheap Policy challenge is control Decisions about implementation will be risk‐risk not benefit‐cost

• Abatement challenge is to get (almost) all to comply.
• SRM challenge may be to constrain the actions of rogues

Fast Decisions about the amount (if any) of SRM can wait until climate sensitivity uncertainty is resolved Given inertia + climate uncertainty SRM may be the only way to ensure global temperatures stay below some given threshold Imperfect SRM cannot obviate the need to cutting emissions (although it may substitute at the margin) Distribution of winners and losers under SRM not the same as distribution from CO2‐ driven climate change

27

Albedo modification CO Concentration

2

Geoengineering instead of mitigation

Radiative Forcing 2100 2000 2050

28

Albedo modification CO Concentration

2

Geoengineering instead of mitigation

Radiative Forcing 2100 2000 2050

Geoengineering to take the edge of the heat

2100 2000 2050

29

Warning: Moral Hazard

Knowledge that geoengineering is possible

• Climate impacts look less fearsome
• A weaker commitment to cutting emissions now

Opposite reaction possible: “if the pointy‐heads think we need to shoot nano dust into the stratosphere then we should get worried & get serious about cutting emissions” Whatever the reaction: it seems reckless and un‐precautionary to avoid looking at something that might help limit the damage of the CO2 already in the air.

30

Mitigation, SRM and CDR

Fact 1: Carbon emissions alter climate on millennial timescales. Assuming we rely on natural processes, that is ignoring CDR. Fact 2: Climate response to carbon perturbation is uncertain. Uncertainty in 2 × CO2 is 2‐3, and it gets larger if you include cryosphere etc. Fact 3: SRM is the only fast‐acting method that might limit climate risks posed by the CO2 already in the atmosphere. No plausible CDR can make a big dent in concentrations in less than about half a

• century. SRM much faster.

Conclusion 1: Cutting emissions is necessary but insufficient to manage climate risk. Necessary because SRM is imperfect. Conclusion 2: The capability for SRM is a necessary for climate risk management. Given carbon inertia & climate uncertainty the only known method to limit climate change to a given threshold, e.g., 2 C involves the capability to do SRM.

31

Questions & Opinions

Opinions

• 1. We need a serious research program
• Impacts, methods

and implications

• International
• Need not be large $$ to make enormous progress.
• 2. Current understanding of climate systems suggests that intelligently executed

climate engineering would reduce climate risks.

• 3. We need to break out of the foolish all‐or‐nothing dichotomy. Ramp up with

learning.

• 4. The science community should expect to loose control.

Questions 1. How can we best avoid the geoengineering mitigation trade off? 2. Should we work toward a treaty? Norms? An alternate mechanism?

32 32

33 33

34 34

35

www.ucalgary.ca/~keith/Geo.html

Shepherd, J., K. Caldeira, J. Haigh, D. Keith, B. Launder, G. Mace, G. MacKerron, J. Pyle, S. Rayner, C. Redgwell and A. Watson (2009). Geoengineering the climate ‐ Science, governance and

• uncertainty. The Royal Society (UK).

David Keith. (2009). Why Capture CO2 From The Atmosphere? Science, 325: 1654‐1655.

• J. J. Blackstock, D. S. Battisti, K. Caldeira, D. M. Eardley, J.
• I. Katz, D. W. Keith, A. A. N. Patrinos, D. P.

Schrag, R. H. Socolow and S. E. Koonin, (2009) Climate Engineering Responses to Climate Emergencies Novim Report. Juan B Moreno‐Cruz and David W Keith (under review). Climate Policy under Uncertainty: A Case for

• Geoengineering. Environmental Research Letters

David Keith. (under review). Photophoretic levitation of aerosols for geoengineering David Keith (2010). Engineering the Planet. Climate Change Science and Policy. S. Schneider and M. Mastrandrea eds. Stephens, J. C. and D. W. Keith (2008). Assessing Geochemical Carbon Management. Climatic Change, 90: 217‐242. David Keith (2001). Geoengineering. Nature, 409: 420. David Keith (2000). The Earth is Not Yet an Artifact. IEEE Technology and Society Magazine, 19: 25‐28. David Keith (2000). Geoengineering the Climate: History and Prospect. Annual Review of Energy and the Environment, 25: 245‐284. David Keith and Hadi Dowlatabadi (1992). A Serious Look at Geoengineering. Eos, Transactions American Geophysical Union, 73: 289‐293.

36 36

Warning: Slippery Slope

“Interest in CO2 may generate or reinforce a lasting interest in national or international means

• f climate and weather modification; once

generated, that interest may flourish independent of whatever is done about CO2 .”

1982 US National Academy study, Changing Climate.

37

Climatic Geoengineering Energy balance (forcings) Energy transport (feedbacks) Short-wave (albedo) Long-wave (emissivity) Ocean Atmosphere & Land Surface Geoengineering

• Space-based scatterers

Atmospheric scatterers Land surface albedo modification

• Ocean fertilization

Terrestrial ecosystem carbon capture Accelerated weathering Ecosystem productivity enhancement by genetic modification

• Large dams: Gibralter
• r bearing strait

OTEC Iceberg transport

• Chemical or physical

control of evaporation Hydrological engineering Weather control Modification of surface roughness

Impacts

• Sulphuric and

carbonaceous aerosols: Direct and indirect effects Surface albedo change: Clearing forests Built structures: Cities & roads ...

Radiativly active

gases: CO CH N O, etc

2 2

4

Secondary effects of

land use change: E.g., salinity changes in Atlantic due to increased evaporation in Mediterranean

• Hydrological

engineering Modification of surface roughness