Title: UNFCCC Article 2, Stabilisation and Uncertainty
1- UNFCCC Article 2, Stabilisation and Uncertainty
- probabalistic interactive exploration of
scenarios using Java Climate Model - ASTR seminar 17th Feb 2004Ben Matthews
matthews_at_climate.be - (with Jean-Pascal van Ypersele
vanyp_at_climate.be) - Institut dastronomie et de géophysique G.
Lemaître,Université catholique de Louvain,
Louvain-la-Neuve, Belgium - www.climate.be (UCL-ASTR)jcm.chooseclimate.org
(interactive model)
2(No Transcript)
3UN Framework Convention on Climate Change
Ultimate objective (Article 2)
- '...stabilization of greenhouse gas
concentrations in the atmosphere at a level that
would prevent dangerous anthropogenic
interference with the climate system. - Such a level should be achieved within a time
frame sufficient - - to allow ecosystems to adapt naturally to
climate change, - - to ensure that food production is not
threatened and - - to enable economic development to proceed in a
sustainable manner.'
(technologies, lifestyles, policy
instruments) Emissions pathways(biogeochemical
cycles) Critical Levels (global temperature /
radiative forcing) Critical Limits (regional
climate changes) Key Vulnerabilities
(socioeconomic factors)
inverse calculation
4European Union 2 C limit
- EU Council Of Ministers 1996
- "...the Council believes that global average
temperatures should not exceed 2 degrees Celsius
above pre-industrial level and that therefore
concentration levels lower than 550 ppm CO2
should guide global limitation and reduction
efforts." - "This means that the concentrations of all GHGs
should also be stabilised. This is likely to
require a reduction of emissions of GHGs other
than CO2, in particular CH4 and N2O" - However, widely varying interpretations of
implications for emissions! - Why? Java Climate Model may help to investigate...
5- Stabilisation scenarios in Java Climate Model
- (Article 2 critical limits gt critical levels
gt emissions pathways) - Inverse calculation to stabilise
- CO2 concentration (as IPCC "S"/ WRE scenarios)
- Radiative Forcing (all-gases, "CO2 equivalent")
- Global Temperature (e.g. to stay below 2C limit)
- (Sea-level -difficult due to inertia in ocean /
ice) - JCM core science very similar to IPCC-TAR models,
but (unlike TAR SYR) JCM stabilisation scenarios
include mitigation of all (21) greenhouse gases
and aerosols, scaled w.r.t. SRES baseline.
6- Stabilisation scenarios in Java Climate Model
- CO2 concentration scenario is a Padé
polynomial(similar to formula of Enting et al
1994 for IPCC S/WRE)defined by - 2000 concentration c
- 2000 gradient dc/dt,
- 2000 second derivative d2c/dt2 (ensures smooth
emissions trajectory), - stabn year concentration
- stabn year gradient (zero if stabilising
concentration)Also define quadratic curve to
continue from stabn year until 2300. - If stabilising radiative forcing or temperature
(or...) iterate to find best concentration and
gradient in stabilisation year. - Iterates 1-10 times, depending on magnitude of
change (reuse of correction factors so efficient
for dragging control). - Explore interactively by dragging target curve
with mouse - Or systematically calculate probabilistic
analysis ...
7- Systematically exploring uncertainty
- 81 Carbon cycle variants
- 3 Land-use-change emissions (Houghton, scaled),
- 3 CO2 fertilisation of photosynthesis ("beta"),
- 3 Temperature-soil respiration feedback ("q10"),
- 3 Ocean mixing rate (eddy diffusivity of
Bern-Hilda model) - 6 Ratios of emissions of different gases
- Emissions of all gases (including CH4, N2O,
HFCs, sulphate/carbon aerosol and ozone
precursors) reduced by same proportion as CO2
with respect to one of six SRES baselines - note atmospheric chemistry feedbacks included,
but not varied - 84 Forcing/Climate Model variants
- 3 Solar variability radiative forcing
- 4 Sulphate aerosol radiative forcing
- 7 GCM parameterisations climate sensitivity,
ocean mixing/upwelling, surface fluxes (W-R UDEB
model tuned as IPCC TAR appx 9.1) - note for sea-level rise, should add more
uncertainty in ice-melt
8Demonstration of JCM
9 Carbon Cycle
Climate Model
Other gases/Aerosols
10Probability from fit to historical data
- Relative probability of each set of parameters
derived from inverse of "error" (model - data) - Measured global temperatures (CRU proxies)
- Measured CO2 concentration (Mauna Loa others)
- Reject low-probability variants (kept 468 / 6804)
- Ensures coherent combinations of parameters, e.g.
- More sensitive climate models with higher
sulphate forcing - High historical landuse emissions with higher
fertilisation factor - Still 2808 curves per plot (including 6 SRES per
set)So show 10 cumulative frequency bands
(using probabilities)
11 Carbon Cycle
Climate Model
Other gases/Aerosols
12Shifting the Burden of Uncertainty
- On average, all sets of scenarios stabilise at
the same temperature level of 2C above
preindustrial level. But their uncertainty
ranges are very different! - A Temperature limit rather than a Concentration
limit reduces the uncertainty for Impacts/
Adaptation... - (assuming we commit to adjust emissions to stay
below the limit, as the science evolves) - ...however this increases the uncertainty
regarding emissions Mitigation pathways. - Which is better?
13What CO2 level stabilises Tlt 2C ?
- note 90 of cum freq means that 90 of variants
weighted by probability fell below this level - note concentrations derived from IPCC-TAR
science are lower than those from SAR,
principally due to less sulphate cooling, and
slightly higher sensitivity - note 550ppm "CO2 equivalent" (all gases) would
bring us close to 2C. However, to keep the
temperature level, total radiative forcing (and
hence CO2 equivalent) must decline gradually.
This is possible while CO2 remains level, due to
declining CH4 and O3 (which have short
lifetimes).
14Is it 'realistic'? check trends change in CO2
emissions per capita per year x-axis from 1950 to
2050, y-axis from 10 to -10 Left
Stabilisation at 450ppm Right SRES A1B
15Why should CO2 concentrations/temperature
constant?What about forcing (all gases), or
sea-level, or ...?Inertia in the climate
systemStabilising CO2 alone doesn't stabilise
temperature (as below from TARSYR Q6) However
stable CO2 may correspond to stable Temperature
if other gases with shorter lifetimes are also
mitigated to a similar extent.
16As natural scientist, am not advocating 2C level,
only that derivations from it should be
consistent with latest science...
- Interpretation of Article 2 needs a global
dialogue (Article 6) - Risk/Value Judgements (including equity
implications) - Impacts Key Vulnerabilities? Acceptable level
of Change? - Risk Target Indicator? Acceptable Level of
Certainty? - (choice of target indicator shifts the burden of
uncertainty) - Such risk/value decisions cannot be made by
scientific experts alone. - There is not yet any global consensus about the
safe levelhow can modellers help
citizens/policymakers to explore this?
17Stabilisation is consistent with sustainable
development, but this is not the only paradigm
for future climate policy...Economists prefer
optimisation (maximising welfare)Suggest
that the optimum level is much higher than EU
policy -e.g. Climneg papers 600-1000ppm (and
continuing to rise).But hidden risk/value
judgements very controversial.And not
considering uncertainty...Others consider
pessimisation (avoiding critical thresholds of
change) -e.g. Tolerable Windows guardrail
approach (ICLIPS)WGBU concludes that we need to
aim well below 450ppm.Yet guardrails too rigid -
also not considering uncertainty!Can these all
be reconciled?gtRisk analysis framework?
18Integrated assessment (for Climneg
project) Assess balance of mitigation and
impacts/adaptation, etc. Apply to Game
Theory (coalition formation etc.) Economic costs
module added to JCM, applying Climneg formulae.
(abatement costs, damage costs, time-integral
with discount rate, etc.) Abatement costs (RICE /
MACGEM) based on comparison to SRES baselines.
Need to incorporate MACs for each gas.
Combine with probabalistic approach to convert
optimisation to Risk Analysis. But I do not
believe simple climate-impact cost functions from
economists - much more work needed (need your
help!) Also should also consider a set of
world-views for impact valuation / equitable
aggregation / discounting / risk-aversion
etc. JCM could help to make the assumptions
transparent.
19(No Transcript)
20Regional Climate Impacts
- GCM climate patterns
- scaled instantly to JCM average
- Latest datasets from IPCC-DDC
- JCM module to be used in DDC website
- Climate is multi-dimensional
- Temperature average, min, max, dtr
- Precipitation, humidity, clouds, sunlight,
pressure, wind - Seasonal cycle, view monthly animation
- Compare GCMs to see uncertainty
- Combine change with baseline climatology
- Calculate averages for any country / region
- (now for various region-sets SRES, JCM, RICE,
CWS15, EDGAR, 50, all nations...) - Next challenge is to derive socioeconomic /
ecological impacts from such data, - using regional socioeconomic models to assess
vulnerability.
21Flexible region sets at both ends of the chain,
to connect a variety of data-sources and
applications. JCM12, JCM50, RICE, CWS15,
SRES4, TGCIA, IMAGE, EDGAR, CDIAC/Houghton,
All Nations, subdivisions... Idea analyse
sub-regions of large diverse countries (eg US
states, Russian oblasts, Chinese provinces), to
consider potential new coalitions if central
govts can't agree policy?
22- Regional Distributions
- JCM can also be used to explore...
- Attribution of responsibility
- Regional climate change patterns
- Abatement and Impact Costs In combination with
stabilisation scenarios,and scientific
uncertainties
23ClimNeg game-theory of coalition-building /
transfersis based essentially on the victim
pays principle.(the more a region anticipate to
suffer from climate change impacts, the more it
should pay others not to pollute)So what about
the polluter pays principle?
24Attribution of responsibility for climate change
(Brazilian Proposal)
- Many potential applications, comparing impacts
due to emissions from - countries (pay for adaptation?)
- projects(CDM)
- timeslices (inter-generational equity)
- gases (replace GWP ?)
- Calculations considered several
- gases CO2 fossil, CO2 landuse, CH4, N2O
- regions (4 SRES, 12 JCM, 15CWS, EDGAR...)
- indicators concn, forcing, temp, sealevel
- time-slices and future scenarios(now including
stabilisation scenarios) - methods for attributing non-linear processes and
Feedbacks
Sensitivity to uncertainties is much less for
relative attribution, compared to absolute(maybe
similar effect applies to other problems in
Climneg? - should test) JCM contributed to
UNFCCC intercomparison ( workshops Hadley centre
2002, Berlin 2003). Next stage of intercomparison
during 2004, report to SBSTA 2005. Process helps
to engage developing countries, who often mention
historical responsibility.
25- Interpretation of Article 2 needs a global
dialogue (Article 6) - Risk/Value Judgements (including equity
implications) - Impacts Key Vulnerabilities? Acceptable level
of Change? - Risk Target Indicator? Acceptable Level of
Certainty? - (choice of target indicator shifts the burden of
uncertainty) - Such risk/value decisions cannot be made by
scientific experts alone.
26- The ultimate integrated assessment model
remains the global network of human heads. - To reach effective global agreements, we need an
iterative global dialogue including citizens /
stakeholders. - The corrective feedback process is more important
than the initial guess. So let's start this
global debate! - But we still need models to provide a
quantitative framework for the discussion. JCM
was developed to make models more accessible and
transparent.
27(game -theory, game practice...)Role-play on
Article 2 with students Louvain la Neuve,
Belgium, Dec 2002, as if COP11, 2005,Presented
at COP9 Milano, Dec 200360 university students
grouped in 17 delegations (Belgium, Denmark,
Russia, USA, Australia, Saudi-Arabia, Venezuela,
Brazil, Burkina-Faso, Marroco, Tuvalu, India,
Greenpeace, GCC, FAO, WB/IMF, Empêcheurs)had the
task to agree by consensus in a UNFCCC-style
process a quantitative interpretation of
Article 2, an equitable formula for funding
adaptation.Delegates used Java Climate Model to
explore options / uncertainties. Can "justify"
diverse positions by selecting parameters /
indicators !
28Conclusions of role-play
- Equity implications were key aspect of discussion
- Final compromise between Russia and Tuvalu (after
US quit) - Quantitative interpretation of Article 2
Temperature rise (lt1.9C 2100-1990)
Sea-level rise (46cm 2100-1990) Principles for
Adaptation funds Tax on emissions trading
Percapita emissions GDP formula Principles
sufficiency/capacity - Such "games" also help us to identify scientific
issues, e.g. Reconciling multi-criteria climate
targets (inconsistency maybe realistic in policy
compromises), Meaning of CO2 "equivalents" in
stabilisation context
29- Future development for global dialogue
- Could we combine such tools and experiences to
link groups from all corners of the world? - JCM also used for teaching in several
countriesUniv Cath de Louvain (BE) Open
University (UK), Univ Bern (CH), Univ Waterloo
(CA),... - Such web models might provide a quantative
framework for a global dialogue. Model can be
shared by saving snapshots of model parameters to
pass to others in asynchronous discussion forum.
30Experiment with Java Climate Model Try JCM at
jcm.chooseclimate.org Trying to combine research
and outreach Works in web browser, very
efficient/compact Instantly responding graphics
show cause-effect from emissions to
impacts, Based on IPCC-TAR methods / data, New
flexible stabilisation scenarios, Regional
emissions, abatement, costs, responsibility
Regional patterns of climate change Transparent,
open-source code, modular, scriptable, Interface
in 10 languages, 50,000 words documentation JCM
also developed with DEA-CCAT Copenhagen,
UNEP-GRID Arendal, KUP Bern
31Demonstration of JCM