1. Introduction Java Climate Model, Scenarios to limit global warming to 2°C, +> implications for Brazilian energy sector
Presentation for COPPE-EPD 24 Aug 2011, Cenergia (+a few updates from later presentations)
Dr Ben Matthews Université catholique de Louvain (UCL) Centre de recherche sur la Terre et le climat Georges Lemaître (TECLIM) in IVIG, COPPE 14 may - 25 august 2011

2. ... who is this ...?
90-93 Univ Edinburgh - focus environmental chemistry +...
93-00 PhD UEA Norwich - Air-Sea CO2 Fluxes, catalysis marine algae + European Study of Carbon in Ocean Biosphere Atmosphere + project Qingdao Ocean Univ China + UNFCCC COP2 (GCI) COP3 Kyoto (SGR)
01-02 early development JCM Denmark DEA, Norway UNEP-GRID, Switzerland Univ-Bern + COP 6,7 ...
02-11 UCL-TECLIM (formerly ASTR - Louvain-la-Neuve Belgium) + projects CLIMNEG, ABCI + support to IPCC vice-chair, IPCC scenarios process + support European UNFCCC science expert groups (SBs, COPs ) + MATCH, project INFRAS (swiss), UNEP paper etc.
now here in IVIG... (until mid august)

3. Overview: Range climate models, introducing JCM Some specific issues
Historical contributions & Alternative metrics
Global 2°C pathways
Regional emissions, pledges, effort-sharing, equity
new Energy module
Demonstration JCM

4. Various types of climate models
(AO)GCMs - Global Climate (General Circulation) Models (AO = coupled Atmosphere and Ocean) 3D models based on real physics (box size about 0.5°-2.5°=> parameterised clouds, vegetation etc.). GCMs evolve towards ESMs (Earth System Models) including coupled biogeochemical cycles and ecosystem responses. Include intrinsic variability. Running multi-century scenarios takes weeks of computation.
RCMs - Regional Climate Models: 3D models with higher resolution for regional impacts studies - better clouds, topography Focus on one region (eg NE Brazil) over short timescale (eg decade). “Nested” within GCM providing boundary conditions.
EMICs - Earth-System models of Intermediate Complexity “2.5D” models first designed to study long paleoclimate timescales (eg glacial cycles), => much faster than GCMs, but incorporating nonlinear physical and biogeochemical feedbacks and “tipping points”. Useful for long-term future, especially sea-level rise.
IAMs - Integrated Assessment Models Designed to explore wide range of mitigation scenarios / climate policy proposals. Simpler parameterised physics and carbon cycle, but better socio-economic projections, maybe including energy sector, land-use change, optimisation, equity etc.
JCM - Java Climate Model - where does this fit ? began as interactive tool for global dialogue, evolved to be between IAMs and EMICs

5. Range of model complexities / specialities
National energy planning models
Future supercomputer clusters / cloud?
complex IAMs (IMAGE, MESSAGE, AIM... )
JCM all-round, interactive
Socioeconomic complexity
simple IAMs (optimisation, game theory - eg RICE)
EMICs (>1000yrs) (CLIMBER, MoBiDic... )
ESMs (HadEM, ...)
GCMs (HadCM, ECHAM, GISS...)
RCMs (Precis CCLM, ...)
algebraic models (old, for teaching)
Physical complexity / resolution
Fast (many scenarios) - Medium - Slow (few scenarios) Regional (+short-term)

6. Model complexity <=> applications
Complex 3D models
Intermediate Complexity
Simple models
Reports (- eg IPCC)
Strengths
Physical realism, High-resolution, variability
Biogeochemical feedbacks, Long timescales (paleoclimate)
Visualisation, Transparency
Accessibility,
Simplicity
Weak-nesses
Slow, not flexible, few scenarios, few people can use
Neither intuitive nor realistic => hard to interpret
Lack physical basis - don’t use alone
Lack flexibility,
Lost information
(esp. coupling)
Role in Integrated Assessment
Calibrating simpler models
Extreme events
Probabilistic Risk analysis
Long-term (SLR)
Tipping points
Explore options /risk-value choices
Stakeholder dialogue
Consensus Synthesis

7. Linking / intercalibrating models and scenarios
So - for whole climate problem we need a suite of models, not one
Need to be consistent, inter-calibrated, with coherent scenarios
IPCC (WG1) new Representative Concentration Pathways GCMs => full range climate response => tune IAMs
IPCC (WG3) new socioeconomic scenario library (?) incorporate into IAMs, compare diff sources / resolutions /timescales
both => Impacts and Adaptation (IPCC WG2) + need iterative feedback => needs more time !

8. Introducing Java Climate Model

9. Special focus of JCM
Global Stabilisation Scenarios (2°C etc.) - multi gas, multi-indicator, flexible pathway shapes sensitivity to climate, & carbon cycle uncertainties - sharing of emissions / effort between countries
Interactive tool for global dialogue - enable people to explore relative sensitivity to policy options and scientific uncertainties: “the ultimate integrated assessment model is the global network of human heads”
Core science copied from other models and IPCC reports. Fast, efficient implementation convenient for both: - interactive exploration useful for teaching - integration over uncertainty - risk analysis

10. Physical science of JCM
Bern carbon model including climate feedbacks and ocean chemistry
Atmospheric chemistry and radiative forcing calculated for >30 gases and aerosols
UDEB climate model (parameters tuned to fit GCMs)
Originally intended to be consistent IPCC- TAR, mostly updated to AR4
Speciality - inverse calculations to stabilise temperature

11. Regional Emissions Scenarios
Regional data CO2, CH4, N2O, CO2eq, population, GDP, ...
Diverse region sets depending application
Calculates by country , by region thereafter (=> 2300+)
Calculates LUC emissions from biome changes
Originally - top-down sharing, convergence, depending population, GDP, etc.
Recently - added “pledges” to gradual participation approach thereafter

12. Other applications of JCM
Historical Contributions to climate change (ACCC/MATCH) + focus historical landuse change with IVIG (revisit? compare “carbon space”, “equitable access to sustainable development?”)
Aviation scenarios incl climate impact cirrus clouds/other gases (ABCI) (comparison short/long-lived gases - reapply to issue GWP, GTP ?)
Economic analyses - sensitivity to scenarios and integration over regions, wealth, time, and uncertainty/risk (Climneg2 project)

13. Overview: Range climate models, introducing JCM Some specific issues
Historical contributions & Alternative metrics
Global 2°C pathways
Regional emissions, pledges, effort-sharing, equity
new Energy module in JCM
Demonstration JCM

14. Historical contributions / Alternative metrics
Link to COPPE - IVIG was via MATCH
Special focus on historical contributions from land-use- change

15. => Integrating over time / gases
MATCH/ACCC - historical contributions to CC still relevant in equity discussions (comparing e.g. Indian proposal “carbon space”) Impact of each ton CO2 / other gas changes with time, * lifetime/decay in atmosphere => changing climate impact * technology evolution => changing contribution to sustainable development
Future contributions of different gases - technically similar question, - alternatives to GWP for comparing short/long-lived gases - depends impact of most concern (eg biodiversity vs sea-level rise) - also depends reference scenario - e.g. 2°C not constant concns - => We should consider new metrics, not only GTP + calculate tables differences variants, write joint paper input IPCC-AR5 ?

16. => Integrating over time / gases (contd...)
Original “Brazilian-proposal”
Transparent “Policymaker model” good concept (but now computer-tools can replace algebra)
“Double integral” was misleading, - Atmosphere has little heat capacity to accumulate warming, - Second integral applies to deep ocean + ice-melt => Sea-level rise, not surface Temperature
Revised Brazilian proposal JM+GM-F, and more recent presentations (eg UNFCCC workshop 2009)
Still greatly over-estimated duration of temperature response,
+ why exclude other gases + LUC ?
=> misleading results now help delay / excuses by China etc.
GWP is different issue - not only temperature in one year
compromise to integrate all impacts including rate of warming, SLR, regional ...
not constructive to tell UNFCCC that IPCC all wrong...

17. Global 2°C stabilisation pathways
Policy compromise to interpret UNFCCC Article 2 European Policy since 1996, global policy since (Copenhagen / Cancun) But is it enough to avoid dangerous impacts ?
What does this imply for emissions pathways? What is acceptable risk of passing this level?
Stabilisation under uncertainty First 2°C probabilistic analysis with JCM in 2003
Pathway Shape New pathways (Swiss INFRAS project)
UNEP “Gap” report and other recent papers

18. 2°C Stabilisation under uncertainty - 2003
Example below from presentation of Matthews & VanYpersele at WCCC 2003 Moscow, also to EU strategy Firenze.
Shifting from concentration to temperature target shifts the burden of uncertainty from impacts to mitigation

19. JCM: 2°C / 1.5°C pathways after the 2020 pledges
Caveat: No rapid ice melt!

20. Relating temperature - concentration
Global average temperature rise
above preindustrial
(for best-guess model parameters)
CO2 equivalent concentration
Including all GHG gases and aerosols
Dotted lines – CO2 only
(from Java Climate Model – jcm.ivig.coppe.ufrj.br or jcm.climate.be)

21. Relating concentration - emissions
CO2 equivalent concentration
Solid lines - including all GHG gases and aerosols
Dotted lines – CO2 only
CO2 equivalent emissions
Solid lines – All gases + LUC
Dotted lines – Fossil CO2 only
(from Java Climate Model – jcm.ivig.coppe.ufrj.br or jcm.climate.be)

22. UNEP “Gap” report, EGSci paper (Cancun)
Both apply probabilistic approach (>66% chance <2°C), compare many IAM scenarios
Key message: 2020 pledges are not enough, about half-way there. If emissions peak higher, they decline at unfeasible rates later.

23. UNEP “Gap” 2010

24. Conclusion
EGSci 2010
The Copenhagen Pledges are indeed a first step in the right direction, but are by themselves not enough for the reductions we need by 2020.
Instead further steps are necessary to enhance these pledges in a binding agreement to bring 2020 emissions to a level that does not burden future generations with potentially unfeasible emission reduction rates.
This document was commissioned and is now provided with the intent to inspire and assist in discussions working towards a post-2012 climate agreement.
The European Union is looking forward to an exchange of views stimulated by this document.
“Scientific Perspectives after Copenhagen” December 1st, 2010 Cancún, Mexico

25. Regional emissions, pledges, effort-sharing, equity
Current emissions distribution
2020 Pledges / NAMAs
Peaking
Historical contributions, vs feasibility
Equity

26. CO2 Emissions (per capita) 1990, 2000, 2010 Bars: width: population, height CO2 per capita, area CO2 lower part Energy CO2 upper part Land-Use CO2 (white = negative) (2010 is estimate based on 2009 data, extrapolation energy intensity trends)
USA
USA
USA
W. Europe
Brazil
Brazil
China
Brazil
China
China
Africa

27. Copenhagen Pledges / NAMAs 2020 + percapita emissions 1990-2020

28. CO2 Emissions/GDP (intensity) - 2005 and 2020

29. 2050 emissions and peaking
Cancun mandate to COP17, Durban 2011: to agree 2050 global emissions goal and peaking timeframe - crucial to give signal for longterm infrastructure investments
global total constrains China more obviously than Brazil
of course, peaking is earlier in some countries than others...
integral of emissions may be better indicator for avoiding 2°C

30. CO2eq Peaking later in developing countries 2°C stabilisation scenario, global peak 2020, CO2eq includes energy+LUC CO2, CH4, N2O
China
India
USA
CFCs
Africa
Russia
Brazil

31. CO2 Peaking later in developing countries 2°C stabilisation scenario, global peak 2020, energy CO2 only
China
India
USA
Africa
Russia

32. CO2eq per capita (same 2°C scenario)
2°C stabilisation scenario, global peak 2020, CO2eq includes energy+LUC CO2, CH4, N2O

33. Historical contribution to temperature rise, per capita (including emissions of CO2 (energy+LUC), CH4, N2O, since 1890, for same 2°C scenario)

34. Emissions Pathways
Per capita CO2 emissions (left) and rate of decline %/year (right) Scenario -50% wrt 1990 by 2050, multiple participation / sharing criteria

35. Equity Issues
The greatest inequity is in the distribution of climate change impacts - think first about the most vulnerable countries /regions
In UNFCCC, “Equity” has become “excuses” (for China) ...? Emphasis historical responsibility / “Equitable Access to Common Atmospheric Space” => delay action => greater warming => greater inequity
=> How to broaden the equity debate? Include both mitigation + impacts Interpret “Equitable Access to Sustainable Development”?
Global 2050 target - Annex 1 target => implied developing country limit? Equal per capita: Ax1 approx -86% wrt 1990 by 2050 if global -50% (European position: Ax1 -80 to -95%)

36. JCM Energy module (under development):
To use JCM to calculate projection for Brazilian energy sector - need sectoral energy module ( JCM regional shares module only based on political / equity criteria )
How to differentiate extra reductions by sector / fuel ? - shift towards lower-carbon energy, some options cheaper / more capacity
Not only energy supply - what about changing demand? energy efficiency, buildings, transport infrastructure planning etc.Relation energy / land-use change? - distribution of effort energy vs LUC, implications of biofuels for landuse, capacity...

37. JCM Energy module - Characteristics:
BUT ...
For pathways avoiding 2°C warming, need long-term global perspective Timescale climate stabilisation at least 150 years, and global emissions reductions >>50%
=> Essential to consider investment in infrastructure, with learning (power plants, new transport networks, buildings and planning...)
Extrapolation of trends can’t capture this, nor can “equilibrium” model or MAC approach => disequilibrium / dynamic partial-equilib / agent based approach
Need all regions, all sectors to get global total (and hence climate, 2°C...) + global trade => fossil fuel prices, energy intensive goods, technology transfer, carbon market
Too much for a few weeks ...

38. JCM Energy module - Progress:
Incorporated data to JCM from IEA- WEO (16 regions, 3 scenarios, , various fuels+renewables) Created energy-data module to view / explore these scenarios
Start to develop dynamic model - start Electricity sector, convenient investment cost data (IEA) + technical potentials (SRREN) - considered challenges Transport sector - ?? how to acccount for buildings, construction, energy intensive goods ....
More time needed to complete...

39. JCM Energy module - IEA scenarios:
Example IEA 450 Scenario
Regional CO2 Emissions, Energy Sources (TPED)

40. JCM Energy module - Electricity:
Regional capacities starting from IEA data, Capacity investment depends expected price
Investment costs from IEA (with learning), regional technical potentials from SRREN, ...
Very sensitive to interest rates, as well as carbon price etc.
To add: Efficiency, dynamic learning, CCS, Extra capacity for peak demand...
Electricity Production, Costs, CO2 Emissions

41. JCM Energy module - Transport:
Transport: more challenging to get data on investments ...
Substitution of fuels is not enough for whole world, (dominant in western hemisphere - but rest of world is building railways...)
Must consider modal shifts (road, air... => rail, ships ...) but modal shift not yet accounted for in most models (eg IEA 2009 shifts scenario - arbitrary - not efficient allocation investments)
needs more time...

42. Contributions to CENPES project
Meetings / course / demonstrations / website...
Energy module of JCM - focus on longterm => infrastructure investments - done electricity, started transport, - buildings + industry + trade & feedbacks to complete... need more time - when done - illustrate sensitivity parameters + feedbacks inc prices, explore “needs-based” / feasibility approach to sharing effort
Sensitivity Analysis 2°C - postponed => quick... in general Brazilian energy emissions per capita low ( << China) => poorly constrained by equity-based approach, unless LUC rises again

43. To develop here: sensitivity analysis 2°C=> Brazilian energy sector
inverse calculation, automated to explore many variants - what makes most difference?
Starting from 2°C stabilisation scenarios (various pathway shapes - approaching 2°C faster or slower)
+ Varying physical climate parameters => Concentrations GHGs (climate sensitivity, ocean mixing, aerosols etc. - probabilistic weighting of sets as ? )
+ Varying carbon cycle parameters => Global CO2 emissions (ocean sink, CO2 fertilisation, climate feedback on soil respiration, etc.)
+ Varying sharing of national emissions /effort => Brazilian emissions (convergence, gradual participation, ... starting from pledges + higher and lower? )
+ Varying distribution between sectors => Brazilian energy sector (uncertainty in LUC will make a big difference for Brazil)

44. To develop later...
Incorporate new IPCC RCP scenarios and GCM results, to test and re-calibrate physical climate system, feedbacks...
Bottom-up socioeconomic scenarios, demographics, local data compare / connect scenarios in new IPCC socioeconomic scenario library?
1.5°C scenarios? (decline after peak - interest many countries)
Alternative metrics for integrating gases (GWP, GTP etc.)
Synthesis of regional & sectoral impacts of climate change - new functions based on AR5 WG2?
Re-develop optimisation module for cost-effective solutions / re-calibration ?
Simpler versions for teaching, for smaller screens, ...?
Recall concept democratisation of climate science-policy interface: “ultimate integrated assessment model is the global network of human heads”

45. Overview: Range climate models, introducing JCM Some specific issues
Historical contributions & Alternative metrics
Global 2°C pathways
Regional emissions, pledges, effort-sharing, equity
new Energy module
Demonstration JCM

46. JCM - Demonstration
recent updates on IVIG site: jcm.ivig.coppe.ufrj.br belgian site:

47. Hints for using JCM
If problem with “launch” button, can try zip package note page about issues with java (+ tell me) be patient on first startup - wait while loading data
Unfold tree => modules => plots, parameters
right-click on plots => stack, table, save... right-click on curve-sets in tree => variants, more plots
parameters only appear if they are useful
increase complexity level => more options
If strange result try: Reset “scientific” parameters
Read how-to-use pages and “help-on-click” doc

48. Dr Ben Matthews matthews@climate.be
Introduction Java Climate Model, Scenarios to limit global warming to 2°C, +> Implications for Brazilian energy sector
Obrigado - Perguntas ?
* Range climate models
* Historical contributions & Alternative metrics
* Global 2°C pathways
* Regional emissions, effort-sharing, equity
* Energy module
* Intro / Demo JCM
Dr Ben Matthews
recent updates on IVIG site: jcm.ivig.coppe.ufrj.br belgian site: