climate science and uncertainty: improving assessments and … · climate science and uncertainty:...

Climate Science and Uncertainty: Improving Assessments and Decision Support

and an Overview of the

IPCC Special Report on Extremes

Dr. Richard H. Moss Joint Global Change Research Institute

Pacific Northwest Nat'l Lab/University of Maryland For

Rutgers University Initiative on Climate and Society Extreme Weather and Climate Change:

How Can We Address Uncertainty?

3/30/12 1 PNNL-SA-86838

Acknowledgements

!   Numerous colleagues including Jae Edmonds, Leon Clarke, Allison Thomson, Jennie Rice, Stephen Unwin, Michael Scott, Anthony Janetos, Elizabeth Malone, John Weyant, Tom Wilbanks, Ken Kunkel, Adam Parris, Holly Hartman, Kathy Jacobs, Gary Yohe, Kris Ebi, Tom Kram, Detlef van Vuuren, Keywan Riahi, Elmar Kriegler, Tim Carter

!   DOE Integrated Assessment Research Program, Bob Vallario !   NASA, Jack Kaye

3/30/12 Richard Moss, Joint Global Change Research Institute 2


How can we inform decisions that need to be made under deep uncertainty?

Assessments

Selected past and ongoing assessments: !   A series of international Stratospheric Ozone Assessments (started

in 1980s) !   Intergovernmental Panel on Climate Change (IPCC) periodic

comprehensive assessments (1990, 1995, 2001, 2007) plus numerous special reports

!   National Assessment of Climate Change Impacts on the United States (2000, 2009, 2013)

!   U.S. Climate Change Science Program (CCSP) Synthesis and Assessment Products (21 reports, 2006-2009)

!   Global Biodiversity Assessment (GBA) (1995) !   Millennium Ecosystem Assessment (MEA) (2004) !   Arctic Climate Impact Assessment (2005) !   Intergovernmental Platform on Biodiversity & Ecosystem

Services (IPBES) proposed IPCC-like assessment (2011-12)

Source: NRC (2007), Analysis of Global Change Assessments: Lessons Learned 3/30/12 Richard Moss, Joint Global Change Research Institute 4

Uncertainty language for assessments

!   Purpose: inform users of confidence levels and uncertainties

!   IPCC (Moss and Schneider, 2001) 3rd assessment, with revisions for 4th and 5th assessments

!   US National Climate Assessment !   Confidence assessment

process and language !   Progress, but consistent

failure to perform serious evaluations


!"#$%&'()($*$+(&,%&-,+-./'#,+0&"$%$"$+-$1&(,&"$2,"(&,"&-3)2($"4&&&&56&7")*#+8&)+1&'()9$3,.1$"&#+%,"*)(#,+&+$$1'&:+$&,"&*,"$&(;2$'&,%&'()9$3,.1$"&1$-#'#,+'&<,"&/'$'&,%&(3$&#+%,"*)(#,+=&3)>$&?$$+&-,+'#1$"$1&#+&%,"*/.)(#+8&(3$&-,+-./'#,+6&&

&

@6&A+#(#).&$>)./)(#,+&,%&$>#1$+-$&B+&$>#1$+-$&")(#+8&3)'&?$$+&)''#8+$10&-,+'#1$"#+8&(3$&(;2$0&)*,/+(0&C/).#(;0&)+1&-,+'#'($+-;&,%&$>#1$+-$6&A+&.#83(&,%&(3$&/'$&,%&(3$&#+%,"*)(#,+0&(3$&$>#1$+-$&#'4&&

&

D6&E"$2)")(#,+&,%&-,+-./'#,+&&F3$&-,+-./'#,+&"$%.$-('&(3$&1#>$"'#(;&,%&$>#1$+-$6&7,"&C/)+(#()(#>$&$'(#*)($'&,%&"$.$>)+(&2)")*$($"'&,"&*$("#-'0&)&")+8$&#'&2",>#1$1&<#+&G3#-3&(3$"$&#'&)&HIJ&-3)+-$&(3$&("/$&>)./$&%)..'=0&)+1&)&K?$'(&$'(#*)($L&#'&8#>$+0&#%&G)"")+($16&M#83&-,+'$C/$+-$&,/(.#$"'&3)>$&?$$+&-,+'#1$"$10&&

&

N6&A1$+(#%#-)(#,+&,%&9$;&/+-$"()#+(#$'&&O,/"-$'&,%&/+-$"()#+(;&)+1&'($2'&%,"&#*2",>#+8&(3$&#+%,"*)(#,+&?)'$&3)>$&?$$+&#1$+(#%#$16&&

&&

P6&B''$''*$+(&,%&-,+%#1$+-$&?)'$1&,+&$>#1$+-$&)+1&)8"$$*$+(&&A+&.#83(&,%&(3$&2,($+(#).&/'$'&,%&(3$&#+%,"*)(#,+0&)&-,+%#1$+-$&.$>$.&3)'&?$$+&)''#8+$16&&

&&&

Q6&A+1#-)(#,+&,%&3,G&.#9$.;&#(&#'&(3)(&)+&,/(-,*$&,"&$>$+(&G#..&,--/"&A%&;,/&#+1#-)($&3,G&.#9$.;&)+&$>$+(&#'&(,&,--/"0&(3$&'()+1)"1#R$1&+/*$"#-).&")+8$'&)+1&.#9$.#3,,1&G,"1'&3)>$&?$$+&/'$16&

&&

S6&F")-$)?.$&)--,/+(4&&&&&&

T$'& U,&

O(",+8& O/88$'(#>$&

7/..;& E)"(#)..;&

7/..;& E)"(#)..;&

M#83& 7)#"& V,G&W,1$")($&

XH&#+&5I&Y$";&V#9$.;&

Z5&#+&@&V#9$.;&)'&U,(&

V#*#($1&$[($+(&

W,1$")($& A+-,+-./'#>$&

X@&#+&D&&

V#9$.;&

\5&#+&D&&

]+.#9$.;&

\5&#+&5I&Y$";&&

]+.#9$.;&


!"#$%&#'&()&*&+ (( ,-./0+&('"/12#.3"#4("5(5.'6"74(68.6('"9+%('"#672196&(6"(6824('"#$%&#'&(&*.+9.3"#(

!"#$% &'()*#%+,"-+*.+%/+0'123"0$+-%'$+)(45%673893+%0)7(.+05%.)*0"0'+*'%(+073'05%:+33%-).76+*'+-%1*-%1..+9'+-%6+'$)-05%+'.;<5%$"#$%.)*0+*070%

=+-"76%!"#$% =)-+(1'+%+,"-+*.+%/0+,+(13%0)7(.+05%0)6+%.)*0"0'+*.45%6+'$)-0%,1(4%1*->)(%-).76+*'18)*%3"6"'+-5%+'.;<5%6+-"76%.)*0+*070%

=+-"76%?):% &7##+08,+%+,"-+*.+%/1%@+:%0)7(.+05%3"6"'+-%.)*0"0'+*.45%6)-+30%"*.)693+'+5%6+'$)-0%+6+(#"*#5%+'.;<5%.)69+8*#%0.$))30%)@%'$)7#$'%

?):% A*.)*.370",+%+,"-+*.+%/3"6"'+-%0)7(.+05%+B'(19)318)*05%"*.)*0"0'+*'%C*-"*#05%9))(%-).76+*'18)*%1*->)(%6+'$)-0%*)'%'+0'+-5%+'.;<5%-"01#(++6+*'%)(%31.D%)@%)9"*")*0%16)*#%+B9+('0%

National climate assessment confidence terms

Issues and questions

!   The rate at which new knowledge becomes available? !   The burden on the scientific community? !   Participation from industry? !   Role of an authorizing environment or mandate from governments? !   Adequacy of budgets? !   Communications strategy? !   Link to decision making? !   Analysis and communication of uncertainty?

Source: NRC (2007), Analysis of Global Change Assessments: Lessons Learned 3/30/12 Richard Moss, Joint Global Change Research Institute 7

Theme and topics


!   Theme: Communication requires more engagement with users – there is some progress to report

1.  Scenarios !   New international scenario process !   US National Climate Change

Assessment scenarios 2.  Integrated regional modeling for

adaptation and mitigation !   Stakeholder driven uncertainty

characterization !   IPCC Special Report on Extremes

What are scenarios and why use them?

!   Scenarios are plausible descriptions of how the future might unfold !   Used to gain insight into the

future, not to "predict" it !   Encourage creative thinking !   Inform decisions

!   Scenarios in climate research: !   Establish consistent inputs to

modeling !   Frame uncertainty (including

risks) !   Communicate

Richard Moss, Joint Global Change Research Institute

Time Today Future Horizon

Alt. Futures

Why we can't predict climate change (and why scenarios are important)

!   Human choices are driving change and aren't predictable

!   Scenarios provide "if-then" insights and a basis for projecting change given assumptions

!   Projecting climate change is difficult due to !   Natural variability !   Numerous processes !   Many parameterizations

!   Climate process research and modeling are the foundation for climate projections

!   Social science research provides foundation for emissions scenarios


Cred

it: USG

CRP

Climate change research depends as much on social science as natural science

!   Drivers !   Resource use and

scarcity !   Exposure !   Sensitivity !   Adaptive capacity !   Capacity for

mitigation !   Decision making

under uncertainty and risk management


Historical perspective on emissions scenarios for climate research

!   Early period: instantaneous 2x (or 4x) increases in CO2 concentrations

!   Early 1990s: transient increase (1%/yr) in CO2 !   1990s: increasing complexity of gases and particles

!   SA90 (included policy cases) !   IS92 (multiple realizations of "business as usual")

!   2000: Special Report on Emissions Scenarios (SRES) !   Narratives of socioeconomic futures drive emissions

!   2009: "Parallel" scenario process !   Shorter development time !   Socioeconomic futures explore vulnerability as well as emissions


Scenario types and sequencing in climate change research

Source: Moss et al. 2010

SOCIO-‐

ECONOMIC SCENARIOS

• PopulaJon • GDP • Energy • Industry • TransportaJon • Agriculture •  …

EMISSIONS SCENARIOS

• Greenhouse gases (CO2, CH4, N2O, …) • ParJcles and chemically acJve gases (SO2, BC, OC, CO, NOx, VOCs, NH30) • Land use & land cover

RADIATIVE FORCING SCENARIOS

• Atmospheric concentraJons • Carbon cycle – including ocean and terrestrial fluxes • Atmospheric chemistry

CLIMATE SCENARIOS

• Temperature • PrecipitaJon • Humidity • Soil moisture • Extreme events • …

IMPACT,

ADAPTATION, VULNERABILITY

STUDIES • Coastal zones • Hydrology and water resources • Ecosystems • Food security • Infrastructure • Human health • …


Schematic illustration of SRES logic

Source: IPCC


Source: IPCC

Socioeconomic narratives to radiative forcing

3/30/12 15

Radiative forcing to impacts and responses

Source: IPCC 3/30/12 16

Motivations for a new process •  Address critiques

–  Overconfidence in scenario details –  Long lead times –  Misperceived one-to-one correspondence between

socioeconomic scenarios and climate futures

•  Evolving information needs –  Increasing focus on

adaptation •  Scientific requirements

–  Improve coordination to manage model overlaps

•  IPCC’s new role


Socioeconomic pathways

Vulnerability: exposure, sensiJvity, adapJve

capacity

Emissions drivers, miJgaJve capacity

Radia8ve Forcing:

Representa8ve Concentra8on Pathways

GHGs, other gases, and parJcle concentraJons over Jme; land cover

W/m2 in 2100

Earth System Model Simula8ons

Climate change, climate

variability

Integrated Analyses

MiJgaJon, adaptaJon, impacts

Current task Ongoing (CMIP5)

New scenarios: "Parallel Process"


FOUR RCPs

Source: Moss et al., 2010

Content of RCP Database

!   Data for climate modelers or atmospheric chemists hap://www.iiasa.ac.at/web-‐apps/tnt/RcpDb/

GHG Emissions and Concentrations from IAMs

! Greenhouse gases: CO2, CH4, N2O, CFCs, HFC s, PFCs, SF6 ! Emissions of chemically active gases: CO, NOx, NH4, VOCs ! Derived GHG’s: tropospheric O3

! Emissions of aerosols: SO2, BC, OC ! Land use and land cover [NEW]

! Extension of scenarios to 2300—ECPs.

FORCING AGENTS

EXTENSIONS

! You will not find an integrated set of detailed socioeconomic storylines and scenarios (e.g., no common reference scenario)

WHAT YOU WON’T FIND

Framing: challenge to adaptation and mitigation in "Shared Socioeconomic Pathways" (SSPs)

3/30/12 21 Richard Moss, Joint Global Change Research Institute

Mi8ga8on challenges Adapta8on challenges

Baseline(no-‐policy) emissions MiJgaJon capacity

Exposure SensiJvity

AdapJve capacity

PopulaJon Carbon intensity

Agricultural producJvity Energy intensity

Energy-‐related tech. change CCS availability

… EffecJveness of policy insJtuJons

Energy tech. transfer Diet

Average wealth Extreme poverty Governance

Water availability InnovaJon capacity Coastal populaJon

EducaJonal aaainment UrbanizaJon

… Quality of healthcare

Availability of insurance

Schweizer & O’Neill, in prep.

Alternative futures: population, social progress, and technology uncertainties

!

"#$%&'()!*+&'&#,$'(-,(#-!.)/0! .)/1! .)/23!

4567!!8"9-,&(%&:(;(,<=!!

!

456>!5$'&(;$?!5$@$;)/A$%,!

!

4567!*)%@$%,()%&;!4&'B$,>)'($%,$?!6')C,+!

!

456>!49??;(%D!E+')9D+!

!

4567!F9-,;$!

!

456>!*+&)-!

!

!

!"#$%&'()#*+%(%,*'(,'+-+(#%*#.%/%(0''!1,2%#/'$),3)4++''!5,6'74)(%/%(0'!8%39'%*(:/'()#&4'#*&'2,,$4)#(%,*'''

!;'+9,2<'&4)#%/+'%*%(%#/'$,+%(%=4'()4*&+'!>2,*,?%2@'$,$-/#(%,*@'#*&'4*=%),*?4*(#/'2,//#$+4'!8%39/0'%*4A-#/%(0'.,(9'6%(9%*'#*&'#2),++'2,-*()%4+''

!1-+(#%*4&'+,2%#/'$),3)4++''!"#$%&'?#)<4(!,)%4*(4&'42,*,?%2'3),6(9''!B,&4)#(4'$,$'3),6(9'!C,*=4*(%,*#/'D7,++%/E'7-4/+'&,?%*#(4'''

!1(#3*#(%,*'!1$,)#&%2'42,*,?%2'3),6(9''!;$#(90'#.,-('(94'/4++'7,)(-*#(4'#*&'(94'4*=%),*?4*('!'B%F4&'(429*,/,3%2#/'$),3)4++''

!G)#&%(%,*#/'2-/(-)#/'=#/-4+'#*&'/%74'+(0/4+''!8%39'$,$'3),6(9''!H,,&'42,*,?%2'3),6(9'!>*3%*44)4&'42,+0+(4?+'!1,2%#/'2,94+%,*'''

!B%)4&'%*'$),./4?+'!8%39'$,$-/#(%,*'3),6(9@'/,6'?%3)#(%,*''!'1/,6'42,*,?%2'3),6(9''!C,*=4*(%,*#/'(429*,/,3%4+''!"4+,-)24'2,?$4(%(%,*'!1/,6'&%77-+%,*',7'(429*,/,30''

' Richard Moss, Joint Global Change Research Institute

Application in user-driven impacts research: nested scenarios – working across scales

Finer scale information needed for impacts, adaptation, and vulnerability (IAV) research !   "Downscaling", or !   "Place-based" scenario process

!   Greater credibility, legitimacy, and salience

!   Incorporate local knowledge !   Degree of coupling can vary from

global to local – can be constructed consistently with global scenarios


Households

Livelihoods

Local communities

Governments

Global Markets

Scenarios in US National Climate Assessment

March 30, 2012 25 Richard Moss, Joint Global Change Research Institute

US National Climate Assessment and its use of scenarios

!  Mandate, process, and near-term deliverable !  Long-term goal: establish an ongoing, distributed process !  Uses of scenarios:

!   Provide context of range of potential future conditions !   Establish common assumptions for modeling

!  Types of scenarios: !   Four sets using existing resources based on SRES A2 and B1 scenarios:

!   Climate outlooks, data, and downscaling !   Sea level – core and extended "risk management" ranges !   Land use – demographic allocation using SRES logic !   Socioeconomic – existing Census and modeled data

!   Participatory scenario planning: inventory and pilot studies

26 Richard Moss, Joint Global Change Research Institute

Now for something completely different: Participatory scenario planning

!   Group visioning and planning process !   Systematic and creative evaluation of objectives and implications of

uncertain forces !   Community/user driven

!   Many approaches/methods, but common steps include... 1.  Discuss values and objectives, prioritize issues, and select focus 2.  Identify "drivers" (including uncontrollable external forces) 3.  Analyze potential impacts and risks; test plausibility of ends and means 4.  Assess implications for decision making

!   E.g.s; National Park Service, Western Lands and Communities, Wildlife Conservation Society, Army Corps of Engineers, Tucson Water, ...

27 Richard Moss, Joint Global Change Research Institute

NCA participatory scenario pilot studies: integrating different types of scenarios

Bring climate change scenarios into a participatory scenario planning process !   Participants conduct planning/visioning and then consider

ability to achieve objectives under two futures !   “The Best Chance You’ll Get” – "B1 world": environmental values, rapid

socioeconomic progress, "smart growth", low climate change !   “Big Problems, Low Capacity” – "A2 world": consumerism, slow

socioeconomic progress, sprawling urban development, high climate change

!   In second stage, participants explore adaptation strategies (not just technologies) for A2 conditions

3/30/12 28

NarraJves: Divergent futures -‐ Forcing -‐ Vulnerability

Scenarios: -‐ Socioeconomic -‐ Emissions -‐ Climate -‐ Environmental

Scenarios: Driving forces -‐ Socioeconomic -‐ Emissions -‐ Climate -‐ Environmental

Visioning Scenarios ("creaJng" the future): -‐ Community development -‐ Resource planning -‐ ...

Studies and scenarios: Climate & impacts

Scenarios: TesJng adaptaJon opJons -‐ QuanJtaJve planning methods

NarraJves: -‐ Strategic thinking (e.g., low pro-‐bability, high impacts outcomes) -‐ AdaptaJon planning responses

Embracing Uncertainty

Characterizing Uncertainty

Reducing Uncertainty

Conceptual relationship among different types of scenarios and their uses

Hartman and Moss, in prep.

Global Local Regional


Stakeholder driven uncertainty characterization in regional modeling

March 30, 2012 30 Richard Moss, Joint Global Change Research Institute

The need for transparent evaluation of uncertainty in regional modeling

31 3/30/12 31


iRESM conceptual framework

32

Key Aaributes:   Open source   Flexible and modular   Capable of simulaJng

interacJons and resolving impacts at high resoluJon

  Uncertainty characterizaJon for stakeholder quesJons and issues

iESM project (PNNL, LBNL, ORNL)

iRESM Ini8a8ve

Global Earth System Model

(CESM)

Global Integrated Assessment Model

(GCAM)

•  Socioeconomics •  Energy-‐Economics •  Agriculture & Land Use •  Water

Regionalized Integrated Assessment Model

Regional Earth System Model

•  Atmosphere •  Land •  Ocean •  Biogeochemistry

•  Energy Infrastructure •  Building Energy Demand •  Crop Produc8vity •  Water Supply •  Water Management •  Land Use, Land Cover

Regional Sectoral Models

Climate

Boundary Condi8ons Regional Capability

Data Exchange Climate

Feedbacks

Feedbacks

Legend

Future Development

FY13-‐14 Development

FY10-‐12 Development

Detailed iRESM framework

33

Building Electricity Demands

Ag and LU

R-GCAM Regional

Integrated Assessment

Crop ProducJvity

Sub-‐basin Water Supply

IrrigaJon Supply/Demand

Hydropower & Cooling Water Supply/Demand

Sectoral Water

Supply/ Demand

BEND Building Energy

Demand

EPIC Crop

Productivity

DCLM Distributed Hydrology

DCLM-WM Water

Management

RESM Regional Climate

Building Stock

Climate

Crop Demand

Electric Infrastructure Supply/Demand Climate

Climate

Climate

Climate

TransportaJon & Industrial Electricity Demands

Agricultural Land Cover

Land Use/ Land Cover (All)

Building Energy Demands

REIF Energy

MELD Demand

SITE Infrastructure

EOM Operations

Land Use / Land Cover

•  Wisconsin Bioenergy Initiative •  Wisconsin Climate Change Initiative (represents a

wide range of stakeholders) •  Nelson Institute for Environmental Studies,

University of Wisconsin •  Center for Sustainability and the Global

Environment, University of Wisconsin •  Center for Science, Technology and Public Policy,

Humphrey School of Public Affairs, University of Minnesota

•  Minnesota Forest Resources Council •  Minnesota Pollution Control Agency •  Iowa State University, Climate Science Program,

Agricultural Meteorology •  University of Iowa, Center for Global and Regional

Environmental Research •  Great Lakes Commission •  Midwest Independent System Operators (MISO) •  International Plant Nutrition Institute •  U.S. Department of Agriculture, ARS •  Illinois Department of Agriculture

•  Chesapeake Energy •  Illinois Energy Office, Illinois Department of

Commerce & Economic Opportunity •  Illinois EPA •  City of Chicago Department of Environment •  Great Lakes and St. Lawrence Cities Initiative •  Metropolitan Water Reclamation District of Greater

Chicago •  Pennsylvania State University, several departments

34

Stakeholder organiza8ons met with as of March 2012:

Numerical experiments – stakeholder perspectives

35

Climate Crops/Land Use Energy Water

  Changes in seasonal average temperatures and precipitaJon

  Increased intensity and/or frequency of extreme events (rainfall, drought, heat waves)

  Crop yield   Land use   Water use   Erosion   Soil carbon

and nitrogen   Climate

feedbacks   Emissions   Crop prices   Management

costs

  Energy demand by end use

  Electricity demand by uJlity zone (peak and total annual energy)

  Electricity reserve requirements

  Electricity generaJon mix   Infrastructure expansion

requirements   Electricity prices   Emissions   Fuel prices   Water use   Land use

  Water availability and conflicts between municipal, agricultural, hydropower, and thermo-‐electric cooling needs

Key iRESM model outputs from stakeholder perspec8ve:

Numerical experiments – stakeholder perspectives

Stakeholder-driven uncertainty analysis

3/30/12

Example of decision support process

!   Select a mitigation decision

!   Level/form of renewable portfolio standard? !   Select a single decision criterion

!   e=Electricity price (could be grid operational reliability, ag impacts, etc.)

!   Select model components; assess runtimes; develop surrogates !   Address uncertainties in relevant models contributing to calculation of

costs and grid reliability !   R-GCAM !   BEND !   REIF !   RESM

Building Electricity Demands

Ag and LU

R-GCAM Energy

Economy

BEND Building Energy

Demands

REIF Electric

Infrastructure Expansion

RESM Regional Climate

Building Stock

Climate

Climate

Other Electricity Demands

Building Energy Demands

Regional Electric Capacity Expansion

Feasibility, Cost

iRESM v 0.1

3/30/12 Richard Moss, Joint Global Change Research Institute

Need for development of UC methods for scientific insight and decision support

!   Estimated runtimes for integrated models can be long, with implications for UC

!   A flexible architecture and surrogate models will need to be developed to make UC tractable !   Facilitate coupling appropriate models for the research question at hand !   Based on research question or decision needs, I-O requirements, and

uncertainty source identification !   Develop and use surrogate models as needed to address runtime issues

!   This approach is reflected in the draft USGCRP strategic plan, but agency programs are still driven by a 'bigger and more detailed is better' philosophy

38

The IPCC Special Report on Managing the Risks of Extreme Events and Disasters to Advance

Climate Change Adaptation

sample

Impacts from weather and climate events depend on:

40

nature and severity of event

vulnerability

exposure

Disaster risk management and climate change adaptation can influence the degree to which extreme events translate into impacts and disasters

41

Increasing vulnerability, exposure, or severity and frequency of climate events increases disaster risk

Effective risk management and adaptation are tailored to local and regional needs and circumstances

42

  Changes in climate extremes vary across regions

  Each region has unique vulnerabilities and exposure to hazards

  Effective risk management and adaptation address the factors contributing to exposure and vulnerability

The most effective strategies offer development benefits in the relatively near term and reduce vulnerability over the longer term

There are strategies that can help manage disaster risk now and also help improve people’s livelihoods and well-being

43

IPCC Assessment Reports: The Process

44

Thank you

Richard H. Moss [email protected]


climate science and uncertainty: improving assessments and … · climate science and uncertainty:...

Documents