Transcript Communicating Uncertainty PPT Slides

Global Climate Change
and Uncertainty
David B. MacNeill
Fisheries Specialist
NY Sea Grant Extension
SUNY Oswego
[email protected]
Global Climate Change and Uncertainty
Apocalypse
Climate models
Decision-making
This Presentation:
• Broad-brush overview of climate change
uncertainties, communication etc. from
literature sources, extension experience
with scientific uncertainty.
• Not an indictment of science or an
admonishment of scientists, policy makers,
government or the lay community!!
Understanding the concepts of risk and
uncertainty with a deck of cards??
The uncertainty:
What poker hand will
I draw next?
The Dead Man’s Hand:
unlucky for Wild Bill Hickok?
The risk: What is the probability
of drawing it? (<1%)
But, the card deck changes unexpectedly……
Death
cards
Other
cards
The Risk ?
Some Climate Change Perspectives
• A complex, multidisciplinary issue of long-
•
term global consequence, that demands:
– Best available information
– New assessment, predictive, decision-making
tools
– A carefully planned extension/outreach strategy
– Better PR for science
An opportunity to:
– Inform communities: climate science, risks,
abatement and science 101
– Assist coastal communities: decision-making
Global Climate Model
Climate Change Complexity:
• Many different disciplines.
• Highly uncertain events; outcomes poorly defined.
• Interactive anthropogenic and natural events.
• Future outcomes sensitive to small changes in current
conditions.
• Incomplete understanding of climate system.
• Imprecise models: feedbacks, interactions, parameter
values.
• Huge jigsaw puzzle having 10s of thousands of pieces.
• Compilation: decades of intensive, international research.
Uncertainty leads to those nagging questions
Is climate change real?, are humans responsible?
• What are the impacts?, What should we do?
• Why:
–
–
–
–
is science uncertain?
do scientists disagree? change their minds?
don’t scientists always have the answer?
do results contradict?
Uncertainty paradigms
•Uncertainty is unwelcome, and needs to be avoided. Science
must eliminate uncertainty through more and better research.
•Uncertainty is undesirable, but unavoidable.
Science must estimate and quantify uncertainty as well as
possible.
•Uncertainty creates opportunities.
Science must contribute to more inclusive, understandable
discussions.
•Uncertainty is an integral part of decision-making. Science
must have more societal influence.
Communicating Science and Uncertainties
Why even bother ???
• PR: The process of science.
• Restore credibility of science: increased
•
•
•
•
•
transparency.
Provide accessible information/knowledge to
decision-makers.
Decision-making: accurate and collaborative.
Increase public support/involvement: decisionmaking
Enhance societal abilities: adaptation & mitigation
GCC interactions: science and human ecology
Three Arguments for Climate Change
• Climate is changing: analyses of many indicators
• Human activities have contributed to increases in
•
•
green house gas emissions
Scientific deliberations and large-scale computer
models suggest potential for climate change from
anthropogenic influences
High degree of confidence: weight of evidence from
expert opinion
Is climate really changing?
Climate proxies
Convincing
evidence
BUT..
Contentious
Points
Natural vs.
anthropogenic
Sea level
Seeing is Believing?
Muir Glacier Alaska, August 2004. photo by B.F. Molnia
Muir Glacier Alaska, August 1940. photo by W.O. Field
An exaggerated view…..
“You just don’t understand.”
“It’s too complicated”.
“We know what is best.”
“It’s not our job to explain it to you”.
“We’re scientists, not interpreters”.
Scientist
“Science is sloppy - a collection of
useless facts”.
“You’re arrogant, out-of touch and
have impractical ideas”.
“You’ve been wrong before.”
“Prove it.”
Non-scientist
Uncertainty
Some major challenges
• Continuing uncertainties on climate system sensitivity to various
•
•
•
•
•
•
feedbacks (e.g., clouds, water vapor, snow).
Several natural modes of climate variability have been identified and
described, but their predictability is uncertain.
Need to improve understanding of whether and how human impacts
may alter natural climate variability.
Do not yet have confident assessments of the likelihood of abrupt
climate changes.
Insufficient understanding of effects of climate variability and change
on extreme events.
Limited capabilities at regional scales.
Need better means for identifying, developing, and providing climate
information required for policy and resource management decisions.
Mac’s Uncertainty Concept Model
Stochastic
(Surprises)
Climate System
Epistemic
(Unknowns)
Scientists
Science
communication
(translation)
Knowledge
Knowledge
Non-Scientists
Human reflexive
(volition)
Decisions
Mac’s Uncertainty Concept Model
Stochastic
(Surprises)
Climate System
Epistemic
(Unknowns)
Scientists
Science
communication
(translation)
Knowledge
Knowledge
Non-Scientists
Human reflexive
(volition)
Decisions
Mac’s Uncertainty Concept Model
Surprises
Climate System
Unknowns
Scientists
Science
communication
(translation)
Knowledge
Non-Scientists
Decisions
Knowledge
Human reflexive
(volition)
Different Roles of Science in GCC Policy
Pure
scientist
interpretation
Politicians
Science
arbiter
Scientific
Knowledge
Policy
makers
Decision
making
Policy
Honest
broker
Issue
opinions
advocate
Advocacy
Roger Pielke Jr.
How does science work, anyway?
Susan Haack
Addressing uncertainties
•
•
•
•
Identify
Characterize: source, magnitude
Solicit expert judgments: level of “confidence”
Sensitivity analysis: range of probable model outcomes assessed
with model using a range of values various inputs, upper and lower
bound
• Quantify: probabilistic analysis (Frequentist and Bayesian),
probabilistic distributions, deterministic analysis and hybrids
• Clarify, document range and distributions
• Articulate and communicate: probabilistic and scenarios
Some predicted impacts of climate change?
In-direct
Direct
 Droughts, crop loss, famine
 Invasive species, new or re-
 Warmer, dryer summers
 Warmer, wetter winters
 Increased spring
flooding
 Changes in sea/lake
levels, water currents,
thermal structure
 Increased storm
frequency, severity




emerging pathogens, parasites
More hyperthermia deaths
Coastal infrastructure/tourism
Habitat damage/loss
Loss of biodiversity, extinctions?
 Technological advances
 Longer growing seasons
 New agriculture/tourism



opportunities.
More snow?
Reduced heating costs
Fewer hypothermia deaths
GCC heretics, infidels, skeptics, nay-sayers, cynics, deniers??
What are they really saying?
•
•
•
•
•
•
•
•
•
Nature: too complex.
Conflicting data.
Models: poor predictors.
Exaggerated impacts.
Doom/gloom vs. facts.
Earth’s resiliency.
Strategies: cost/benefits?
Consensus: evidence supports GCC
Less consensus: drivers, impacts,
strategies, policies
What is the matter with science?
The debate continues……
• Dyson (1993)
– Consensus: peer pressure (entrepreneurial science) vs. debate
– Public fear drives funding priorities = politicization of science
– Science's failure to address global welfare vs. unrealistic expectations
• Rubin (2001)
–
–
–
–
–
Science is not the sole repository of the truth
Little self-limitation on deliverable truths
Get the facts straight vs. overselling science
Scientific authority fosters hidden agendas that short-circuit debate
Participatory decision making impeded by science education shortfalls
• Commoner (1971)
– Illusion of scientific objectivity
• Grant et al. (2004)
– Popper’s vs. psychological v
– Benedikter (2004) basic ideologies and mechanisms not fully visible (psychologically)
• Malnes (2006)
– Mixed messages: duplicity vs. extraneous diversions
Classical, Modern & Post-Normal Science
Classical:
•Observations
•Sense experiments
•Subjective judgments
•Past experience
Modern / Normal:
•Exclusive, remote
•Non-interdisciplinary
•Experiments/models
•Data analysis/interpretation
•Hypothesis testing
the Truth!
•predictions
•probabilities
•possible explanations
•disconnected policy
•adversarial
•communication gaps
Absolute
Reductionist,
“puzzle-solving”
“Post-Normal”
•Inclusive
•Natural & social sciences
•Complexity/risk/urgency
•Systems approach
•Cost/benefits
•Public debate
•shared decision making
•problems solving
•confidence/trust building
•Anti-science perception
Precautionary,
risk management
Classical, Modern & Post-Normal Science
Classical:
•Observations
•Sense experiments
•Subjective judgments
•Past experience
Modern / Normal:
•Exclusive, remote
•Non-interdisciplinary
•Experiments/models
•Data analysis/interpretation
•Hypothesis testing
the Truth!
•predictions
•probabilities
•possible explanations
•disconnected policy
•adversarial
•communication gaps
Absolute
Reductionist,
“puzzle-solving”
“Post-Normal”
•Inclusive
•Natural & social sciences
•Complexity/risk/urgency
•Systems approach
•Cost/benefits
•Public debate
•shared decision making
•problems solving
•confidence/trust building
•Anti-science perception
Precautionary,
risk management
Perceptions of Science
God-like?
Elitist?
Crusading knight?
Mad/evil?
Two Opposing Metaphors for Science:
God-like or Golem?
• “Ultimate source of knowledge/wisdom.
• Operates in unencumbered, controlled
environment.
• Strives for perfection.
• Accountable, held to high standard.
Truth
• A creature of our own design, neither good or
•
•
•
•
bad.
Powerful, protective, follows orders.
Clumsy and dangerous, must be controlled.
Fallible = low expectations.
Can’t be blamed for mistakes if it is trying.
“Other”
Uncertainties
Climate Science
Uncertainties
The Snowball Effect
Cascading Uncertainties in Climate Science
Adapted from Schneider 1983
Emission
scenarios
Carbon cycle
response
Global
climate
sensitivity
Regional
climate change
scenarios
Range of
possible
impacts
Scientists face important challenges in communicating
science to non-scientists
• The nature of ‘normal’ scientific investigation and debate
– logic vs. cognitive processes
– adversarial, not focused on consensus development
– debate primarily within disciplines
• Isolationism
– “too busy” to talk to non-scientists!
– rift between physical and social scientists
• Inadequate training in communication skills
– dealing with media
– addressing misinformation
– understanding policy development process
Can complex science be understood by the
public?
• Yes, many successful examples !
• Knowledge from Scientific process
• “Step-back”, discuss and debunk science myths
– Myth 1: science as a collection of established facts
– Myth 2: conflicting science presented in a balanced
way
– Myth 3: science jargon as chief obstacle
Interpretations of Global Climate Science
Uncertainties
• Scientists:




intrinsic part of science
too many variables to eliminate
can be reduced with more scientific information
general support of a “precautionary” approach”.
• Policymakers:




science is sloppy
“burden of proof”
lack of/incomplete knowledge = bad science
must have all the facts: decision making/policy
implementation
 little/no support of precautionary steps
The Climate Uncertainty “Toolbox”
Communicating Uncertainties of Climate Change
•
•
•
•
•
•
•
•
Increase science literacy
Outreach materials: Hypothetical scientific investigations.
Develop vivid narratives of potential harm
Address/communicate uncertainties to stakeholder communities.
Understand decision making mechanics, assess values and attitudes
Develop an integrative (social-natural science), participatory decisionmaking process
Psychometric paradigm: people (focus on a range of qualitatively
distinctive factors that are irreducible by numbers) show a richer
rationality than experts (focus on quantity), risk perception in social
sciences, used to explain divergence between risk related judgments
People influenced by whether risk is catastrophic , future generations,
involuntary incurred, , uncontrollable, delayed vs immediate, and
particularly dreaded.
Cass Sustein 2007: Columbia Law Review 107: 503-557
What are the likely climate changes over the
next century, or so??
• Most global warming projections are for a 4-10 F
increase by 2100
• Virtually certain: ~ 95 to 100%
– Warmer days and nights, fewer cold periods over most land areas
• Very likely: ~ 67-95%
– Warm spells/heat waves, frequency increasing over most land
areas
– Heavy and more frequent precipitation events
• Likely: ~ 33-67%
– Area affected by drought increases
– Intense tropical cyclone activity increases
– Increased incidence of extreme high sea level (exclude tsunamis)
Communicating Uncertainty:
Examples from Weather Forecasts
• Numerical probabilities:
– A 30 % chance of rain.
• Qualitative or categorical forecasts:
– Today’s weather will be “fine”.
Handmer et al. 2007
Communicating Uncertainty:
Examples from Weather Forecasts
• Numerical probabilities:
– high likelihood, tangible events
– can be misinterpreted: where? when? how long?
– example: 30% chance of rain
• a 30% chance of rain in the forecast area.
• a 30% chance of rain at a specific location in forecast area.
• only 30% of the forecast area will be affected, if it does rain.
• it will rain 30% of the day.
• it will rain 3 out of 10 days when rain is forecasted
– not useful when:
i.e. 0.0001% chance of as a severe event
• Abstract, “invisible”, even catastrophic events
• Public more concerned with issues of control, trust and equity
•
Handmer et al. 2007
Decision-making Under Uncertainty
Decisions:
• based on likelihood of uncertain events
– Uncertainties expressed
• numerical form (odds)
• subjective probabilistic statements
• heuristics
– Representativeness – degree of relationship, causality
– Availability – ease of instances/consequences imagined
– Adjustment/Anchoring –initial value adjusted to yield final answer
(problem formulation or partial computation)
Tversky, A. and D. Kahneeman. (1974). Judgment under Uncertainty: Heuristics and
Biases. Science 185: 1124-1131
Decision-making Under Uncertainty
• Task of choice
– Framing
• Relate decision making to similar problems
• Used to determine outcome loss or gains
– Evaluation
• Act to reduce loss probability, maximize gains
• Adopt risk averse stance
• 3 subconscious processes (heuristics):
– Representativeness – degree of relationship, causality
– Availability – ease of instances/consequences imagined
– Adjustment/Anchoring –initial value adjusted to yield final answer
(problem formulation or partial computation)
Patt, A. and S. Dessai. (2005). Communicating Uncertainty: lessons learned and suggestions for climate change
assessment. C. R. Geoscience 337: 425-441
Tversky, A. and D. Kahneeman. (1974). Judgment under Uncertainty: Heuristics and Biases. Science 185: 1124-1131
Decision-making Under Uncertainty
• Stochastic uncertainties (unpredictability/surprises)
–
–
–
–
Framing: (usually) in frequentist terms
Uncertainty: probability expressed relative frequencies
Heuristic: Availability = analogy
Evaluation: Less risk averse, under-estimate risk, less prone to
illogical choice
• Epistemic uncertainties (structural/ignorance)
– Framing (often) in Bayesian terms
– Uncertainties: ambiguous probability estimates, numerical ranges
confidence, expert opinion
– Heuristic: Representativeness = common, familiarity
– Evaluation: More risk averse, over-estimate risk, more prone to logic
errors
Patt, A. and S. Dessai. (2005). Communicating Uncertainty: lessons learned and suggestions for climate change
assessment. C. R. Geoscience 337: 425-441
Tversky, A. and D. Kahneeman. (1974). Judgment under Uncertainty: Heuristics and Biases.
1131
Science 185: 1124-
Decision-making Under Uncertainty
Decisions:
• based on likelihood of uncertain events
– Uncertainties expressed
• numerical form (odds)
• subjective probabilistic statements
• heuristics
– Representativeness – degree of relationship, causality
– Availability – ease of instances/consequences imagined
– Adjustment/Anchoring –initial value adjusted to yield final answer
(problem formulation or partial computation)
Tversky, A. and D. Kahneeman. (1974). Judgment under Uncertainty: Heuristics and
Biases. Science 185: 1124-1131
*
*
9 graphical representations of the same snow fall predictions
Communicating Uncertainty:
Examples from Weather Forecasts
• Qualitative or categorical forecasts:
– “Fine”
– Also misinterpreted: does it mean
•
•
•
•
•
•
•
No rain?
Sunny/sunshine?
Not too hot/moderate temperature?
Clear day/ not cloudy or overcast?
Lovely weather/a nice day?
No wind/light winds?
Some cloud/may be overcast?
Handmer et al. 2007
Communicating Uncertainty:
When Uncertainties are Insurmountable
• Scenarios
– Coherent, plausible, alternative representations of future climate
– Projections/modeled responses (not forecasts) from climate
“drivers”.
– Descriptions: current states, drivers, step-wise changes, future
images.
– Assessments future climate conditions (very high uncertainties).
– Assist in designing adaptation/mitigation strategies
– Provide better understanding of interactions/dynamics
Outreach: Uncertainties of Climate Change
•
•
•
•
•
•
•
Increase science literacy
vivid narratives of potential harm/benefits
Communicate uncertainties to stakeholder communities.
Assess values and attitudes
Develop an integrative (social-natural science) decision-making
process
Psychometric paradigm: people (focus on a range of qualitatively
distinctive factors that are irreducible by numbers) show a richer
rationality than experts (focus on quantity), risk perception in social
sciences, used to explain divergence between risk related judgments
People influenced by whether risk is catastrophic , future generations,
involuntary incurred, , uncontrollable, delayed vs immediate, and
particularly dreaded.
An Interesting Expert Opinion:
An Essay: Divergent American Reactions to Terrorism and Climate Change
Cass Sustein 2007: Columbia Law Review 107: 503-557
Similarities: potentially catastrophic outcomes, difficulty assigning probabilities to risks
Divergence: simple facts and political responses to each risk:
•
•
•
•
•
•
•
•
Terrorism:
low probability, palpable, catastrophic
risks are immediate, short term
Concern to US, Britain an allies.
Perceived high risk recurrence
neglect probability visual anger, fear,
Huge costs justified to protect national
security benefits unimportant
2005-2006: $255 $318 billion committed
to war on terror vs $312 billion for entire
Kyoto protocol.
Public opinion
–
–
2004 48% Britons: top global priority
2006 80% Americans top global priority
•
•
•
•
•
•
•
•
Climate change:
high probability, impalpable, catastrophic
Long-term risk, affect future generations.
Concern to other nations only
serious mitigative/adaptive action unlikley
climate change causes obscure
(uncertainties)
people lack experience make risks
apparent, real or impending,
cost benefits,
Public opinion
–
–
2000 CC: ranked environment as 16th most
important issue and 12th out of 13 top
environmental problems
2004: 63% Britons: top global
environmental issue.
An Interesting Expert Opinion:
An Essay: Divergent American Reactions to Terrorism and Climate Change
Cass Sustein 2007: Columbia Law Review 107: 503-557
“We have to deal with this new type of threat [terrorism] in a
new way we haven’t yet defined.. With a low-probability, high
impact event like this.. if there is a 1% chance that Pakistani
nuclear scientists are helping Al Qaeda build or develop a
nuclear weapon, we have to treat it as a certainty in terms of
our response”
-- Dick Cheney, Former Vice-President
“Climate change is the most severe problem we are facing
today - more serious than the threat from terrorism”
– Sir David King Director, Smith School of Environment, Oxford; Research Director,
Dept. of Physical Chemistry, Cambridge; Former Chief Scientific Advisor to Blair
Administration.
Epilogue
“Any philosophy that in its quest
for certainty ignores the reality
of the uncertain in the ongoing
processes of nature, denies the
conditions out of which it
arises.”
John Dewey, The Quest for Certainty, 1929
And now, the punch line(s)……
• Climate change uncertainties: tremendous outreach challenges
• Uncertainties are cumulative: science to policy
• Climate change predictions: probabilistic context where
possible.
• Scenarios: address insurmountable uncertainties.
• Integrative natural and social science approach to decisionmaking.
• Outreach: science mechanics, sources of uncertainty, restore
faith in science, assess/understand heuristics, facilitate
improved decision-making, craft a responsible, informative and
useful message.