The Wisconsin Initiative on Climate Change Impacts U.W. Program on Climate Change October 20, 2009 Daniel J. Vimont University of Wisconsin - Madison Nelson.

The Wisconsin Initiative on Climate Change Impacts

U.W. Program on Climate Change October 20, 2009

Daniel J. VimontUniversity of Wisconsin - Madison

Nelson Institute for Environmental StudiesCenter for Climatic ResearchUniversity of Wisconsin - MadisonUnderstanding Earth’s Past, Present, and Future

Atmospheric and Oceanic SciencesUniversity of Wisconsin - Madison

Center for Sustainability and the Global Environment (SAGE)University of Wisconsin, Madison

The Wisconsin Initiative on Climate Change Impacts

Daniel J. VimontUniversity of Wisconsin - Madison

Thanks to:Steve Vavrus, David Lorenz, Michael Notaro (CCR)

Chris Kucharik (SAGE)Jack Williams (UW Geography)

Wisconsin State Climatology OfficeWisconsin Initiative on Climate Change Impacts (WICCI)

Wisconsin Focus on Energy Program

Nelson Institute for Environmental StudiesCenter for Climatic ResearchUniversity of Wisconsin - MadisonUnderstanding Earth’s Past, Present, and Future

Atmospheric and Oceanic SciencesUniversity of Wisconsin - Madison

Outline

• IntroductionA history of the Wisconsin Initiative on Climate Change Impacts (or WICCI)

• WICCI Organizational StructureWICCI as a boundary organization

• Downscaling Methodology and ResultsDownscaling for Climate Impacts Assessment

• Conclusions

Outline

• IntroductionA history of the Wisconsin Initiative on Climate Change Impacts (or WICCI)

• WICCI Organizational StructureWICCI as a boundary organization

• Downscaling Methodology and ResultsDownscaling for Climate Impacts Assessment

• Conclusions

Wisconsin Initiative on Climate Change Impacts

WICCI: Partnership between the UW Nelson Institute for Environmental Studies, the Wisconsin DNR, and other state groups

Goal: Assess and anticipate climate change impacts on specific Wisconsin natural resources, ecosystems and regions; evaluate potential effects on industry, agriculture, tourism, and other human activities; and develop and recommend adaptation strategies…

http://www.wicci.wisc.edu

Wisconsin Initiative on Climate Change Impacts

History:• June, 2007: Initial meeting between

DNR and UW. Outlined idea.• Summer, 2007: Organizational

structure outlined• Fall 2007 – Present: Science Council

formed and operational• Fall 2007: First working group(s)

formed. Climate Working Group• Spring / Summer, 2008: First major

funding: FOE downscaling (~$183K)• Jan 12, 2009: First WG meeting

(>100 participants)• Feb 2, 2009: Advisory Council formed

http://www.wicci.wisc.edu

Wisconsin Initiative on Climate Change Impacts

History:• Spring, 2009: “Bracing for Impact”

series of lectures for public and WPT• August, 2009: Submitted proposal

for Great Lakes RISA• Sept. 2, 2009: Brief the Secretary of

the DNR, suggested coordination with the Governor’s staff.

• Sept. 14, 2009: Second Advisory Council Meeting (downscaling results released)

• Sept. 15, 2009: Public (media) release of downscaling results

• Sept. 21, 2009: Second working group meeting (~150 participants)

• Fall, 2010: First Assessment Reporthttp://www.wicci.wisc.edu

Outline

• IntroductionA history of the Wisconsin Initiative on Climate Change Impacts (or WICCI)

• WICCI Organizational StructureWICCI as a boundary organization

• Downscaling Methodology and ResultsDownscaling for Climate Impacts Assessment

• Conclusions

WICCI Organizational Structure

Needs in Climate Impact Assessment• Broad expertise across disciplines• Stakeholder engagement• Science and policy representation and

expertise• Legitimacy within and across scientific and

policy communities• Public engagement and support

WICCI Organizational Structure

Boundary Objects: “Objects” that sit between social worlds, like science

and nonscience…. they can be used by individuals within each for specific purposes without losing their own identity

(from Guston, 2001)

WICCI Organizational Structure

Boundary Organizations: • Provide the opportunity and sometimes the

incentives for the creation and use of boundary objects

• They involve the participation of actors from both sides of the boundary, as well as professionals who serve a mediating role

• Exist at the frontier of the two relatively different social worlds of politics and science, but they have distinct lines of accountability to each

(from Guston, 2001)

WICCI Organizational Structure

WICCI as a Boundary Organization

Physical Climate

Impact Systems

Policy

WICCI

WICCI Organizational Structure

Science Council: ~20 members representing a variety of expertise in Wisconsin. Primary function is to organize and coordinate Working Groups that have the scientific expertise to assess climate change impacts pertinent to specific issues or areas of concern.

WICCI Organizational Structure

Advisory Committee: Representatives of business interests, non-governmental organizations, municipalities, agencies, state and local government, and other stakeholders. Advises WICCI, and provides engagement / link to stakeholders.

WICCI Organizational Structure

Operations and Outreach: Provides logistical support to the Science Council and performs outreach functions related to the mission of WICCI (e.g. media release, public lectures, etc.)

WICCI Organizational Structure

Working Groups: created by the Science Council to conduct science-based assessments of climate change impacts pertaining to specific topics or areas of concern and to make recommendations on adaptation strategies

WICCI Working GroupsWater

Resources

SoilConservation

Agriculture

Adaptation

Plants & NaturalCommunities

Central SandsHydrology Forestry

CoastalCommunities

Green Bay

Wildlife

Stormwater

Coldwater Fish

Milwaukee

HumanHealth

WisconsinClimate

WICCI Future DirectionsAssessment Reports: First Assessment Report to be completed Fall, 2010. This will include an assessment of physical climate change in Wisconsin and specific vulnerabilities.

Funding: Some sort of steady funding is needed. We have applied for a Great Lakes RISA, and will be submitting a proposal for a USGS Midwest Regional Center.

Public / Political Connections: We have launched a (brief) media campaign, and will continue public lecture series. We also are actively travelling around the state to give talks, including to political groups (e.g. the Public

Outline

• IntroductionA history of the Wisconsin Initiative on Climate Change Impacts (or WICCI)

• WICCI Organizational StructureWICCI as a boundary organization

• Downscaling Methodology and ResultsDownscaling for Climate Impacts Assessment

• Conclusions

Needs for Downscaled Data

Characterize UncertaintyUncertainty from: large-scale model physics, emissions scenario, transition from large to small scale, additional uncertainty (from subjective assessment)

High resolution (spatial and temporal)8-10km resolution, daily time scale

Need to represent extremesExtreme precipitation is necessary for hydrology; extreme temperature for human health / forestry / others

FLEXIBILITY!!!Numerous potential applications, so flexibility is needed!

Global Climate Change Moving from Global to Regional

Downscaling: Interpret global projections on a scale relevant to climate impacts.

WICCI Climate Working Group / Focus on Energy

Problems with simple interpolation:

Global to Local Climate Change Moving from Global to Regional

Downscaling Method:

Downscale Probability Distribution, instead of actual variable.

Advantages:

• PDF is large-scale, so method is “more true” to technique• Extreme events are better characterized• PDFs are more flexible – allows a variety of applications

Work by David Lorenz - WICCI Climate Working Group / Focus on Energy

Downscaling Precipitation and Temperature

Downscaling Precipitation and Temperature

• The large-scale predictors do not contain all the information we need to know to predict the precipitation (P) and temperature (T) at a particular point.

• This uncertainty in predicting P and T has important implications for generating downscaled P and T with realistic variance and extremes.

• We must predict more the just the most likely value for P and T given the large scale fields, but also the distribution of the errors from this value.

• The downscaled P and T are the sum of 1) the most likely value and 2) a random number generated from the distribution of the errors. Both components are required to give realistic variance and extremes.

Downscaling Precipitation and Temperature

• Train the downscaling on station data (COOP stations) and NCEP reanalysis precipitation (it’s like a GCM)

• Debias daily CDF of large-scale predictors from each global climate model to NCEP CDF

• Use downscaling relationships from observations on global climate models. Estimate parameters of sub-gridscale distribution at each station location.

• Interpolate distribution parameters (not actual data) to fine-scale grid -

• Final product: daily varying probability distribution on a high-resolution grid, for each climate model, and for each emissions scenario.

Temperature:

Temperature is downscaled using a standard normal distribution:

Temperature:

Temperature is downscaled using a standard normal distribution (valid because residuals are normal):

Large Scale

Small Scale (raw)

Precipitation:

Two steps: 1) Bernoulli distribution for rain / no rain. 2) Generalized gamma distribution for rain amount.

The histogram of precipitation amount when the large-scale predictor is in a) the 25th to 27.5th percentile and b) the 97.5th to 100th percentile

Precipitation:

Gamma:

Generalized Gamma:

How does it perform?

Ways to use the data:1. Classic Risk Assessment

Use actual probability distributions to identify Risk as the product of probability and consequence

2. Spatio-temporal DataGenerate spatial data using a “weather generator” type noise pattern.

3. Historical RescalingRescale an existing time series from a present-day PDF to a future PDF.

Actual Probability Distributions

Adaption

Climate Space

Pro

bab

ilit

y

Present Climate

Predicted Climate

Impact threshold

Climate Space

Pro

bab

ilit

y

Risk Assessment:• Identify threshold /

response surface• Define present day risk

with present day probability distribution

• Compare future risk with future probability distribution

• Explore how adaptation strategies can impact risk

Intense Precipitation Events

Global Climate Change

Downscaling: Focus global projections to a scale relevant to climate impacts.

WICCI Climate Working Group / Focus on Energy

Thanks to D. Lorenz

Temporal Correlation between Stations• Expect downscaled station data to be correlated with each other

because a portion of the stations’ variability is controlled by the large-scale.

• However, if the spatial scale of the "random component" of the P (or T) variability is larger the separation between stations, then one expects the downscaled P (or T) to under-estimate the correlation between stations.

Temporal Correlation between StationsTo remedy this situation, the random numbers used to generate the precipitation at the different stations are not independent but instead are correlated with each other. Let R be a nstat X ntime matrix of independent random numbers used to generate the P occurrence. The new R to generate the P occurrence is: W•R, where W is a nstat X nstat matrix of weights.

Spatial and / or temporal data

Annual Temperature Change

Winter Temperature Change

>90° Days, and <0° Nights

Winter Precipitation Change

Rescale a historical time series

MaxT (e.g.)

Pro

bab

ilit

y

Present Climate

MaxT (e.g.)

Pro

bab

ilit

y

Why to use this approach:• You’ve already done some

analysis with historical weather data

• Impact is “event-like”• Covariates are important

(e.g. warm, wet, and windy on a given day)

• Policy decisions can be compared to historical decisions

Outline

• IntroductionA history of the Wisconsin Initiative on Climate Change Impacts (or WICCI)

• WICCI Organizational StructureWICCI as a boundary organization

• Downscaling Methodology and ResultsDownscaling for Climate Impacts Assessment

• Conclusions

Climate Change Impacts in Wisconsin

The Wisconsin Initiative on Climate Change Impacts (WICCI)WICCI is set up as a boundary organization that includes climate sciences, impact sciences, and policy makers.

The organizational structure engages stakeholders and climate / impact scientists alike, while preserving the identities of each group.

Working groups allow focused and efficient efforts at understanding specific impacted systems

The Operations and Outreach arm actively engages the public, and works to build support for the group in new and existing stakeholder communities.

Climate Change Impacts in Wisconsin

The Wisconsin Initiative on Climate Change Impacts (WICCI)Future directions:

Base funding is needed as interest in the group snowballs.

Assessment reports to be completed annually.

Continued public outreach and engagement needed to build support at the state and local levels.

Climate Change Impacts in Wisconsin

Downscaling Climate over WisconsinDownscaled projections of precipitation and maximum and minimum temperature for Wisconsin have been completed.

The downscaling methodology predicts the (daily) probability distribution for a specific station based on large scale inputs.

The advantages of the downscaling technique include

(a) it works well

(b) interpolation of distribution parameters avoids bias in extremes or discrete events

(c) uncertainty is characterized across various dimensions

(d) the resulting data are very flexible

Resources:Wisconsin Initiative on Climate Change Impactshttp://www.wicci.wisc.edu

Governor’s Task Force on Global Warminghttp://dnr.wi.gov/environmentprotect/gtfgw/

UW Atmospheric and Oceanic Scienceshttp://www.aos.wisc.edu

Nelson Institute for Environmental Studieshttp://www.nelson.wisc.edu

Center for Climatic Researchhttp://ccr.aos.wisc.edu

Center for Sustainability and the Global Environmenthttp://www.sage.wisc.edu

Intergovernmental Panel on Climate Changehttp://www.ipcc.ch