Transcript winkler

Current Research
Julie Winkler
Department of Geography
April 5, 2010
Research Interests
• Synoptic climatology
• Impacts of climate variability and change
Current Funded Research
•
“A row crop ecosystems in a changing climate: Enhancing ecosystem services at field
farm and watershed scales”
–
–
–
–
•
•
•
USDA/EPA
A. Kravchenko, S. Snapp, A. Grandy, J. Winkler, and J. Andresen.
$475,400
March 1, 2010 – February 28, 2014
“Towards an Integrated Framework for Climate Change Impact Assessments for
International Market Systems with Long-Term Investments”
–
NSF, Dynamics of Coupled Natural and Human Systems Program
–
J. Winkler, S. Thornsbury, P.-N. Tan, J. Andresen, J.R. Black, S. Loveridge, S. Zhong, J. Zhao, N. Rothwell, and A. Iezzoni
–
$1,499,763
–
October 1, 2009 – September 30, 2014
“Toward an Improved Understanding of the Characteristics, Processes, and Impacts of
Northerly and Southerly Low-Level Jets in the Central United States”
–
NSF
–
J.A. Winkler and S. Zhong (Michigan State University); C. Walters (University of Michigan-Dearborn)
–
$421,610 (MSU portion )
–
September 1, 2009 – February 28, 2013
“Great Lakes Regional Integrated Sciences and Assessments Center”
–
NOAA
–
–
–
D. Scavia, T. Dietz, J. Andresen, M. Lemos, R. Rood, J. Winkler, M. Huntley, C. Pistis, and M. Staton
$3,495,180
Beginning October, 2010
Traditional Climate Change Assessments
(following Carter et al. 2007)
Single Location or Region
Isolated time slice
Specific System, Process or Industry
• Local/regional in scale
• Isolated time slice(s)
– assessments for different time slices are
not informed by earlier time periods
• Focus on a specific system, process, or
industry
• Local/regional climate projections
downscaled from simulations from
global climate models
– Also referred to as a “top-down”
approach
Static Modeling
Output is an assessment of potential
impacts for a SYSTEM or ACTIVITY
for a LOCATION/REGION
• Static modeling
– often used a series of linked models
– “feed forward” approach to downstream
models without interactions and
feedbacks
• Spatial interactions and
interdependencies are not considered
Comprehensive Integrated Assessments
• Global viewpoint
• Sectoral or cross-sectoral interactions
• Often use dynamic modeling
– Complex integrated models
– Include system components and
feedbacks
– Continuously running models
– Examples include IMAGE, DICE,
PAGE, etc.
• Limitations
– often not fully integrative across all
aspects of a system
– relatively simple characterizations for
some if not all of the system
components
Global
Sector (Multiple Industries)
Dynamic Modeling
Continuously running models
Output is an assessment of
potential impacts for a
SECTOR or multiple sectors
at the GLOBAL scale
What is missing?
• Assessment methods for sub-sectors (i.e., specific
industry) with international markets
• These assessments require:
– a broader spatial perspective than a traditional
assessment
– more detail than the “broad brush” approach of a
comprehensive integrated assessment
– greater incorporation of temporal dynamics
• changing patterns of international trade, consumption and
production
• adaptation (spatially differentiated between production
regions) by stakeholders groups within the industry
Approach Used in Pileus Project
Single Location or Region
Specific System, Process or Industry
Static Modeling
Isolated time slice(s)
Output is an assessment of potential
impacts for a SYSTEM or ACTIVITY
for a LOCATION/REGION
•
•
•
•
Michigan
Tart cherry industry
End-to-end assessment
Early, mid, and late century
time slices
• No consideration of
– Climate impacts on tart
cherry production outside of
Michigan
– Adaptation
End-to-End Assessment Approach
Climate
Observations
or
Scenarios
Industry,
Ecological
or Activity
Model
Economic
Model(s)
Risk Management Decision Making Tools
Uncertainty Evaluation limited to
Climate Scenario Ensembles
• Many sources of
uncertainty need to be
considered
• Ensemble approach
where multiple
scenarios are used to
estimate the
“quantifiable range of
uncertainty”.
Source: IPCC, 2001
Beyond Pileus
• Impetus came from stakeholders
– 2002 freeze event “opened the door” to foreign imports
• Develop a conceptual framework for climate change
assessments for international market systems
– Emphasis on “tractable”
• data requirements reasonable/obtainable
• methods for temporal and spatial scaling (both upscaling and
downscaling) transferable to multiple regions
• procedures for evaluating the sensitivity of the assessment
outcomes to uncertainty
Example Industry: The
Tart Cherry Industry
• Highly sensitive to weather and climate extremes
• Requires long-term capital investment decisions
– orchard life cycles ~ 20-30 years
• Limited adaptation options
• Undergoing a substantial evolution with large potential regional and
international shifts of production and international trade
• Small enough in size and scope that it is possible to build a research
team familiar with the different industry components and production
regions
– major production areas are Michigan and central Europe
• Subject of previous intensive efforts to understand the impacts of
weather and climate at the local/regional level
– Pileus Project (www.pileus.msu.edu)
– KliO (www.agrar.hu-berlin.de/agrarmet/forschung/fp/KliO_html)
Expanded Assessment Approach
• Hybrid Modeling
– Combines dynamic and static modeling
• Continuous, evolving projections for system components
where this is possible
– e.g., climate
• Static modeling for time slices where dynamic modeling is not
feasible
– Dynamic models of economic components are either overly abstract
or the modeling requires enormous resources and is not tractable
• Time Slices
– Short enough that while climate is changing the amount
of change is relatively small so that the focus within
each time slice is on the impact of climate variability.
– Later time slices are informed by outcomes of earlier
time slices
Traditional
Impact Assessment
Expanded
Impact Assessment
Comprehensive
Integrated Assessment
Specific Industry
Single Location or Region
Specific Industry
Multiple Regions/Global
Sector (Multiple Industries)
Global
Static Modeling:
Models for individual processes
“feed forward” to downstream
models without interactions and
feedbacks
Time slice 1
Time slice 2
Time slice 3
Static modeling within time slices
Output is an assessment of
* potential impacts for a
LOCATION or REGION
Output is an assessment of
potential impacts for an
INDUSTRY (sub-sector)
Outcomes of earlier time slices
inform future time slices
Isolated time slice
Dynamic Modeling:
Complex, integrated models which
contain all system components and
feedbacks
Hybrid Modeling (dynamic, static):
Models with different complexity
(individual processes and complex
processes with interactions)
Dynamic modeling
Differences between assessment types
Types of Climate Change Impact Assessments
Continuously
running models
Output is an assessment of
potential impacts for a
SECTOR or multiple sectors
Winkler, J.A., S. Thornsbury, M. Artavio, F.-M. Chmielewski, D. Kirschke, S. Lee, M. Liszewska, S. Loveridge, P.-N. Tan, S. Zhong, J.A. Andresen, J.R. Black, R. Kurlus, D. Nizalov,
N. Olynk, Z. Ustrnul, C. Zavalloni, J.M. Bisanz, G. Bujdosó, L. Fusina, Y. Henniges, P. Hilsendegen, K. Lar, L. Malarzewski, T. Moeller, R. Murmylo, T. Niedzwiedz, O. Nizalova, H.
Prawiranata, N. Rothwell, J. van Ravensway, H. von Witzke, and M. Woods, 2010: Multi-regional climate change assessments for international market systems with long-term investments: A
conceptual framework. Climatic Change, DOI 10.1007/s10584-009-9781-1 .
Expanded Framework
Our Current Objective
• Can we demonstrate that it is possible, in
spite of numerous constraints, to conduct an
industry-wide assessment of the potential
impacts of climate change that is
meaningful to industry stakeholders and to
do so within a framework that allows
comparison and integration among different
industries particularly within the same
sector?
Climate Projections
• Local scenario ensembles for locations in Europe and
Michigan
– Development of combined dynamic/empirical downscaling
methods
– Improved methods for simulating extremes
– Introducing landscape temperature variability
Emissions
Scenario
GCM
RCM
Empirical
downscaling
function
Scenario
Suites
Emissions
Scenario
GCM
RCM
Empirical
downscaling
function
Weather Dependency Modeling
• Modification, development and evaluation
of tart cherry phenology and yield models
– Output will be a joint probability distribution of
production across regions
Adaptation
• Identify adaptation options for tart cherry production
– change in cultivars
– input mix (e.g., more/less frost protection), irrigation
– land use changes (e.g., converting orchards to alternative
uses)
– use of insurance instruments
• Availability of adaptation options and willingness to
adapt need to be considered
• Real Options Model
– Link productivity projections with market equilibrium
generated from decision-making across a set of adaptation
options
• Define decision trees to identify conditional decisions faced by
growers
• Assign joint probability distributions to prices and yields.
Trade Models
• Trade links markets between production regions
both within a country and internationally
• Adjust multi-regional supply and demand model for
climate impact analysis
– Supply functions
• capture regional differences of impact of climate change on
productivity
• reflect quantity of product supplied per year at alternative
prices for a “typical” year within a time slice.
– Demand equations
• functions of commodity prices and income at the beginning of
a time slice.
Regional Economic Development
• Future scenarios of macro-economic
variables consistent with emissions
scenarios used in the climate model
projections
• Between time-slice projections of regional
economic variables (e.g., income)
“Meta-Uncertainty”
• Two broad categories of uncertainty:
– calibration error
• introduced by the short period of observations available to calibrate a
model
– model structure error
• arising from how a model is formulated
• Meta-uncertainty
– aggregated uncertainty due to differences in the functional form, or
structure, of the suite of linked models.
• Ensemble of “final” outcomes from the linked climate, yield,
and economic models
– The ensemble members reflect different combinations of
alternative model structures.
Summary
• Traditional local/regional climate impact
assessments, while useful, do not consider
important spatial and temporal interactions.
• An “expanded” assessment approach is
needed, particularly for international market
systems.