Forecasting Resilience of Social Ecological Landscapes

tailpillowΔιαχείριση

9 Νοε 2013 (πριν από 3 χρόνια και 10 μήνες)

58 εμφανίσεις

Forecasting Resilience of Social
Ecological Landscapes

Some Tools to Help Us Understand This Thing
Called “Sustainability”

Lilian Alessa,
Andrew Kliskey, Mark Altaweel

Resilience and Adaptive Management Group, Water
and Environmental Research Center, University of
Alaska; Center for Social Dynamics and Complexity,
Arizona State
University, University of Chicago,
Argonne National Lab

A Few Sustainability Myths

1.
Sustainability is about the environment.

2.
Consumer choices and grassroots activism
works.

3.
There is no single critical piece of the
sustainability challenge.


Lemonik
, 2009. Princeton, New Jersey.

Humour

Me…..


Sustainability is possibly one of the most
misunderstood words in common usage.


Social structure, particularly agent types,
are powerful determinants of emergent SES
patterns.


The environment has become synonymous
with “green” but we are more of a STS than
an SES (something I’d like you to jot down
for your Immersion Experience tomorrow).


Water is
the

Critical Piece


Entering the “Century of Water”


Most issues depend on water availability,
distribution and/or quality.


Transitions from common pool resource to
trade commodity.


Several “solutions” are not possible unless
water is factored in.


Trends in Water Resources

Also see White, Hinzman,
Alessa,
JGR
Biogeosciences
, 2007


Not just availability but also quality, we can only “clean” so well.


FishR_SP
0
100
200
300
400
500
600
1960
1965
1970
1975
1980
1985
1990
1995
Year
Discharge (ft3)
Dealing with Future Change Requires a
Paradigm Shift in “Science”

Growing evidence that technological interventions
alone are not
effective and may drive critical
changes in water use patterns.

--
UN
Commission on
Sustainable Development (1995).

1
.
Our understanding of the social dynamics in
social
ecological systems

is poor
.

2. Our incorporation of scale is sloppy.

3. Our treatment of SES is oversimplified.

These
may represent
some of our
greatest
vulnerabilities
to effectively coping with change.



ArcticRIMS_UNH

Scale

Alessa et al. 2009. In Press, Sustainability

Screen shot of SES types paper


How Could We Possibly Fail?


Scale


Messy Social Ecological Systems


Underestimation of Social Dynamics


Hubris: we will engineer a solution or
‘sustainability as a hobby’

Desire,

Means

Technology

Perceptions,

Values

Exposure

Networks

Learning

Vulnerable

Resilient

Resources

Disasters/Conflicts

Policies

Gaining an
Edge
:The

Tools


Social Ecological Hotspots Mapping.


The Arctic Water Resources Vulnerability
Index (AWRVI).


Forecasting Environmental Resilience of
Arctic Landscapes (FERAL).

Map

䅳獥獳

䵯摥M

Social Ecological Systems Hotspot
Mapping


Takes social and biophysical values and
uses GIS to map the coupled social
ecological landscape.


Gives us information about where specific
dynamics exist.


Was highlighted as innovative science by
NSF in Spring 2008.

Screen shot of paper



Screen shot of paper


Adapting to Change:
AWRVI


The Arctic Water Resources Vulnerability Index:
AWRVI (“
Ar
-
Vee
”).


Tool to assess
status of water resources at the
watershed scale.


Unifies western and traditional knowledge systems.


Can be used to determine resilience and best
strategies for development.

First and only of its kind for high latitudes and
local scales.

Environmental Vulnerability Indices


EVI: United Nations Environment
Programme

(2001).


UN Commission on Sustainable
Development (1995).


Global Commission on Fresh Water
Resources (2004).


Water recognized as single
-
most important
variable in rapid change.


Emergent tools

agent
-
based models
(ABM)


Agent Based Models


Specify the rules of behavior of individuals
(agents) as well as rules of interaction


Simulate many agents using a computer
model


Explore the consequences of the agent
-
level rules on the population as a whole


“Simple” models to produce complex
behaviors



“How could drops of water know themselves to be a river? Yet the river flows on”


--
Antoine du Sainte
-
Exupery

Agents and Systems


agents have connections to each other, and
form a system and operate in an
environment with feedbacks


agents behave autonomously thus they each
have their own parameters (data) and
behaviors


systems change once the agents affect the
threshold in a significant way

Agent Based Models

Are not


An attempt to perfectly reproduce reality (usually)


Are


Are a tool to gain intuition about the system of interest
without needing to know all of the details


A tool to run “experiments” which cannot be performed
in real life


A tool to generate and test hypotheses about what is
occurring


A tool to refine data collection foci


Big Questions


What drives the human hydrological
system?


How do societies ‘overshoot’ their
resources (both social and physical)?


How can we learn to avoid this fate?
(Should we? If so, why?)


Move beyond rhetoric.

Source: Alessa , Kliskey,
and Altaweel. 2009, In Press, Sustainability

Desire,

Means

Technology

Perceptions,

Values

Exposure

Networks

Learning

Vulnerable

Resilient

Resources

Disasters/Conflicts

Policies

Forecasting Environmental Resilience of Arctic
Landscapes (FERAL)


Developing “Real” Rules


Too often, ABMs rely on ‘artificial’ rules
(e.g., games).


Or ….”what ifs”.


It is critical that rules be derived from the
messy, real world.


Humans are not logical but they are
predictable.

"Man is a complex being; he makes the deserts bloom and lakes die."



Gil Stern

Developing Real Rules


There are three rules of thumb to
successfully developing rule sets for ABM.

1.
Observe your system to the point of intimacy.

2.
Establish colleagues in it who will assist you
with field work and data collection.

3.
Include modelers at the outset, not once you
think it would be “nice” to model.

Screen shot of JASSS 2 paper


Your Immersion Experience


Tomorrow you will go out into three “SES”
(two being primarily “STS”).


As yourself “who/what are the objects in the
landscape” (e.g., people, terrain,
interventions, others?).


For each of these objects, what would you
need to know about them to develop
meaningful rule sets?

Applying Agent
-
Based Modeling


Source:
Altaweel, Alessa, and Kliskey
,
JASSS, Forthcoming

Values held toward water


drinking
washing and cleaning
subsistence uses
travel
recreation
biological
cultural activities
Most important value of water
18- 39 years
40- 59 years
60- 99 years
Source: Alessa , Kliskey, Williams. Society & Natural Resources. 2008
.


Current Social ABM

in

FERAL

Step 1: Assess water source selection process

with observed trends and

determine consequences of water

selection choices.

Integrated Models: Example Runtime Output 2

River Discharge

Quantity change belief

Maximum

Mean

How People Make Choices: Why We Need to Know This

People make decisions according to their life experiences,

social relationships, and perceptions of what is around them.


Different people have different influence and goals that influences

other people around
them: three agent types, alpha, beta and gamma
.

A person’s ideas

Person’s decision

The thought process

Social influence and behavior affects water use

Decision Making: Divisions in the Decision Process

1

25

10

Plot points show agents.

Red=Reject

Blue=Accept

Different agent types affect

whether decisions made result

In collective or individual

benefits


Results show
cliques

forming and
social position

of those rejecting

an idea.

Decision Making: Representation in Social Space

25

1

10

Social network

representation of

relationships.

Over a few ticks, more

people agree to accept the

initial idea. However, this

often occurs if leaders agree

initially and coordinate their

efforts.

Black=Reject

Light Blue=Accept

25

Negative Relationships

Changing Viewpoints: Effect on Decision Making


Group vs. Individual

goals

FERAL: White Mountain Scenario


agent

White Mountain

Fish River

Municipal Water source

World Wind 3D

visualization view

10
-
Year Scenario: Travel To River


agents

White Mountain

Fish River

Agents concentrate at river

sources nearest to

White Mountain.

10
-
Year Scenario: Tracking Total Movements

Aggregate agent movements during each

Time tick.

Concentration of movements

over entire simulation and time.


Municipal and non
-
municipal sources fluctuate seasonally.

Colors in water sources indicate relative levels, blue

colors indicate high volume, while red is lower volume.


agents accessing

the municipal source


house icon varies in size based

on population levels

10
-
Year Scenario: Travel To River


agents

White Mountain

Fish River

0
2000
4000
6000
8000
10000
12000
14000
16000
18000
20000
1900
1910
1920
1930
1940
1950
1960
1970
1980
1990
2000
Population
0
20
40
60
80
100
120
140
160
180
200
Water Use (litres per capita per day)
Population
Water use
In:
Alessa, Kliskey, Busey, Hinzman, White.
Global Environmental Change,
2008.

Evolution of Water
Use on the Seward
Peninsula

Take Home Messages


Many of the challenges in sustainability are
not ‘fixable’ using technologies or good will.


Agents drive the system from the bottom up
and some dynamics simply aren’t pretty.


A powerful approach to understanding
consequences is to use agent based models.


ABMs allow the unpredictable outcomes of
simple choices and changes in patterns of use
to be visualized in virtual worlds.

Acknowledgements


The RAM Group at UAA


My colleagues at the International Arctic Research Center,
and the Institute for Northern Engineering, UAF


My colleagues at the Center for Social Dynamics and
Complexity, Arizona State University


Fabrice Renaud, Head, Environmental Assessment and
Resource Vulnerability Section, United Nations University


Volker Grimm, Director, Center for Environmental Research,
Leipzig
-
Halle, Germany.


The National Science Foundation.