Transcript An Introduction to Complex Systems and Complex Systems Science

Complex Dynamics of Urban
Systems – Some Reflections
David Batten
IIASA, IFS, Temaplan Group & CSIRO
[email protected]
Summary
IIASA’s comparative work in the eighties
Nested Dynamics of Metropolitan Processes and Policies
Cities – planned or self-organizing systems?
“Booster” theories of selective urban growth
Large ABMs – e.g. TRANSIMS (Albuquerque), EPISIMS
The new drivers
Global markets – Space versus place, land, water, ecosystems
Climate change – GHG emissions, warming, sea rise
Peak oil – Low emissions transport, new ways of interacting?
Where to next and with what toolkit?
Nonlinear human/climate/ecosystems interface
CSS Working Groups and Interaction Tasks
CABM/HEMA, CDUS, integrated mega-models
Adaptive capacity of Australian cities (Climate Adaptation Flagship)
Fragility of critical infrastructures (with IIASA again)
Nested Dynamics of Metropolitan
Processes and Policies (IIASA)
Initiated in 1982
Aims:
To enhance our primitive understanding of interacting metropolitan
change processes which are operating at very different speeds (“slow
and fast” dynamics)
To develop new concepts and tools that could probe beyond familiar
lifecycle theories of urbanization, suburbanization and de-urbanization
Approach:
Systematic comparison of changes and simultaneous interactions
between 5 metropolitan subsystems in about 20 major cities:
Population
Housing
Transportation and infrastructure
Economy and workplaces
Institutional management
Key Subsystems and Interactions
TYPICAL
INTERMEDIATE LINKAGE
FINAL
DEMAND PARAMETERS DEMAND
SUPPLY
SYSTEM
(STOCKS)
CAPACITY
CHANGES
Housing
System
Changes
in housing
capacity &
location
Dwellings
Transport
System
Changes
in transport
capacity &
location
Transport
Services
Production
System
Changes in
production
capacity &
location
Workplaces
Household
size (-)
Vehicle
density (+)
Population
Rate of
employment (-)
Capacity Tensions
Tension signals arise when a state of excess demand or excess
supply grows larger, owing to inconsistent directions or speeds
of change of the supply and demand components.
e.g. Letting yD denote demand for and xD supply of dwellings at
time t, we can formalize the definition of a capacity tension as a
state in which:
dxD/dt > dyD/dt when xD > yD
or
dxD/dt < dyD/dt when xD < yD
In the eighties, most urban management decisions were seen as
necessary responses or adjustments to signals of imbalances and
capacity tensions in the urban system.
However, such signals can be misleading if the underlying
dynamics are not well understood.
Planned or Self-Organized?
For much of the twentieth century, cities were thought
to be the result of premeditated planning alone
Some urban scientists believed that their geographical
location and design could even be optimized
Views on urban evolution changed in the 80s and 90s:
“Booster” theories – feedback loops (William Cronon)
Self-organizing human settlements (Peter Allen)
Cities may behave more like human brains
Self-maintaining and self-sustaining
Self-repairing
New set of drivers have emerged
“Booster” Theories of Urban Growth
Climate,
the natural
environment
and other
attractors
Migration
and Trade
Greater
Specialization
Selective
Growth of
Settlements
Increasing
Returns to Scale
& Agglomeration
GROWING
CIRCULATION OF
GOODS AND PEOPLE
(POSITIVE FEEDBACK LOOP)
New Drivers of Urban Dynamics?
Global Markets (How and where we produce)
Space versus place?
Resource scarcities – e.g. water, energy (see below)
Land degradation
Threatened ecosystems
Climate Change (How and where we live/consume)
GHG emissions and air pollution
Global warming
Sea rise
Peak Oil (How we interact)
Low emissions transport?
New ways of moving and interacting?
Where Next and What Toolkit?
Human/Climate/Ecosystems Interface
CSIRO-CCSS Working Groups and Interaction Tasks
ABM WG (David Batten) + HEMA network (Pascal Perez)
e.g. NEMSIM, Rangelands model, Barrier Reef model et al
Complex Dynamics of Urban Systems IT
Mega-models – e.g. TRANSIMS, EPISIMS, EPICAST
Integrating social processes in climate & earth system
models (John Finnigan) – possibly involving ABM
Adaptive Capacity of Cities
Climate Adaptation Flagship (Liveable cities, coasts & regions)
Audit of adaptive capacity of Australian cities and towns?
Fragility of Critical Infrastructures
IIASA (http://www.iiasa.ac.at/Research/FCI/index.html?sb=8)
Climate Adaptation Flagship
Theme 2: Liveable cities, coasts and regions
Our urban and coastal populations are exposed to climate change
through:
declining water availability
increasing extreme weather events
sea level rise.
The four focus areas of this Theme of Flagship research are:
new building and infrastructure design, and adaptation of built
infrastructure at building, development and urban system scales
infrastructure planning at larger scales (cities, coastal development)
that takes into account policies, codes, regulation, and demands for
emergency services
integration of social, economic and environmental analyses to help
communities, industry and governments adapt to the impacts of climate
change at regional scales
human health and diseases, extreme temperatures and spatial shifts in
vector-borne diseases.
Some Useful References
Michael Batty (2005): Cities and Complexity:
Understanding Cities with Cellular Automata,
Agent-Based Models and Fractals, MIT Press.
Juval Portugali (2000): Self-Organization and the
City, Springer Series in Synergetics.
David Batten (2000): Discovering Artificial
Economics: How Agents Learn and Economies
Evolve, Westview Press.
Pascal Perez and David Batten (2006): Complex
Science for a Complex World: Exploring Human
Ecosystems with Agents, ANU ePress.
I am currently reviewing
NEMSIM = National Electricity Market
Simulator
Goal: To evolve “would-be” worlds of new
agents, new micro-grids and new rules
Simulation is changing the frontiers of science
We can explore “What-if” scenarios of really
complex systems
Like cities, our National Electricity Market
(NEM) is a Complex Adaptive System
Our NEM as a Complex Adaptive System
Market of
Adaptive
Agents
Physical
Energy
Network
Changes in climate and
weather forecasts,
contribute to
price volatility
and demand
uncertainty
in the NEM
Stationary
energy
accounts
for about 60%
of all GHG emissions
Socio-Technical System
Climate
Scenarios
GHG
Emissions
Calculator
Natural System
What kind of Simulator is it?
Agent-based simulation (or MAS)
NEM participants are the software agents
Agents’ behaviours programmed via rules
Action evolves in 3 simulated environments
Collective outcomes (and surprises) emerge
from the bottom up.
Examples are price volatility, market power,
network congestion, regional blackouts
and excessive GHG emissions.
Smart Generator Agent: Re-bidding
Tuesday, 24/06/2003
$/MWh
10
P10
300.0
350.0
400.0
450.0
500.0
550.0
600.0
650.0
700.0
9126.00
P9
6300.00
P8
8
P7
252.00
P6
136.00
P5
82.00
P4
36.00
P3
Axis Title
bands
TenYprice
2200.00
6
MW
4
18.20
P2
8.61
P1
2
0
4:00
2 8:00
4
12:00
16:00 6
20:00
8 00:00
10
4:00
(48 trading intervals)
X Axis Title
Generating Unit (Thermal – coal)
Re-bid stack
submitted at 22:00
on the previous day
Capacity Withholding
Price
($/MWh)
Evening
peak
12000
04.30 18.00 14.00 22.00 09.30
10000
8000
6000
This 09.30 band
was shifted down
three times in the
morning via rebids
4000
2000
0
2000
-2000
2500
3000
Quantity Offered (MW)
3500
An Overview of NEMSIM
Typical Graphical Output
Regional Summary Window for
GHG Emissions
Thank you
David Batten
Coordinator, CSIRO Agent-Based Modelling Working Group
CSIRO Marine & Atmospheric Research
[email protected]