Adaptive policy for urban planning:Adaptive policy for ...€¦ · Adaptive policy for urban planning: ... The state of a ... SWOT analysis Map SWOT & reclassify SLEUTH’s input

Adaptive policy for urban planning:Adaptive policy for urban planning: operational models in support of

planning policy.

Venue: Hardy Building Rm 101. Time: 1-2pm

Dr. Elisabete A. [email protected]@

Department of Land EconomyUniversity of CambridgeUniversity of Cambridge

Geography, Dep. Cambridge, UK5th November, 2015

OUTLINE

1. The argument in favor of adaptive policy2 Complexity theory: The right moment in time to link2. Complexity theory: The right moment in time to link

planning decision and urban modelsM t hi th k tt ib t f PPS DSS ith D i-Matching the key attributes of PPS -DSS with Dynamic Simulation and Planning Decisions (Policy Support)

3. ‘Wicked problems’ and the wrong “decision makers model4 Key areas to address: Calibration Validation4. Key areas to address: Calibration, Validation,

Randomness , uncertainty, data-mining5. The examples of models: The SLEUTH model; The

CVCA model; CCID model; The DG-ABC model6. Concluding remarks


1. The argument in favor of adaptive policypolicy


Cities and landscapes evolve in time and spaceCities and landscapes evolve - in time and space (across scales and along the same scale)

The rational models of the 50-70s - systems theory orThe rational models of the 50 70s systems theory or participative theory they are both based on the ‘presumption of certainty’ they provide one answerpresumption of certainty - they provide one answer to the decision maker (static snapshot of time)

Historical evolution is due to theory practiceHistorical evolution is due to theory, practice, professional qualifications/numbers, computation, data constraintsconstraints


The simplified reality of static world resulting from p y goverlays of data is not enough

Today is a result of complex physical and social interactions that have in account past events and future expectationsexpectations

Pure causation is not enough and cumulative effects, ‘carrying capacity’ , self-organization, etc. play important y g p y g p y proles

Complexity theoryGeography, Dep. Cambridge, UK5th November, 2015

Element/variable

What is complexity?

Element/variable attribute (i.e. temperature)

In terms of representing various phenomenaA – yes/noA yes/noB – only complexity(i.e. water convection cells– at the edge of “chaos”)C – yes/no (non complex phenomena)time

A and C can be dealt with other models, but CM can be very importantB can only be understood by using Complexity modelsWhy: because B represents a “transient reality” something that will become different(the trajector is not “linear” 1+2 might not be eq al to a same realit (e en if more/less(the trajectory is not “linear” 1+2=might not be equal to a same reality (even if more/less

intense))


Identification of complex behaviour for an element or phenomena through timeIdentification of complex behaviour for an element or phenomena through time

Identification of complex behaviour for multiple elements or variables through time and space


p p g p

How a particular phase transition is deployed – the vortex of time


2- ‘Wicked problems’ and the wrong “decision makers model”

Rittel and Webber’s 1973 conception of “wicked problems” to explain why conventional scientific approaches failed to solveexplain why conventional scientific approaches failed to solve problems of pluralistic urban societies. Try to confront (urban) social problems with the wrong toolsTry to confront (urban) social problems with the wrong tools

because we have misunderstood the very nature of the problems“wicked problems have no stopping rule ” and “wicked problems wicked problems have no stopping rule, and wicked problems

do not have an enumerable (or an exhaustively describable) set of potential solutionspotential solutions

The concept of certainly in an uncertain world The timing of Lee’s requiem and Rittel and Webber’s Wicked problems


3. Complexity theory: The right moment in time to link l i d i i d b d lplanning decision and urban models

Mismatch between technology, theory, data of the 70s gy, y,resulted in Lees’ “Requiem for large scale models”XXI century of Big Data high computation capabilityXXI century of Big Data, high computation capability,

vast numbers of experts, more data-aware policy Key contributions: Von Neumann and Morgenstern (1944, 1966), Ulam (1960, 1974), Prigogine (1977,(1944, 1966), Ulam (1960, 1974), Prigogine (1977, 1999, 1984), Tobler and Burks (1979), Kauffman (1984, 1993) Wolfram (1994) Holland (1995 1999) and1993),Wolfram (1994), Holland (1995, 1999), and Crutchfield (1995); John Nash exploring research

lt b M ill Fl d d M l i D h t RANDresults by Merrill Flood and Melvin Dresher at RAND corporation (1950s);


Theory Processes Modelsffocus

time50s

Deterministic

Systems theory Rational Planning

Physical/regions

Beyond modernity

Advocacy of Planning

People/

Participative PlanningIncremental Planning

Mix Scanning

People/social

Mix-Scanning (Zoom in-out, Top/down-b/up)

Stochastic

Complexity Theory

Top/down b/up)

Feedback loopsf l / d t/ i h k l dof learn/adapt/enrich_knowledge


CAs(i) A grid or raster space – organised by cells which are the smallest units in that(i) A grid or raster space organised by cells which are the smallest units in thatgrid/space;(ii) (ii) Cell States – cells must manifest adjacency or proximity. The state of a

ll h di l t t iti l hi h d fi d i t fcell can change accordingly to transition rules, which are defined in terms ofneighbourhood functions;(iii) (iii) The neighbourhood and dependency of the state of any cell on the state( ) g p y yand configuration of other cells in the neighbourhood of that cell;(iv) (iv) Transition rules that are decision rules or transition functions of the CAmodel and can be deterministic or stochastic;model and can be deterministic or stochastic;(v) (v) Sequences of time steps. When activated, the CA proceeds through aseries of iterations

study of random complex CA came an understanding of its basic patterns: as they appear to fall into four qualitativebasic patterns: as they appear to fall into four qualitative classes, in what concerns one-dimension (1-D) CA evolution leads to: (i) a homogenous state; (ii) a set of separated

( ) ( )simple stable or periodic structures; (iii) a chaotic pattern; (iv)complex localised structures, sometimes long-lived (Wolfram, 1984:5)


1984:5)

ABM-GAs are constituted of:

(i) agents that do not have the constraints of neighbourhoodeffects,

(ii) b h i l l t d th i t it lf (iii) (ii) behavioural roles among agents and the environment itself, (iii)independence from central command/control, but able to act if actionat a distance is required,q , (iv) states of agents tend to represent behavioural forms.

The most basic model environment of an ABM-GA will have a setof attributes per agent (or group of agents), (one)a set of decision treesand trigger points that will allow to set the context for a new movementand trigger points that will allow to set the context for a new movement(upgrade of the spatial/temporal environment) in time/space .


Starting the study of complex systems in Spatial Analysis ….

Waldo Tobler in contact with Arthur Burks was exposed to Von Neumann’s ‘C G ’ (19 9)works, and published ‘Cellular Geography’ (1979).

At NCGIA S t B b H l C l li d K ith Cl k bli h dAt NCGIA-Santa Barbara, Helen Couclelis and Keith Clarke, published respectively ’Cellular Worlds’ (Couclelis, 1985) and develop the first fully operational and implementable CA (Clarke and Gaydos, 1998). While et. ope at o a a d p e e tab e C (C a e a d Gaydos, 998) e etsince the 1990’s focus in the ‘adaptive’ CA as a basis basis of integrated dynamic regional analysis (1997 )

Michael Batty initially at NCGIA-Buffalo and afterwards at CASA-UCL, developed the theory and practice that culminated in the publication of thedeveloped the theory and practice that culminated in the publication of the seminal books ’Fractal Cities’ (1994) and ‘Cities and complexity’ (2005). Recently, Wolfram’s book ‘A New Kind of Science’ (2002) y, ( )

ES = 3rd Generation (consolidation, reassemble, expansion, validation)


Operational Dynamic Urban Models


urban19751976 urban

roadsslope

excludedhilshade

1976...

test mode

coarse

hilshade1997 calibratefine

finaldata

i iti

SLEUTH Urban Model

forecastDNAacquisition

metricsimages

19981999

.

.

reclass excluded

calculate LA metrics

apply LA strategies

1998

1999SLEUTH resultsCVCA

?CVCA

.

.2025

results

workshop’s morning

Keep existent DNA?

yes?

no?forecastEnvironmentalModel

images metrics

morning

urbanroadsslope

new DNA

E I l iworkshop’s afternoon

SWOT analysis

Map SWOT &

reclassify SLEUTH’s input data

slopeexcludedhilshade

Expert Inclusion‘people’s model’


-Map-SWOT & analysis-critic

2. SLEUTH


100120

cells

20406080

umbe

r of u

rban

Identification/Quantificationof the metrics that control

0D

iffus

ion

Bre

ed

Spr

ead

Slo

pe

Roa

ds

Diff

usio

n

Bre

ed

Spr

ead

Slo

pe

Roa

ds

nu of the metrics that control the behavior of the system

AML

D D

AML AMP

final fine coarse

future

AMLAMP

Identification futureIdentification/Quantificationof phase

pastof phase

transitionsConfirmation of fractal

Urban growthof fractal dimension


urban roads slope

1975 1976

. test mode slope

excluded hilshade

.

.

1997 calibrate

coarse

fine

f t

final

DNA

data aquisition

t ii forecastDNA

1998

metrics images

199819981999

.

.

reclass excluded


apply LA strategies

1998

1999

.

. 2025

CVCA - CA - Environmental Model

3. CVCA Model

images metrics


Transition Rules:

Number of pixels (pixels with a probability of change to urban)

Action stepA. Protective

Desired network elements are identified and protected through planning policy and land change to urban)

1 Protective 0 but NN > than add protective pixels B. Defensive

use control in advance of negative landscape matrix changes.

Isolated core area in ‘non-supportive 1. Protective 0 but NN > MNND

than add protective pixels around all outer patch and add protective pixels until arriving at closest neighbor

landscape matrix’ is subject to isolation from disturbance to corridors and to incremental reduction in size of the core area that can be protected through a new buffer zone.neighbor

2.Defensive <=50% *,** than add defensive pixels to all outer patch cell where transition cell exists

C. OffensiveIsolated core area is protected with a buffer zone and linked into a greenway network with corridors that are newly developed within a non-supportive landscape matrix

3.Offensive >50% add offensive pixel to all outer patch cells and add offensive cells until nearest neighbor D Opportunisti

within a non supportive landscape matrix context. The offensive strategy employs a range of tactics, including nature development, to achieve a desired landscape configuration.

nearest neighbor

4.Opportunistic 0 but NN = NNI (and no transition cell nearby) than link to nearest neighbor

D.Opportunistic

Isolated core area is linked with an existing corridor, buffered, and anew supporting landscape matrix is developed. The opportunistic strategy takes advantage of unique circumstances that may only support nearest neighbor

5. Grow Goal or Result

some greenway uses, e.g. recreation.

Existing Landscape

Core AreaBuffer Zone

Corridor

Supporting Landscape MatrixNon-Supporting Landscape Matrix


Metric - AMP ValueMetric - AML Value Metric AMP Value

Edges 14964

Area 24204

Metric AML Value

Edges 35171

Area 106460

Num Clusters 708

MCS 34

MPS 275

Num Clusters 1134

MCS 93

MPS 577 MPS 275

LSI 7.7

MNND 1.5

MPS 577

LSI 9.9

MNND 1.6


CVCA SimulationCVCA Simulation


Protective cells

Defensive corridor

Promoting big patches, by avoiding divide on big patch in t t


to two

urban19751976 urban

roadsslope

excludedhilshade

1976...

test mode

coarse

hilshade1997 calibratefine

finaldata

i itiforecastDNA

acquisitionmetricsimages

19981999

.

.

reclass excluded


apply LA strategies

1998

1999SLEUTH resultsCVCA

?..

2025

results

workshop’s morning

Keep existent DNA?

yes?

no?forecast

h l d limages metrics

morning

urbanroadsslope

new DNA 4. The People’s model

workshop’s afternoon

SWOT analysis

Map SWOT &

reclassify SLEUTH’s input data

slopeexcludedhilshade


-Map-SWOT & analysis-critic

TWO MAP DRAWINGS RESULTING FROM THE WORKSHOP’S AFTERNOON


SWOT RESULTSStrengths Votes

%Weakness votes

%Opportunities Votes

%Threats votes

SWOT RESULTS

Transport system (road

19.5 Mobility, accessibility

32.6 Improve transportation

17.6 Uncontrolled urban sprawl

29.9y (

network, airport, harbor)

yand transport

psystem

p

Tourism and world heritage (Lisbon and Porto)

17.8 Lack of urban quality

17.0 Urban renewal 15.7 Natural risks (e.g. coastal, flooding, earthquake)

16.4

Porto) earthquake)

Capital city 13.0 Uncontrolled 11.3 Cultural 11.8 Urban violence 14.2p yurban sprawl tourism/

eventsand drugs



1. The image of the city

The image of a city-regionIdentification/quantificationU b f ( i t t / ibl )Urban forms (existent /possible)

“same future” – different simulations same future – different simulations

Geography, Dep. Cambridge, UK5th November, 2015 28

7. DG-ABC MODEL

Concept model of DG-ABC model

Intelligent agents Cellular automata TPB model Genetic algorithmIntelligent agents Cellular automata TPB model Genetic algorithm

Dynamics capturing

a-spatial dynamics spatial dynamics behavioural regulations

behavioural optimizations

Factors social-economic infrastructures/ behaviours of agents behaviours of agentsFactors oriented

social-economic influences

infrastructures/ecosystems

behaviours of agents behaviours of agents

Level individual individual individual level high level

h lt b h i b i hb h d N/A l ti bchanges alter behaviours by GA and themselves

neighbourhoods navigation

N/A evolution by themselves

Data requirement

social-economics /policies quantifying

GIS data agent’s beliefs/ profile information

strategies/optionsg

Integrated model


DG-ABC MODEL

3.2 DG-ABC model

1. Model Environment2. Heterogeneous agents3 CA (SLEUTH)3. CA (SLEUTH)4. Decision behaviors5. Interactions6. Synchronization

h k d blThe key decision tables:

• The Resident agents’ utility table.• The developer agents’ development

spatial�data

application table. • The government agent’s approving table. • Synchronization decision table.

Source: Ning Wu and Elisabete A. Silva 2010a


DG-ABC MODEL

3.3 Theory of Planned Behavior

a i ia Ap AW WA Behavioral Beliefs

Attitude toward the Attitude toward the

i p Ctd ctAp AW WA

1( ) /

na a ii ij ji neighbor

jSN M Inf N

m

I

Normative Beliefs

Subjective Norm Intention

behavior

Behavior

behavior

Subjective Norm Intention Behavior

atraffic environment convenience tijEI a E b E c E

1

ma a ai ik ki

kPBC Cb P

Perceived Behavioral Control

Control Beliefs

Actual Behavioral Control

Perceived Behavioral Control

Actual Behavioral Control ff j

Be W I W AbC W I W EI • A: the degree to which the performance of the behaviour is

TpB model (Icek Ajzen 2006) 2 311 2 3( )highway citycenterroad B D B DB Dt t t

trafficE w A e w A e w A e

1 2 3( ) max{ , , .......... }nD f B Be Be Be Be

1 2 1 2Be W I W AbC W I W EI g ppositively or negatively valued.

• SN: an agent’s perception of social normative pressures, orrelevant others’ beliefs that the agent should (not) performsuch behaviour.

• PBC: an individual’s perceived ease or difficulty of performing p y p gthe particular behaviour.

• I: an indication of a agent’s readiness to perform a givenbehaviour.


Properties of government developer agentsp g p g

Properties of resident agents

Properties of property developer agents


p p p y p g

Spatial synchronization in the model

Temporal synchronization in the model


(a) run CA standby (b) run agents standby

(a) run CA standby (b) run agents standby

(c) run integrated model(d) real urban data


(c) run integrated model


An Integrated Spatial Analysis Environment for Urban-Building Energy Analysis in Cities (I-UBEA)

• To track energy changeE h f L d

RESEARCH OBJECTIVES

Analysis in Cities (I-UBEA)

Energy change of London Energy change per local authority / buildings

• To estimate and analyze Energy Usagey gy gIntensities (EUI) EUI of Local Authorities

EUI of sub categories of buildings EUI of sub-categories of buildings• To explain energy consumption

Explain how the distribution of land use influences energy consumption in local authorities and the entire city of London

Explain how the distribution of floor area pinfluences energy consumption

• To evaluate energy performanceEvaluate energy performance of London while Evaluate energy performance of London while adapting different energy change policies.

Interactive Simulation Model for predicting of energy performance in the future on the basis of

Geography, Dep. Cambridge, UK5th November, 2015HIT / ### Conference /

Sep 25 2013

energy performance in the future on the basis of population change.

Department of Engineering Department of Land Economy

FUNCTION OVERVIEW OF I-UBEA

GIS Data Visualization & Data query in different spatial scales, for both polygon and point data

Policy impact simulation and energy

GIS Boundary Data

I-UBEAI-UBEA

point data prediction (with population dynamic using statistics model)

Data

GIS Boundary I UBEAI UBEA

Attribute Data

yData

Energy & EUI

Attribute Data

Explain energy ti ith tt ib tconsumption with attribute

data by using Statistics AnalysisAttribute

Data

Geography, Dep. Cambridge, UK5th November, 2015The integrated model assembles functions from GIS, R, EnergyPlus and Netlogo

Energy scenario analysis (what‐if analysis)

tion

Baseline scenario

This method is more flexible compared to regression analysis because it is suitable for the situation where energy consumption data are unavailable in some spatial scales or some areas of a city.

Elec

trifi

cat

This scenario is to explore how the total energy or carbon emissions would change if building sector would use electricity instead of gas for heating

• Baseline scenarioThe baseline scenario is based on prior knowledge on

(1

) Eer

sion

electricity instead of gas for heating. energy use intensity in terms of building types. Three methods are provided:

ype

conv

e Three methods are provided: benchmark values median values

This Scenario investigates the change of building types on the energy consumption in cities

uild

ing

ty

Monte Carlo methodAnalysts may choose one or all these three methods based

consumption in cities.

(2

) Bic

ient

all these three methods based on data availability.

This scenario allows analysts studying the influences of change of energy use

nerg

y ef

fi the influences of change of energy use intensities for different building types.

Geography, Dep. Cambridge, UK5th November, 2015 (3

) En

Note: The benchmark values are based on low,

typical, and high values (EUI) from previous literature.

Model function: Energy optimization

Data: UKMap, floor area of London, Gas Energy consumption for each MSOA in London.p

Calculate EUI for MSOA and find polygon with best Calculate EUI for MSOA and find polygon with best EUI.

calculate area percentage for all polygons.II--UBEAUBEAII--UBEAUBEA pick best performing ratio and calculate mean

ratio.

calculate difference to best performing ratio and p gdifference to mean ratio, respectively, for compositional data.

Part 1. display overall Part 2. display

calculate overall difference to best performing ratio and to mean ratio, respectively.

p ydifference to best performing ratio with scaled color

p yoverall difference to mean ratio with scaled color


Note: Each part of step 6 can be executed independently.

Result: Energy optimization based on building types at MSOA level in London

• To compute the overall difference of floor area percentage to best‐performing area in terms of gas use intensity at London MSOA level.terms of gas use intensity at London MSOA level.

• The overall differences of floor area percentage based on gas use intensity still presents the characteristics of spatial distribution to some extent, though not clustered as large area.

• Spatial distribution demonstrates that energy consumptions in an area are to some extent influenced by where the area is located in a city.

best-performing MSOA

MSOA ith tMSOA without gas consumption data

MSOAs with similar overall difference are likely to clusterlikely to cluster together.

small difference

Geography, Dep. Cambridge, UK5th November, 2015 large difference

5. CI Agent Base Model

demand for office

individual citizens

existing

new housing estate

locational determinants for CWs’housing

to accept

variation in CIs firms’ number & size

supply

derelict factories

existing housing estate

spatial

housing choosing

compensationadvocating, regulating &

to refusesize

interaction & feedback

land expropriation

development of CIs derelict factories, warehouses

slum, illegal buildingsb

structure of CIs firms

influence upon land‐use type

controlling

new policies for CIs, CWs interaction

expectation & feedback

expropriation, urban regeneration dilapidated housing

area

urban government

spatial

and citizens of neighbouring plots

for CIs, CWs and new land‐use plan


expectation & feedback

demand for CWsfarmland, unworked acres

structure of CWs’habitation

i ti

compensation

advocating, regulating & controlling


demand for housing

determinants for CIs firms’location choosing

existing housing estate

new housing estate

individual citizens

to refuse

variation in CWs population

supply

to accept




Agents and terms forAgents and terms for negotiationnegotiation

• Agents:196 countries– Annex1 (42)

No Annex1 (149)– No Annex1 (149)– Others (5)

• Terms for negotiation– Technology trade– Carbon trade– GDP growth supportg pp

• Negotiation rules

Geography, Dep. Cambridge, UK5th November, 201505/11/2014

Condition-action rulesCondition action rules

Not included in the Negotiation process, Focusing on GDP growth

Negotiation, aiming to lift GDP th tGDP growth rate

Negotiation, to lift GDPNegotiation, to lift GDP growth, consider carbon reduction

Negotiation, following current agreements andcurrent agreements and strategies

Negotiation, aiming to carbon reduction

Geography, Dep. Cambridge, UK5th November, 201505/11/2014



Some Papers

2013 Simulating the dynamics between the development of creative industries and urban spatial g y p pstructure: an agent-based model (with H. Liu). S. Geertman et al. (eds.), Planning Support Systems for Sustainable Urban Development, Lecture Notes in Geoinformation and Cartography, DOI: 10.1007/978-3-642-37533-0_4, Springer-Verlag Berlin Heidelberg pp. 51-72_ p g g g pp2013 Selecting artificial intelligence urban models using waves of complexity. Urban Design and Planning. 166 (1): 1-22012 Surveying Models in Urban Land Studies (with N Wu) Journal of Planning Literature 27 (2012 Surveying Models in Urban Land Studies. (with N. Wu) Journal of Planning Literature.27 ( ): 1-142010 Artificial intelligence solutions for Urban Land Dynamics: A Review (with N. Wu). Journal of Planning Literature 2010 24: 246-265Planning Literature. 2010 24: 246-265.2008 Strategies for Landscape Ecology in Metropolitan Planning: Applications Using Cellular Automata Models. (with J. Wileden, J. and J. Ahern), Progress in Planning, 70(4):133-177 - ISSN: 0305 90060305-90062005 Complexity, Emergence and Cellular Urban Models: Lessons Learned from Appling SLEUTH to two Portuguese Cities. (with K. Clarke) European Planning Studies, 13 (1): 93-115 –ISSN: 0965 4313ISSN: 0965-43132004 The DNA of our Regions: artificial intelligence in regional planning. Futures, 36(10):1077-1094. – ISSN: 0016-32872002 Calibration of the SLEUTH Urban Growth Model for Lisbon and Porto, Portugal. (with K. Clarke) Computers, Environment and Urban Systems, 26 (6): 525-552 - ISSN: 0198-97150965-4313


Some book chapters

2014 Measuring space: a review of spatial metrics for urban growth and shrinkage (with J. Reis). In: The Routledge Handbook of Planning Research Methods. (Eds. Patsy Healey, Neil Harris and Pieter van den Broeck), RoutlegePieter van den Broeck), Routlege2014 DG-ABC: An Integrated multi-agent and cellular automata urban growth model (with N. Wu). Technologies in Urban and Spatial Planning: Virtual Cities and Territories? (Eds. Nuno Norte Pinto, José António Tenedório António Pais Antunes and Josep Roca ) IGI-Global pp 57-92José António Tenedório, António Pais Antunes and Josep Roca ), IGI Global, pp.57 922013 Simulating the dynamics between the development of creative industries and urban spatial structure: an agent-based model (with H. Liu). S. Geertman et al. (eds.), Planning Support Systems for Sustainable Urban Development Lecture Notes in Geoinformation and Cartography DOI:for Sustainable Urban Development, Lecture Notes in Geoinformation and Cartography, DOI: 10.1007/978-3-642-37533-0_4, _ Springer-Verlag Berlin Heidelberg, pp. 51-722011 Cellular Automata Models and Agent Base Models for urban studies: from pixels, to cells, to Hexa Dpi’s In: Urban Remote Sensing: Monitoring Synthesis and Modeling in the UrbanHexa-Dpi s. In: Urban Remote Sensing: Monitoring, Synthesis and Modeling in the Urban Environment. Edited by: Dr. XiaojunYang. Wiley-Blackwell. pp. 323-345. ISBN: 978-0-470-74958-62010 Waves of complexity. Theory, models, and practice. In: Roo, Gert de, and Elisabete A. Silva (2010) A Planner’s Encounter with Complexity Ashgate Publishers Ltd Aldershot (UK) pp 309(2010), A Planner s Encounter with Complexity, Ashgate Publishers Ltd, Aldershot (UK). pp. 309-331.. ISBN: 978-1-4094-0265-72010 Complexity and CA, and application to metropolitan areas. In: Roo, Gert de, and ElisabeteA Silva (2010) A Planner’s Encounter with Complexity Ashgate Publishers Ltd AldershotA. Silva (2010), A Planner’s Encounter with Complexity, Ashgate Publishers Ltd, Aldershot (UK). pp..187-207. ISBN: 978-1-4094-0265-7


Elisabete (es424@cam ac uk)Elisabete ([email protected])www.landecon.cam.ac.uk/directory/esilva

LISA Labwww.landecon.cam.ac.uk/research/lisa

New Book: "The Routledge Handbook of Planning Research Methods"Methods http://www.routledge.com/books/details/9780415727952/


Documents

Adaptive policy for urban planning:Adaptive policy for ...€¦ · Adaptive policy for urban planning: ... The state of a ... SWOT analysis Map SWOT & reclassify SLEUTH’s input