Decision Support Systems in a Spatial Planning and Policy context

1
Decision Support Systems in a Spatial Planning
and Policy context

• Guy Engelen
• RIKS bv
• Papenstraat 8
• P.O. Box 463
• 6200 AL Maastricht
• Tel. [phone]
• Fax. [phone]
• e-mail. guy_at_riks.nl
• http//www.riks.nl

2
Who / What is RIKS bv ?

• Founded in 1987 as an institute for applied AI
  research.
• Since 1990 involved in research and development
  of High-resolution Spatially-dynamic models and
  Decision Support Systems
• Since 1995 a company 100 owned by Maastricht
  University
• In 2004 14 employees, 10 Full time equivalents
• In 2003 turnover of /- 750 Kilo
• RIKS clients 40 Netherlands, 60 abroad
• Developers of some 30 spatial DSS products (from
  50 - 2000 k)
• Cities and regions
• Metronamica, Murbandy, Moland, LOV, Xplorah,
  SimLucia, SenSim, CellCity, COHESIE,
• Watersheds and Coastal zones
• MODULUS, Medaction, SimDelta, LADAMER,
  Lobster-IBM, WadBOS, Elbe-DSS, Krim-DSS, RamCo,
  Durme-DSS,
• Nature reserves and Protected areas
• EcoVisie, Strandplevier-BOS,

3
Contents

• Concepts DSS for Spatial Policy-making and
  Planning
• Cellular Automata Land use models
• DSS example XPLORAH (Puerto Rico)
• Background of the project
• The models
• Application of the DSS
• Some conclusions.
• Developing the DSS teamwork.

4

• Decision Support Systems for Spatial
  Policy-making and Planning
• Some concepts

5
Integrated Spatial Planning is

• An explicitly spatial - temporal domain. Planners
  need insight in interacting processes at diverse
  spatial and temporal scales
• Multi-disciplinary, multi-sectoral, and multiple
  spatial agents. Planners need to intervene in
  complex and highly dynamic systems
• Policy-interventions cause irreversible change.
  Planners need to be able to anticipate the
  effects of their interventions, prior to taking
  action
• Policy-interventions require resources. Planners
  need to be able to evaluate and rank alternative
  solutions and set priorities.
• this requires a set of models, methods,
  techniques, data, , available in an integrated
  environment rather than in isolation.

6
A Decision Support System ?

• (Adapted from M. Scott Morton, 1971 and George M.
  Marakas, 1999)
• Is a computer-based information processing system
  with advanced capabilities
• Is employed in weakly-structured decision
  contexts
• Makes use of quantitative techniques
• Supports all phases of the decision making
  process
• Is interactive and user-friendly
• Is readily accessible to the (high-level)
  decision maker
• Facilitates learning on the part of the decision
  maker
• Is intended to support, not to replace decision
  makers
• Is generally developed using an evolutionary,
  iterative process.
• Spatial Decision Support System DSS has spatial
  dimension.

7
Weakly-structured decision context
Uncertainty relative to theknowledge for
solving the problem
weakly structuredproblem
unstructured problem
Model A
Model D
Model K
Model Y
Model V
Model L
weakly structuredproblem
structured problem
Model X
Model Q
Example New road construction Structured
problem How many lanes are required to handle a
peak of 5,000 vehicles per hour in an urban
environment ? Weakly structured problem Can we
guarantee congestion free traffic every day for
the next 15 years ? Unstructured problem Is the
construction of a new motorway a good decision
for all concerned now and in the future ?
Conflicting views on values, goals and measures
relative to the solution of the problem
(After Hoppe and Peterse, 1998 in Van Delden,
2000)
8
Supports the completeDecision Making Process
(Mintzberg et al. 1976)
9
Functions of a DSS

• Structuring and Formalising the decision making
  process
• Knowledge management gather and link all
  relevant pieces of knowledge
• Presenting an integral view of a complex system
• Different functions (economic, social, ecologic)
• Different actors
• Analysis and Exploration of alternative solutions
  in the identification, development and selection
  phase
• Autonomous developments
• Impact of external influences
• Consequences of policy measures
• Learning about the behaviour of the system
  analysed
• Facilitating the Communication between different
  parties and stakeholders.

10
The components of a DSS
Graphical, interactive, enables to access the
Models, the Data and the Tools while hiding the
technical complexity for the end-user
User interface
Front-end
Kernel
Model- base
Data- base
Tool- base
Instruments, techniques and methods to work with
models and data
Formal models that are relevant to represent the
problem domain
Data --GIS and other--required to describe the
problem domain
11
Types of DSSManagement vs. Policy oriented
DSS DecisionSupportSystem
PSS PolicySupportSystem
12
Integrated models are very oftenSimulation models

• Like any other model, simulation models are built
  to address a problem. They mimic reality
• They are descriptive or explorative, not
  prescriptive like optimization models. They
  include a representation of the physical world
  system under study and portray the behavior of
  actors in the system
• In policy design, they support designing
  strategies or organizational structures and
  evaluating their effects on the behaviour of the
  system
• They can easily incorporate feedback effects,
  non-linearities, and dynamics.
• They are not rigidly determined in their
  structure by mathematical limitations as
  optimization models often are.

13
Integrated models are very oftenSimulation models

• Limitations due to
• Accuracy of the process representation. Not to
  be obtained from aggregate statistics, but should
  be obtained first hand.
• The majority of data are soft variables. What we
  know about the world is descriptive and
  qualitative.
• Definition of the model boundary. What is
  exogenous and endogenous.
• Calibration and sensitivity analysis

14

• Cellular Automata
• Dynamic Land Use models

15
Cellular AutomatonExample Conways Life
(Gardner, 1970)
16
RIKS Cellular Automata Land use model (1992)

17
RIKS CA land use model (1992)

• Transition rules representing
• Locational preferences of spatial agents in
  competition for space
• Willingness to develop or give-up activity in a
  particular location
• Appreciation of the proximity of other competing
  or befriended activities and spatial elements in
  the immediate neighbourhood.

Forest
Water

• 32 land uses
• Dynamic functions
• Vacant states
• Fixed features
• Neighbourhood with radius of max. 8 cells, 196
  cells

Commerce
Industry
Housing
18
Simulating the growth of Cincinnati from 1840
till 1960
19
Cincinnati 1960Simulation (left) - Reality
(right)
20
Evolving land use in a heterogeneous geographical
environment
Transition Rule Change cells to land-use for
which they have the highest transition potential
untill the demands are met.
Time Loop
21
ConstrainedCellular Automata

• The Cellular Automata dynamics evolve in a
  non-homogeneous geographical space defined by GIS
  attributes and layers
• Their overall dynamics are not determined by the
  micro Cellular Automata transition rules, but
  by processes at a larger macro scale
• Cellular Automata models have been integrated
  with more traditional dynamic models, which in
  the most general case are regionalised (spatial
  interaction based) (Engelen et al., 1993).

22

• DSS Example
• Xplorah (Puerto Rico)

23
Puerto RicoA small island state

• Area with limited physical dimensions and limited
  non-renewable resources, fresh water, , small
  economic base due to limited population
• Land area 8,959 km2 Total area 9,104 km2
• Population 1999 3,889,507
• Population density 434 inh./km2

24
Puerto RicoA small island state (Cont.)

• Isolated but open systems distant markets and
  players, high transportation costs,
• Large area is coastal zone with high
  concentration of people and activities
  competition for space, high pollution and waste
  generation
• Unique but fragile ecosystems.

25
Puerto RicoOn the path of Hurricanes
26
XplorahObjectives and State of the art

• Faced with an increasing amount of planning
  problems, some caused by Climate Change, the
  decision was taken to develop Xplorah. It is to
  enable a formal and better informed spatial
  planning practice
• Initiated in 2000 by the Graduate School of
  Planning (University of Puerto Rico).
• First prototype available since March 2002,
  second in March 2003
• February 2003, law signed by the Governor after
  unanimous decisions on behalf of the House of
  Representatives, and the Senate to support the
  further development of Xplorah and its use to
  evaluate mid to long term spatial policies
  (horizon 2025)
• End-users PR Planning Board, PR Department of
  Transportation
• In the next 5 years, Xplorah is to evolve into a
  Spatial Decision Support System for the
  Integrated Assessment of Socio-economic and
  Environmental Spatial Policies
• Autonomous developments (dynamics) of the system
• Effects of past policies
• Effects of current policies now and in the
  future
• Effects of alternative and potential future
  policies now and in the future.

27
XplorahProcesses at 3 strongly coupled spatial
levels
National1 Nation
Regional 17 Regions
Local225000 cellular units
28
Xplorah Processes at 3 strongly coupled spatial
levels
29
National Input-output economics linked to
Demography and Climate scenarios
This model calculates on a yearly basis the total
population of the island, the inputs, outputs and
employment of 5 aggregated economic sectors,
corrected on the basis of climate change scenarios
30
RegionalDynamic spatial interaction based
For each of the 17 regions, this model calculates
on a yearly basis the changing number of
inhabitants and jobs in the 5 economic sectors.
31
LocalConstrained Cellular Automata
Transition Rule Change cells to land-use for
which they have the highest transition potential
until Regional demands are met.
Time Loop
This model calculates on a yearly basis the
changing land use for 225,000 cells (250 m
resolution, 18 land use categories)
32
Exploring planning options
33
Spatial Planning with Xplorah
What are effects to be expected of zoning this
area for housing starting in 2005 ?
34
Xplorah for infrastructure planning

• Prototype of an integrated land
  use-transportation model based on the databases
  and requirements of the Puerto Rico DOT Highway
  and Transportation Authority.

35
Measuring the effects of traffic at high spatial
and temporal detail

• Noice pollution (gt 40dBA) in protected and
  silence zones
• Air pollution (NOx) due to private vehicles on
  motorways.

36
Infrastructure is more than roads!

• Example expansion of the drinking water pipeline
  system and its effects on urban sprawl.

37
Tools for the end-user
38
Monte Carlo-Tool The single run is not what
counts. Working with uncertainty

• Probability that the cell is occupied by
  particular land use as the result of uncertainty
  in parameter(s).

Not 1,but 10, 100, , runs, fluctuating 1, 2, ,
allparameters
39
OVERLAY-ToolGenerating suitability zoning
maps interactively
40
ANALYSE-ToolComparing and analysing map output
interactively
41
User-interface of the DSS
42
Extensive sensitivity analysis enables Goal
seeking
43
Development pains

• Access to high quality data
• Consistent and high quality map material is
  needed for historic period (identical mapping
  methodology, classes, grids, etc.)
• Historic high quality land use maps are missing
  in Puerto Rico and need to be produced
• Difficulties with linking regional local
  levels
• No easy match between classification schemes of
  census data and GIS / land use data
• Frequent need for (re)calibration
• Model is partly calibrated on the basis of
  historic data and analysis (period 1990-2000)
• Finding appropriate goodness-of-fit measures
• Fractal dimensions (White and Engelen, 1993)
• Fuzzy Inference System (Power et al., 2001)
• Fuzzy Kappa (Hagen, 2003)
• Major research effort for developing
  semi-automatic calibration routines
• Straatman, et al. 2003
• Hagen, Van Loon and Maes, 2004

44
An evaluation

• Is evaluated positively because of
• Its strategic, explorative, interactive and very
  explicit nature,
• Added value as a tool for analysis, discussion
  and communication
• Provides better insight in the interrelated
  nature of functions, processes, cause and effect
  relations
• Provides insight in the effects of policy
  interventions in the own discipline and that of
  others
• Enables the objective evaluation of the relative
  value of more alternatives than would otherwise
  be considered in a policy exercise
• Is evaluated less positively because of its
  complex nature.
• It models a complex reality and requires a
  minimum of knowledge of the domains represented
  by those using it. For many actively involved in
  the planning field this is beyond their capacity.

45
The END

• To find out more about
• RIKS, check
• http//www.riks.nl
• Xplorah, check
• Reports, publications, brochures,
• http//www.riks.nl/projects/Xplorah
• MOLAND, check
• http//www.riks.nl/projects/MOLAND