Ronald C. Estoque
(Student ID Number: 201030243)
D1 – Division of Spatial Information Science
2011
GIS-based Multi-Criteria Decision Analysis
(in Natural Resource Management)
Learning objective
 The aim of this lecture presentation is to present the basic
principles, methods, & some applications of GIS-based Multicriteria
Decision Analysis (MCDA).
2
Outline
 Introduction
 Basic principles of GIS-based MCDA
 Methods & applications of multi-criteria evaluation
 Remarks
3
Introduction
  Land is a scarce natural resource
  The demand is continuously growing
  to satisfy human beings’ basic needs, & insatiable wants & desires
  People must select the best use of this resource
  uphold sustainability to be able to sustain the benefits this resource provides for the
next generations to come
  However, people have different behaviors, beliefs, knowledge, priorities,
goals, interests & concerns
  decision-making on how a particular resource should be utilized is not an easy task
4
Basic principles of GIS-based MCDA
MCDA – Multi-Criteria Decision Analysis
  Problem
  spatial decision problems typically involve a large set of feasible alternatives & multiple
evaluation criteria
o  most of the time, these are conflicting
  alternatives & criteria are often evaluated by a number of individuals (decision-makers,
managers, stakeholders, interest groups).
o  most of the time, they also have conflicting ideas, preferences, objectives, etc.
  many spatial decision problems give rise to the GIS-based MCDA
o  to aid in the decision making process
  GIS
  techniques & procedures have an important role to play in analyzing decision problems
o  recognized as a decision support system involving the integration of spatially referenced data in a
problem solving environment
  MCDA
  provides a rich collection of techniques & procedures for structuring decision problems,
& designing, evaluating & prioritizing alternative decisions
  GIS-MCDA
  can be thought of as a process that transforms & combines geographical data & value
judgments (the decision-maker’s preferences) to obtain information for decision making
5 See Malczewski (2006)
Basic principles of GIS-based MCDA
6
Definitions
  Decision – is a choice between alternatives
  i.e. best land use among different land use alternatives
  Criteria
  are set of guidelines or requirements used as basis for a decision
  Two types: factors & constraints
o  A factor is a criterion that enhances or detracts from the suitability of a
specific alternative for the activity under consideration
  i.e. distance to road (near = most suitable; far = least suitable)
o  A constraint serves to limit the alternatives under consideration;
element or feature that represents limitations or restrictions; area that is
not preferred in any way or considered unsuitable.
  i.e.protected area, water body, etc. (usually represented by a Boolean mask)
See Eastman et al. (1995)
Basic principles of GIS-based MCDA
7
Definitions
  Decision rule
  procedure by which criteria are combined to arrive at a particular evaluation.
  1) Choice function – provides a mathematical means for comparing alternatives;
numerical, exact decision rules
  2) Choice heuristic – specifies a procedure to be followed rather than a function
  Objective
  the measure by which the decision rule operates (i.e. identify suitable areas for a
housing project)
  in a single-objective multi-criteria evaluation, it is also considered as a‘goal’
  Suitability
  is the characteristic of possessing the preferred attributes or requirements for a
specific purpose
  Suitability analysis
  is a GIS-based process used to determine the appropriateness of a given area (land
resource) for a specific use, i.e.agriculture, forestry, business, urban development,
livelihood projects, etc.
See Eastman et al. (1995)
Basic principles of GIS-based MCDA
o  Multi-Criteria Decision Making/Analysis (MCDM/MCDA)
o  Multi-Criteria Evaluation
o  sometimes it is also referred to as multi-attribute evaluation or Multi-Attribute Decision Analysis (MADA)
o  Example:site suitability analysis for housing development (specific single objective)
o  Multi-Objective Evaluation
o  sometimes it is also called as Multi-Objective Decision Analysis (MODA)
o  Example:analysis for best land use (forest,agriculture,residential,etc.) - multiple objectives
o 1) suitability analysis per land use; 2) Multi-Objective Land Allocation (MOLA)
8
GIS-based MCDA (Adopted from Eastman et al., 1995)
Criteria Factor1 Factor2 Factor3 Constraints
Suitable
land for
housing
project
Goal
Attributes Attributes Attributes Attributes Attributes
Example of Multi-Criteria Evaluation (MCE) structure
Methods of Multi-Criteria Evaluation (MCE)
Steps:
1. Set the goal/define the problem
2. Determine the criteria (factors/constraints)
3. Standardize the factors/criterion scores
4. Determine the weight of each factor
5.Aggregate the criteria
6.Validate/verify the result
9
 Steps
1. Set the goal/define the problem
  As a general rule, a goal must be:
S
M
A
R
T
– specific
– measurable
– attainable
– relevant
– time-bound
Source: Haugey, D.“SMART Goals”. Project Smart. www.projectsmart.co.uk/smart-goals.html
10
Methods of Multi-Criteria Evaluation (MCE)
Methods of Multi-Criteria Evaluation (MCE)
 Steps
2. Determine the criteria (factors/constraints)
  how much details are needed in the analysis affects the set of criteria to
be used
o  i.e. main roads only vs. including minor roads; no. of houses vs. no. of
residents; etc.
  Criteria should be measurable
  If not determinable, use proxies
o  i.e.slope stability can be represented by slope gradient
11
Methods of Multi-Criteria Evaluation (MCE)
 Steps
3. Standardize the factors/criterion scores
  Set the suitability values of the factors to a common scale to
make comparisons possible
o  it is hard to compare different things (i.e. mango vs. banana)
For example
  Elevation (m)
  Slope (%)
  Convert them to a common range, i.e. 0 – 255
  0 = least suitable; 255 = most suitable
high
most suitable - 255
least suitable - 0
Elevation (m)
low
Suitability
Slope (%)
most suitable - 255
least suitable - 0 low high
Suitability
12
Methods of MCE
 Steps
3. Standardize the factors/
criterion scores cont’d…
  Fuzzy Membership
Functions are used to
standardize the criterion
scores.
  Decision-makers have to
decide based on their
knowledge & fair judgment
which function should be
used for each criterion.
Source (Figures): Eastman (1999; 2006)
Linear Membership Function
13
Methods of Multi-Criteria Evaluation (MCE)
 Steps
4. Determine the weight of each factor
  There are several methods
o  Ranking
 i.e.3 factors: rank the factors with 1, 2, & 3, where 1 is the least important
while 3 is the most important
o  Rating
  i.e.3 factors: rate the factors using percentile – Factor 1 with the lowest
percentage as the least important & Factor 3 with the highest percentage
as the most important
o  Rankings & ratings are usually converted to numerical weights on a scale 0
to 1 with overall summation of 1 (normalization).
o  i.e.Factor 1 = 0.17; Factor 2 = 0.33; & Factor 3 = 0.50;
14
Methods of Multi-Criteria Evaluation (MCE)
 Steps
4. Determine the weight of each factor
  There are several methods (cont’d)
o  Pairwise comparison
Analytical Hierarchy Process (AHP) (Saaty, 1980)
  A matrix is constructed, where each criterion is compared with the other
criteria, relative to its importance, on a scale from 1 to 9.
  where 1 = equal preference between two factors; 9 = a particular factor is extremely
favored over the other
  a weight estimate is calculated & used to derive a consistency ratio (CR) of the
pairwise comparisons
•  If CR > 0.10, then some pairwise values need to be reconsidered & the process is
repeated until the desired value of CR < 0.10 is reached.
  Like in ranking & rating,AHP weights are also expressed in numerical weights
that sum up to 1.
15
4. Determine the weight of each factor (using AHP)
F1 F2 F3
F1 1 9 3
F2 1/9 1 1/5
F3 1/3 5 1
Σ 1.4444 15.0000 4.2000
Example: Using 3 factors
F1
13579 3 5 7 9
F2
F1
13579 3 5 7 9
F3
F2
13579 3 5 7 9
F3
Extreme
favors
Very
Strong
favors
Strongly
favors
Slightly
favors Equal
Slightly
favors
Strongly
favors
Very
Strong
favors
Extreme
favors
Note:2, 4, 6 & 8 are intermediate values. F1, F2, & F3 are factors.
Step 1 – Compare the factors
Basic rules:
1.  If the judgment value is on
the left side of 1, we put
the actual judgment
value.
2.  If the judgment value is on
the right side of 1, we put
the reciprocal value .
Step 2 – Complete the matrix
F1 F2 F3 Priority
vector* or
Weight
F1 0.6923 0.6000 0.7143 0.6689
F2 0.0769 0.0667 0.0476 0.0637
F3 0.2308 0.3333 0.2381 0.2674
Step 3 – Normalization & weight determination
*Priority vector is also called normalized principal Eigen vector.
• To normalize the values, divide the cell value by its column total.
• To calculate the priority vector or weight, determine the mean value
of the rows.
Step 4 – Calculate the Consistency Ratio (CR)
CR = Consistency index (CI)/Random Consistency Index (RI)
• CI = (λmax – n)/n – 1
• λmax is the Principal EigenValue; n is the number of factors
• λmax = Σ of the products between each element of the priority vector
and column totals.
• λmax = (1.44*0.67) + (15*0.06) + (4.20*0.27) = 3.0445
• CI = (3.0445 – 3)/3-1 CI = 0.0445/2 CI = 0.0222
• CR = 0.0222/0.58 CR = 0.04 < 0.10 (Acceptable)
Random Consistency Index (RI) (Saaty, 1980).
• n - 1 2 3 4 5 6 7 8 9 10
• RI - 0 0 0.58 0.90 1.12 1.24 1.32 1.41 1.45 1.49
16
Note: Values in blue are reciprocals.
Methods of Multi-Criteria Evaluation (MCE)
  Steps
5.Aggregate the criteria
  Weighted Linear Combination
o  is the most commonly used decision rule
Formula:
Where:
S – is the composite suitability score
xi – factor scores (cells)
wi – weights assigned to each factor
cj – constraints (or Boolean factors)
∑ -- sum of weighted factors
∏ -- product of constraints (1-suitable, 0-unsuitable)
S = ∑wixi x ∏cj
•  Example:
Applying it in GIS raster calculator
S =((F1 * 0.67) + (F2 * 0.06) + (F3 * 0.27)) * cons_boolean
Note:F1, F2, F3 & cons_boolean are thematic layers representing the factors & constraints.
17
Methods of Multi-Criteria Evaluation (MCE)
 Steps
6.Validate/verify the result
  to assess the reliability of the output
o  Ground truth verification
  i.e. conduct a field survey to verify sample areas
o  Sensitivity analysis
  How do the following affect the result?
  altering the set of criteria (plus or minus)
  altering the respective weights of the factors
  Is the result reasonable?
  Does the result reflect reality?
 Etc.
18
Examples: Applications of GIS-based MCE
1. Suitability analysis for beekeeping sites in La union, Philippines, using GIS &
MCE techniques
  Goal
– to produce a map showing the suitable areas for beekeeping
  Criteria
Factors
Categorical data
o  Land use/cover – scores/suitability values (0-255) assigned to the different land
uses/covers were based on the availability of source of nectar and pollen
Continuous data
o  Distance to river – standardized to 0-255 scale: suitability values decreases with the
distance to river
o  Distance to road – standardized to 0-255 scale: suitability values decreases with the
distance to road (starting from a buffer of 25 m).
o  Elevation – standardized to 0-255 scale: suitability values decreases with elevation
Constraints
o  built-up areas, sand, water body, riverwash, & areas within 25 m from the roads
For more details,see: Estoque, R C & Murayama,Y (2010). Suitability analysis for beekeeping sites in La union, Philippines, using GIS & MCE techniques.
Research Journal of Applied Sciences, 5, 242 – 253.
Also in: Murayama,Y &Thapa, R B (Eds), Spatial Analysis and Modeling in GeographicalTransformation Process:GIS-based Applications.
(as a book chapter) Dordrecht: Springer Science +Business Media B.V., isbn:978-94-007-0670-5
19
Suitability Map for Beekeeping
in La Union, Philippines
Using Raster Calculator: (Weighted Overlay tool can also be used)
S =((elevation * 0.0553) + (dist_river * 0.2622) + (dist_road * 0.1175) + (luc * 0.5650)) * cons_boolean
1. Suitability analysis for beekeeping sites using GIS & MCE techniques cont’d…
20
For more details,see: Estoque, R C & Murayama,Y (2010). Suitability analysis for beekeeping sites in La union, Philippines, using GIS & MCE techniques.
Research Journal of Applied Sciences, 5, 242 – 253.
Also in: Murayama,Y &Thapa, R B (Eds), Spatial Analysis and Modeling in GeographicalTransformation Process:GIS-based Applications.
(as a book chapter) Dordrecht: Springer Science +Business Media B.V., isbn:978-94-007-0670-5
Examples: Applications of GIS-based MCE
2. Suitability analysis for best site for a new school
  Goal
– to produce a map showing the best site for a new school
  Criteria
Factors
o  Distance to recreational sites – areas near to recreational facilities are
preferred
o  Distance to existing schools – areas away from existing schools are
preferred
o  Slope – areas on flat terrain are preferred
o  Land use – agricultural land is most preferred followed by barren land,
bush/transitional areas, forest, & built-up areas.
Constraints
o  Water & wetlands
For more details,see: ESRI (2007). Using the conceptual model to create a suitability map.ArcGISTutorial. webhelp.esri.com/arcgisdesktop/9.2/index.cfm?TopicName=tutorials
21
For more details,see: ESRI (2007). Using the conceptual model to create a suitability map.ArcGISTutorial.
webhelp.esri.com/arcgisdesktop/9.2/index.cfm?TopicName=tutorials
Distance to recreational
sites
Distance to existing
schools
Slope
Land use
Suitability map for best site for a new school
Using Raster Calculator: (Weighted Overlay can also be used)
S =((dist_rec * 0.50) + (dist_school * 0.25) + (slope * 0.125) + (luc * 0.125)) * cons_boolean
Note: In this
presentation,
“cons_bool” was
added to represent
the constraints.
2. Suitability analysis for best site for a new school cont’d…
22
Standardization method:
Continuous data – re-classed (10 classes) & assigned a suitability value per class using a scale of 1 – 10 (10 = most suitable; 1 = least suitable)
Categorical data (land use) - assigned a suitability value per class using a scale of 1 – 10 (10 = most suitable; 1 = least suitable)
Remarks
  GIS-based MCDA particularly MCE is good for complex scenarios.
  i.e. site/land suitability analysis – involves multiple criteria & a lot more
considerations
  GIS packages
  IDRISI Andes/Taiga have the following decision support modules: Fuzzy
(used to standardize factors),Weight (used to calculate the AHP weights),
MCE (for the actual evaluation), & a lot more.
  The whole MCE process can also be done inArcGIS (model builder)
although it may not have the standardization functionalities like what IDRISI
has.
  However, MCDA/MCE’s nature of being “participatory”
sometimes raises subjectivity.
  i.e.in choosing the criteria & defining the weights of each factor
23
References & Suggested Readings
 Eastman, R J (2006). Idrisi Andes: Guide to GIS and image processing. Clark Labs, Clark University,Worcester,
USA.
 Eastman, R J (1999). IDRISI 32 – Guide to GIS and Image Processing,Volume 2. Clark Labs, Clark University,
Worcester, USA.
 Eastman, R J, Jin,W, Kyem, P A, &Toledano, J (1995). Raster procedure for multi-criteria/multi-objective
decisions. Photogrammetric Engineering and Remote Sensing, 61, 539 – 547.
 ESRI (2007). Using the conceptual model to create a suitability map.ArcGISTutorial.Accessed January 11, 2011.
webhelp.esri.com/arcgisdesktop/9.2/index.cfm?TopicName=tutorials
 Estoque, R C and Murayama,Y (2011). Beekeeping sites suitability analysis integrating GIS and MCE techniques. In
Murayama,Y andThapa, R B (Eds), Spatial Analysis and Modeling in GeographicalTransformation Process:GIS-based
Applications. Dordrecht: Springer Science +Business Media B.V., isbn:978-94-007-0670-5  
 Estoque, R C & Murayama,Y (2010). Suitability analysis for beekeeping sites in La union, Philippines, using GIS &
MCE techniques.Research Journal of Applied Sciences, 5, 242 – 253.
 Haugey, D.“SMART Goals”. Project Smart.Accessed January 11, 2011.
www.projectsmart.co.uk/smart-goals.html
 Malczewski, J (2006). GIS-based multicriteria decision analysis: a survey of the literature. International Journal of
Geographical Information Science, 20, 703 – 726.
 Malczewski, J (2004). GIS-based land-use suitability analysis: a critical overview. Progress in Planning, 62, 3 – 65.
 Saaty,T L (1980). The Analytic Hierarchy Process. NewYork: McGraw Hill.
 Teknomo, K (2006).Analytic Hierarchy Process (AHP)Tutorial .Accessed January 11, 2011.
http://people.revoledu.com/kardi/tutorial/ahp/
24