Spatial Modeling with GIS Longley et al., Chapter 16.

Spatial Modeling with GIS

Longley et al., Chapter 16

Spatial Modeling with GIS

• Introduction

• Types of Model

• Modeling Technology

• Multicriteria Methods

• Accuracy and Validity

Spatial modeling

• Modeling: An overworked term • data model a template for data

relational, object-oriented, coverage, shapefile

• Model concerned with how the world looks

• Model also a representation of some real-world process

• Concerned with how the world works

Spatial modeling

• Manipulation of geographic information in multiple steps

• Steps may represent stages in some complex analysis

• Calculation of indicators over space (potentials)• Steps may represent time • Dynamic model • Iterative analysis • Geocomputation (see www.geocomputation.org)

Analog or Digital Modeling?

• Analog use of a scale model

• Analogous process

• Varignon frame

• Need a digital process represented in 0s and 1s

• program in C • GIS script in VBA• Python

Scaled Real Models

Army Corps of Engineers:WES

Varignon Frame

“Live” table: Pollution demo

Scale in a digital model?

• Spatial resolution/extent

• Temporal resolution/extent

• Define what is left out of the model

• Leave out uncertainty about model data, predictions

• Model must run faster than the real world

• Ecological fallacy

Why model?

• Support some design process

• Allow the user to experiment with a replica

• Investigate what-if scenarios

• To understand change and dynamics

• Test sensitivity and confidence

Analysis vs. Modeling

• To analyze or model?

• Evacuation scenarios– Tom Cova's analysis – Church's simulations– LANL

Analysis

Modeling

LANL TRANSIMSIndividual vehicle-based traffic simulationof entire cities

Limits of Analysis

• Static, one point in time

• Search for patterns, anomalies

• Generating hypotheses

• Revealing what would otherwise be invisible

• Form vs. process

Modeling multiple stages

• Perhaps different points in time

• Implementing ideas and hypotheses

• Experimenting with policy options

• Scenario based planning

Types of Model

• Static models and indicators

• Combining GIS layers through overlay e.g., using ModelBuilder

• Universal Soil Loss Equation

• A = R x K x LS x C x P

• DRASTIC model of groundwater vulnerability

• Karst groundwater protection model

• DRASTIC

Santa Barbara Regional Impacts of Growth Study: 2040 forecasts

Karst groundwater protection model in Model Builder

Model result

Modeling Approach

• Individual vs. Aggregate models

• Is it possible to model every individual element in the system?

• Every molecule of groundwater? Every person in a crowd?

• Autonomous agent models

Mass Behavior: Problems

Twenty-one Hajj pilgrims trampledWednesday, February 12, 2003 Posted: 2:33 PM

EST (1933 GMT)MINA, Saudi Arabia -- Another 21 people were trampled to death Wednesday on their way to one of the rituals of the Hajj, the annual Muslim pilgrimage to Mecca, Saudi officials said. Wednesday's deaths happened on a bridge as the throngs of pilgrims were heading to throw stones at one of three pillars representing Satan's temptation of Abraham, the officials said. The stoning represents a rejection of evil deeds. On Tuesday, a similar incident killed 14 pilgrims.

Notting Hill Carnival

Cellular Models

• Work on a raster: Good match to GIS

• Initial conditions

• Each cell in one of a number of states

• Rules of state transition at each timestep based on states of cell and neighbors

• Conway’s Game of Life

• SLEUTH land use change model

(Universal) Turing machine

Cellular automata

• Framework for systems experiments

• Simplest way to demonstrate complex systems behavior

• Wolfram: Formal framework

• {Cells, States, Initial conditions, Neighborhood, Rules, Time}

• Conway’s LIFE

The game of life

• Grid of square cells extending infinitely in every direction. • A cell can be live or dead. • Each cell in the grid has a neighborhood consisting of the

eight cells in every direction including diagonals. • To apply one step of the rules, we count the number of live

neighbors for each cell. – A dead cell with exactly three live neighbors becomes a live cell

(birth). – A live cell with two or three live neighbors stays alive (survival). – In all other cases, a cell dies or remains dead (overcrowding or

loneliness).

Some examples

More examples

Urban Growth as a CA

Behavior Rules

T0 T1

For i time periods (years)

spontaneousspreading

center organicroad

influenced deltatron

f (slope resistance, diffusion

coefficient)

f (slope resistance,

breed coefficient)

f (slope resistance,

spread coefficient)

f (slope resistance, diffusion coefficient,

breed coefficient,road gravity)

SLEUTH applied to Santa BarbaraUrban growth to 2040

No new roads

Upgrade all local roads

Technology for Modeling in GIS

• Graphic user interface e.g. GISMO in ERDAS• ModelBuilder

– access to all ArcGIS functions – no looping at present

• Scripts ARC/INFO AML • ArcView 3.x Avenue • ArcGIS

– Visual Basic for Applications – Perl – Python – JScript – ArcScripts

Model Coupling

• linking model software to GIS • Loose coupling

– Exchanging files– Entering results

• Tight coupling – Common files– Common interface– Common code

• Modeling languages

Multcriteria Methods

• Multiple factors affect decisions• Weighted by difference levels of importance• Karst case

– slope > 5%– land use = cropping – distance from stream < 300m

• Simple binary decision • How to assign weights to each factor?• Stakeholders may disagree on weights • MCDM = multicriteria decision making

Analytical Hierarchy Process

• Devised by Thomas Saaty

• Each stakeholder compares each pair of factors

• Assigns comparative weights – e.g., slope 7 times as

important as land use – e.g., distance from stream

1/2 as important as slope • forming a complete matrix• Weights must sum to one

Slope Land use Distance from Stream

Slope 7 2

Land use 1/7 1/3

Distance from Stream

1/2 3

AHP example:Idrisi

Model accuracy and validity

• How do we know if the model is correct?• How do we know that forecasts are accurate?• Results from a computer are often trusted

implicitly • How to calibrate the model?

– Hindcasting– Boostrapping

• A model is never more than an approximation to reality but how good/bad is the approximation?

• Important to provide measures of confidence in results

Sensitivity testing

• Varying the inputs to observe effects on outputs • Some inputs affect outputs more than others • These are the inputs that most need to be

correct • Error propagation • Examining the impacts of input errors on outputs • Mostly by simulation