1. 16. Spatial Modeling with GIS

2. Outline Why model? Types of model Technology for modeling Multicriteria methods Accuracy and validity

3. What Is a Model? A data model provides a template for representing the real world in a GIS For modeling how the world looks This chapter is about models that represent real processes Or how the world works

4. Spatial Models Represent variation over the Earth's surface Produce results that change when the locations of features change GIS provides a suitable platform Manipulate geographic information in multiple stages Representing either a single point in time, or predictions about future points in time

5. Spatial and Temporal Resolution Models work at defined levels of resolution All variation at finer resolutions is ignored The model must leave the user uncertain to some degree about the real world A model of a dynamic process has a temporal resolution equal to the size of its time steps

6. The Value of Models They allow experiments to be conducted on simulated systems, rather than the real thing Cheaper, less invasive They allow alternative scenarios to be evaluated Different policy options and their impacts on the future

7. To Analyze or to Model? Analysis is static Revealing what would otherwise be invisible Looking for anomalies, patterns Leading to hypotheses about systems Modeling can be dynamic Testing hypotheses about systems Evaluating alternative scenarios

8. Simulation of driver behavior in an evacuation of the Mission Canyon area of Santa Barbara, California, U.S.A., some minutes after the initial evacuation order. The red dots denote the individual vehicles whose behavior is modeled in this simulation.

9. Types of Model Static models and indicators Take multiple GIS inputs, and compute useful indices The Universal Soil Loss Equation predicting erosion from five inputs The DRASTIC groundwater vulnerability model predicting an index of vulnerability to pollution

10. The results of using the DRASTIC groundwater vulnerability model in an area of Ohio, U.S.A. The model combines GIS layers representing factors important in determining groundwater vulnerability, and displays the results as a map of vulnerability ratings.

11. Graphic representation of the groundwater protection model developed by Rhonda Pfaff and Alan Glennon for analysis of groundwater vulnerability in the Mammoth Cave watershed, Kentucky, U.S.A.

12. Results of the groundwater protection model. Highlighted areas are farmed for crops, on relatively steep slopes and within 300m of streams. Such areas are particularly likely to generate runoff contaminated by agricultural chemicals and soil erosion, and to impact adversely the cave environment into which the area drains.

13. Individual and Aggregate Models Individual models simulate the behavior of every individual in the system Every person in a crowd Aggregate models are used when there are too many individual elements to model Impossible to model every molecule of water, grain of sand

14. Simulation of the movement of individuals during a parade. Parade walkers are in white, watchers in red. The watchers (A) build up pressure on restraining barriers and crowd control personnel, and (B) break through into the parade.

15. Cellular Models Model a system using a raster Each cell can be in one of a number of states Change through time is represented by change of cell state Change is defined by a series of rules Depending on the state of the cell and its neighbors

16. Cellular automaton (CA) A cellular automaton is a discrete model studied in computability theory, mathematics, physics, complexity science, theoretical biology and microstructure modeling. It consists of a regular grid of cells, each in one of a finite number of states, such as "On" and "Off". The grid can be in any finite number of dimensions. For each cell, a set of cells called its neighborhood (usually including the cell itself) is defined relative to the specified cell. For example, the neighborhood of a cell might be defined as the set of cells a distance of 2 or less from the cell. An initial state (time t=0) is selected by assigning a state for each cell. A new generation is created (advancing t by 1), according to some fixed rule (generally, a mathematical function) that determines the new state of each cell in terms of the current state of the cell and the states of the cells in its neighborhood. For example, the rule might be that the cell is "On" in the next generation if exactly two of the cells in the neighborhood are "On" in the current generation, otherwise the cell is "Off" in the next generation. Typically, the rule for updating the state of cells is the same for each cell and does not change over time, and is applied to the whole grid simultaneously, though exceptions are known. Cellular automata are also called "cellular spaces", "tessellation automata", "homogeneous structures", "cellular structures", "tessellation structures", and "iterative arrays". [2] [2]

17. The Game of Life, also known simply as Life, is a cellular automaton devised by the British mathematician John Horton Conway in 1970.Game of Life The "game" is a zero-player game, meaning that its evolution is determined by its initial state, requiring no further input. One interacts with the Game of Life by creating an initial configuration and observing how it evolves. E.g.: Modern Cellular Automata

18. Cellular Models of Urban Growth Each cell is either undeveloped or developed Transitions may occur at each time step from undeveloped to developed Depending on the state of neighboring cells And on the characteristics of the cell slope, access to transportation, protected status, etc.

19. Simulation of future urban growth patterns in Santa Barbara, California, U.S.A. (Upper) Growth limited by current urban growth boundary; (Lower) growth limited only by existing parks.

20. Cartographic Modeling and Map Algebra Organizes all GIS operations on a raster into four types Local operations are determined by the attributes of each cell alone Focal operations are determined by a cell's neighbors Global operations compute properties of the entire raster layer Zonal operations apply to all contiguous cells with the same value

21. Technology for Modeling Scripts Sequences of GIS operations that can be stored and shared written in a scripting language such as Visual Basic for Applications, Perl, Python, or JScript a model can be written and executed as a script Scripts can be manipulated visually e.g., through ESRI's ModelBuilder

22. Coupling Models often exist as separate, stand-alone programs Can be used in conjunction with a GIS through various forms of coupling Loose coupling The GIS and the model exchange data in the form of files Close coupling Both GIS and the model read and write to the same file through a common interface

23. Multicriteria Methods Many decisions depend on numerous factors The factors must be combined in some fashion Often stakeholders disagree on how the factors should be combined Multicriteria methods attempt to reconcile such differences and to reach consensus

24. Saaty's Analytical Hierarchy Process (AHP) Each stakeholder compares each pair of factors Indicating relative importance as a ratio The ratings are combined to produce a consensus set of weights Along with a measure of the strength of agreement or disagreement

25. One Stakeholder's Weights An example using the three factors in the Pfaff and Glennon groundwater protection model SlopeLand useDistance from Stream Slope72 Land use1/71/3 Distance from Stream 1/23

26. Screen shot of an AHP application using IDRISI (www.clarklabs.org). The five layers in the upper left part of the screen represent five factors important to the decision. In the lower left the image shows the table of relative weights compiled by one stakeholder. All of the weights matrices are combined and analyzed to obtain the consensus weights shown in the lower right, together with measures to evaluate consistency among the stakeholders.

27. Accuracy and Validity How can the results of a model be assessed? Is the vulnerability index correct? Are the predictions of a scenario correct? A model is only as accurate as its inputs And the rules used to emulate real processes And subject to the limitations of its spatial and temporal resolution Models help to reduce uncertainty about the future But can never reduce it to zero

28. Error Propagation Methods for assessing the effects of known degrees of error in a model's inputs Producing measures of confidence in model outputs Normally by simulation

29. Sensitivity Analysis Used to assess the effects of uncertainty in model parameters and assumptions Systematically raise and lower the value of each parameter Observe the effects on model predictions Helps in identifying those parameters that are more important and need to be carefully examined