Вы находитесь на странице: 1из 60

Upcoming

Conferences
URISA’s 46th Annual Conference &
Exposition
October 7–10, 2008 — New Orleans, LA
w w w. u r i s a . o r g

URISA Leadership Academy


December 8–12, 2008 — Seattle, WA

13th Annual GIS/CAMA Technologies


Conference
February 8–11, 2009 — Charleston, SC

URISA’s Second GIS in Public Health


Conference
June 5–8, 2009 — Providence, RI

URISA/NENA Addressing Conference


TBD

URISA’s 47th Annual Conference &


Exposition
September 29–October 2, 2009 — Anaheim, CA

GIS in Transit Conference


November 11–13, 2009 — St Petersburg, FL
Volume 20 • No. 1 • 2008
Journal of the Urban and Regional Information Systems Association

Contents
Refereed

5 Automatic Generation of High-Quality Three-Dimensional Urban Buildings


from Aerial Images
Ahmed F. Elaksher and James S. Bethel

15 Robust Principal Component Analysis and Geographically Weighted Regression:


Urbanization in the Twin Cities Metropolitan Area of Minnesota
Debarchana Ghosh and Steven M. Manson

27 Where Are They? A Spatial Inquiry of Sex Offenders in Brazos County


Praveen Maghelal, Miriam Olivares, Douglas Wunneburger, and Gustavo Roman

35 Tools And Methods For A Transportation Household Survey


Martin Trépanier, Robert Chapleau, and Catherine Morency

45 Mapping Land-Use/Land-Cover Change in the Olomouc Region, Czech


Republic
Tomáš Václavík

Plus

53 Mapping the Future Success of Public Education


Journal

Publisher: Urban and Regional Information Systems Association

Editor-in-Chief: Jochen Albrecht

Journal Coordinator: Scott A. Grams

Electronic Journal: http://www.urisa.org/journal.htm


EDITORIAL OFFICE: Urban and Regional Information Systems Association, 1460 Renaissance Drive, Suite 305, Park Ridge, Illinois 60068-1348;
Voice (847) 824-6300; Fax (847) 824-6363; E-mail info@urisa.org.

SUBMISSIONS: This publication accepts from authors an exclusive right of first publication to their article plus an accompanying grant of non-
exclusive full rights. The publisher requires that full credit for first publication in the URISA Journal is provided in any subsequent electronic or
print publications. For more information, the “Manuscript Submission Guidelines for Refereed Articles” is available on our website, www.urisa.
org, or by calling (847) 824-6300.

SUBSCRIPTION AND ADVERTISING: All correspondence about advertising, subscriptions, and URISA memberships should be directed to:
Urban and Regional Information Systems Association, 1460 Renaissance Dr., Suite 305, Park Ridge, Illinois, 60068-1348; Voice (847) 824-6300;
Fax (847) 824-6363; E-mail info@urisa.org.

URISA Journal is published two times a year by the Urban and Regional Information Systems Association.

© 2008 by the Urban and Regional Information Systems Association. Authorization to photocopy items for internal or personal use, or the internal
or personal use of specific clients, is granted by permission of the Urban and Regional Information Systems Association.

Educational programs planned and presented by URISA provide attendees with relevant and rewarding continuing education experience. However,
neither the content (whether written or oral) of any course, seminar, or other presentation, nor the use of a specific product in conjunction there-
with, nor the exhibition of any materials by any party coincident with the educational event, should be construed as indicating endorsement or
approval of the views presented, the products used, or the materials exhibited by URISA, or by its committees, Special Interest Groups, Chapters,
or other commissions.

SUBSCRIPTION RATE: One year: $295 business, libraries, government agencies, and public institutions. Individuals interested in subscriptions
should contact URISA for membership information.

US ISSN 1045-8077

2 URISA Journal • Vol. 20, No. 1 • 2008


Editors and Review Board
Kingsley E. Haynes, Public Policy and
URISA Journal Editor Article Review Board Geography, George Mason University
Eric J. Heikkila, School of Policy, Planning, and
Editor-in-Chief
Peggy Agouris, Department of Spatial Information Development, University of Southern California
Jochen Albrecht, Department of Science and Engineering, University of Maine Stephen C. Hirtle, Department of Information
Geography, Hunter College City University Science and Telecommunications, University of
of New York Grenville Barnes, Geomatics Program, University
Pittsburgh
of Florida
Gary Jeffress, Department of Geographical
Michael Batty, Centre for Advanced Spatial Analysis,
Information Science, Texas A&M University-
University College London (United Kingdom)
Thematic Editors Corpus Christi
Kate Beard, Department of Spatial
Richard E. Klosterman, Department of
Editor-Urban and Regional Information Information Science and Engineering, Geography and Planning, University of Akron
Science University of Maine
Vacant Robert Laurini, Claude Bernard University of
Yvan Bédard, Centre for Research in Geomatics, Lyon (France)
Editor-Applications Research Laval University (Canada) 
Thomas M. Lillesand, Environmental
Lyna Wiggins, Department of Planning, Barbara P. Buttenfield, Department of Remote Sensing Center, University of Wisconsin-
Rutgers University Geography, University of Colorado Madison
Editor-Social, Organizational, Legal, Keith C. Clarke, Department of Geography,
and Economic Sciences
Paul Longley, Centre for Advanced Spatial Analysis,
University of California-Santa Barbara University College, London (United Kingdom)
Ian Masser, Department of Urban Planning
and Management, ITC (Netherlands) David Coleman, Department of Geodesy and Xavier R. Lopez, Oracle Corporation
Geomatics Engineering, University of New
Editor-Geographic Information Science David Maguire, Environmental Systems Research
Brunswick (Canada) Institute
Mark Harrower, Department of Geography,
University of Wisconsin Madison David J. Cowen, Department of Geography, Harvey J. Miller, Department of Geography,
University of South Carolina University of Utah
Editor-Information and Media Sciences
Michael Shiffer, Department of Planning, Massimo Craglia, Department of Town & Zorica Nedovic-Budic, Department of Urban
Massachusetts Institute of Technology Regional Planning, University of Sheffield and Regional Planning,University of Illinois-
(United Kingdom) Champaign/Urbana
Editor-Spatial Data Acquisition and
Integration William J. Craig, Center for Urban and Atsuyuki Okabe, Department of Urban
Gary Hunter, Department of Geomatics, Regional Affairs, University of Minnesota Engineering, University of Tokyo (Japan)
University of Melbourne (Australia) Robert G. Cromley, Department of Geography, Harlan Onsrud, Spatial Information Science
Editor-Geography, Cartography, and University of Connecticut and Engineering, University of Maine
Cognitive Science Kenneth J. Dueker, Urban Studies and Jeffrey K. Pinto, School of Business, Penn State Erie
Vacant Planning, Portland State University Gerard Rushton, Department of Geography,
Editor-Education Geoffrey Dutton, Spatial Effects University of Iowa
Karen Kemp, Director, International Masters
Max J. Egenhofer, Department of Spatial Information Jie Shan, School of Civil Engineering,
Program in GIS, University of Redlands
Science and Engineering, University of Maine Purdue University
Manfred Ehlers, Research Center for Bruce D. Spear, Federal Highway Administration
Geoinformatics and Remote Sensing, University of
Section Editors Osnabrueck (Germany)
Jonathan Sperling, Policy Development &
Research, U.S. Department of Housing and
Manfred M. Fischer, Economics, Geography & Urban Development
Software Review Editor
Geoinformatics, Vienna University of Economics David J. Unwin, School of Geography, Birkbeck
Jay Lee, Department of Geography, Kent State
and Business Administration (Austria) College, London (United Kingdom)
University
Myke Gluck, Department of Math and Stephen J. Ventura, Department of
Book Review Editor
Computer Science, Virginia Military Institute Environmental Studies and Soil Science,
David Tulloch, Department of Landscape
Michael Goodchild, Department of Geography, University of Wisconsin-Madison
Architecture, Rutgers University
University of California-Santa Barbara Nancy von Meyer, Fairview Industries
Michael Gould, Department of Information Barry Wellar, Department of Geography,
Systems Universitat Jaume I (Spain) University of Ottawa (Canada)
Daniel A. Griffith, Department of Geography, Michael F. Worboys, Department of Computer
Syracuse University Science, Keele University (United Kingdom)
Francis J. Harvey, Department of Geography, F. Benjamin Zhan, Department of Geography,
University of Minnesota Texas State University-San Marcos

URISA Journal • Vol. 20, No. 1 • 2008 3


Automatic Generation of High-Quality Three-Dimensional
Urban Buildings from Aerial Images
Ahmed F. Elaksher and James S. Bethel

Abstract: High-quality three-dimensional building databases are essential inputs for urban area geographic information systems.
Because manual generation of these databases is extremely costly and time-consuming, the development of automated algorithms
is greatly needed. This article presents a new algorithm to automatically extract accurate and reliable three-dimensional build-
ing information. High overlapping aerial images are used as input to the algorithm. Radiometric and geometric properties of
buildings are utilized to distinguish building roof regions in the images. This is accomplished with image segmentation and
neural network techniques. A rule-based system is employed to extract the vertices of the roof polygons in all images. Photogram-
metric mathematical models are used to generate the roof topology and compute the three-dimensional coordinates of the roof
vertices. The algorithm is tested on 30 buildings in a complex urban scene. Results showed that 95 percent of the building roofs
are extracted correctly. The root-mean-square error for the extracted building vertices is 0.35 meter using 1:4000 scale aerial
photographs scanned at 30 microns.

Introduction objects. Each model has several pose and shape parameters. The
parameters are estimated by fitting the model to the images using
Three-dimension building information is required for a variety least-squares adjustment. The algorithm is limited to parametric
of applications, such as urban planning, mobile communication, models only. In Shmid and Zisserman (2000), lines are extracted
visual simulation, visualization, and cartography. Automatic in the images and matched over multiple views in a pair-wise
generation of this information is one of the most challenging mode. Each line then is assigned two planes, one plane on each
problems in photogrammetry, image understanding, computer side. The planes are rotated, and the best-fitting plane is found.
vision, and GIS communities. Current automated algorithms Planes then are intersected to find the intersection lines.
have shown some progress in this area. However, some deficien- Henricsson et al. (1996) presented another system to extract
cies still remain in these algorithms. This is particularly apparent suburban roofs from aerial images by combining two-dimensional
in comparison to manual extraction techniques, which, although edge information together with photometric and chromatic
slow, are essentially perfect in accuracy and completeness. attributes. Edges are extracted in the images and aggregated to
Recent research covers extracting building information from form coherent contour line segments. Photometric and chromatic
high-resolution satellite imageries, high-quality digital elevation contour attributes for adjacent regions around each contour are
models (DEMs), and aerial images. For example, QuickBird and assigned to it. For each contour, attributes are computed-based
IKONOS high-resolution satellite imageries are used to acquire on the luminance, color, proximity, and orientation, and saved
planemetric building information with one-meter horizontal ac- for the next step. Contour segments then are matched using their
curacy (Theng 2006, Lee et al. 2003). However, aerial images are attributes. Segments in three dimension are grouped and merged
the primary source used to acquire accurate and reliable geospatial according to an initial set of plane hypotheses.
information. Lin and Nevatia (1998) proposed an algorithm to Fischer et al. (1997) extracted three-dimensional buildings
extract building wireframes from a single image. However, a from aerial images using a generic modeling approach that depends
single image does not provide any depth information. A pair of on combining building parts. The process starts by extracting
stereo images could also be used to extract building information low-level image features: points, lines, regions, and their mutual
(Avrahami et al. 2004, Chein and Hsu 2000). Using one pair relations. These features are used to generate three-dimensinal
of images is insufficient to extract the entire building because of building part hypotheses in a bottom-up process. A step-wise
hidden features that are not projected into the image pair. model-driven aggregation process combines the three-dimensional
Kim et al. (2001) presented a model-based approach to building feature aggregates to three-dimensional parameterized
generate buildings from multiple images. Three-dimensional building parts and then to a more complex building descriptor.
rooftop hypotheses are generated using three-dimensional roof The resulting complex three-dimensional building hypothesis
boundaries and corners extracted from multiple images. The gen- then is back-projected to the images to allow component-based
erated hypotheses then are employed to extract buildings using an hypothesis verification.
expandable Bayesian network. Wang and Tseng (2004) proposed A semiautomated approach is used in Förstner (1999) to
a semiautomatic approach to extract buildings from multiple solve the building-extraction problem. First, the user has to define
views. They proposed an abstract floating model to represent real the building model and find the building elements in one image

URISA Journal • Elaksher, Bethel 5


by a number of mouse clicks. Then the algorithm finds the cor- to provide surface texture.
responding features in other images and matches them to build In this article, a new algorithm to extract building wire-
the three-dimensional wireframe of the building. This approach frames using more than two images is presented. The algorithm
supports the extraction of more complex buildings; however, it has the capability to extract a wide range of buildings with dif-
requires the user to spend a great amount of time interacting ferent shapes, orientations, and heights. The human operator
with the system. Rottensteiner (2000) presented a semiautomated needs only to select an image patch for the building in the first
building-extraction technique in which the user can select an image, specify the location of the input data files, and set up the
appropriate building primitive from a database and then adjust thresholds. The algorithm will then find corresponding patches
the parameters of the primitive to the actual data by interactively in other images, using the input data, and start the extraction
measuring points in the digital images, and determine the final process. The algorithm is implemented using computer vision
building parameters by automated matching tools. techniques, artificial intelligence algorithms, and rigorous pho-
High-quality DEMs such as those available from light togrammetric mathematical models. Computer vision techniques
detection and ranging (LIDAR) have been used to generate are employed to extract image regions using a modified version
three-dimensional building models. Tse et al. (2006) proposed an of the split-and-merge image segmentation technique. Artificial
algorithm based on segmenting the raw data into high and low intelligent algorithms are used to discriminate roof regions and to
regions, and then modeling the walls and roofs by extruding the convert the image regions to two-dimensional polygons. Photo-
triangulated terrain surface (TIN) using CAD-type Euler opera- grammetric mathematical models are employed to simultaneously
tors. Tarsha-Kurdi et al. (2006) proposed another algorithm that and rigorously match image polygons and vertices across all views.
discriminates terrain and off-terrain clouds in a LIDAR point One of the powerful tools that photogrammetry provides is to
cloud. Then it categorizes the off-terrain points to building and simultaneously match features across multiple images. Inputs
vegetation subclasses. Building points then are detected via seg- are four images for the building. The algorithm has been tested
menting the original LIDAR three-dimensional point cloud. In on a large sample of buildings selected quasi-randomly from the
Brunn and Weidner (1997), the digital surface models (DSMs) Purdue University campus. Four images are used for each build-
are extracted using mathematical morphology. The differences ing and the automatically extracted wireframes of the extracted
between the DEMs and the DSMs are computed and building buildings are presented. Results show significant improvement in
points are detected by thresholding these differences. In the next the detection rate and accuracy and suggest the completeness and
step, building wireframes are generated using parametric and accuracy of the proposed algorithm. The remainder of this paper
prismatic models depending on the complexity of the detected is organized as follows. First, the process of extracting building
building. Morgan and Habib (2002) proposed another algorithm polygons in aerial images is presented. Then the generation of
for building detection that has the following steps: segmenta- three-dimensional building models is proposed. Results are given
tion of laser points, classification of laser segments, generation in the next section, followed by discussions and conclusions.
of building hypothesis, verification of building hypothesis, and
extraction of building parameters. Extracting Building
Several researchers worked on integrating LIDAR and
aerial images for building extraction. The approach presented
Polygons in Aerial Images
in Hongjian and Shiqiang (2006) is based on aerial images and
sparse laser scanning sample points. Linear features are extracted Image Region Extraction
in the aerial images first. Bidirection projection histogram and Several researchers worked on segmenting aerial and satellite im-
line matching then are used to extract the contours of buildings. ages in urban environments. Muller and Zaum (2005) proposed
The height of the building is determined from sparse laser sample an algorithm to detect and classify buildings from a single aerial
points that are within the contours of the buildings extracted image using a region-growing algorithm. Lari and Ebadi (2007)
from the images. proposed another segmentation algorithm to detect building re-
The presented systems display many deficiencies. Satellite gions in satellite images. The results, although applied to a single
imageries still do not provide high-quality elevation data. Several image, showed the significance of implementing segmentation
systems require human interaction. Using a parametric model strategies to detect buildings in aerial and satellite images. In this
to represent buildings limits some systems to specific building research, a modified split-and-merge image segmentation process,
models. Another problem is the excessive reliance on primitive Horowitz and Pavlidis (1974) and Samet (1982), is applied to
features such as corner points or line segments. Naive matching segment the aerial images. This technique obtains good results if a
of such primitive features yields numerous false matches, and high contrast exists between the foreground and the background
misses many correct ones. Systems using more than a pair of im- objects and if the segmented objects are internally homogenous.
ages perform the feature matching in a pair-wise mode. Systems For urban aerial images this is not always the case. Objects such
utilizing only DEM in building extraction start by segmenting the as roads, cars, trees, and buildings are common to a typical urban
DEMs. This process is problematic because of outliers and spikes. aerial image. Although this wide variety of objects is expected to
In addition, such DEMs are expensive to collect and insufficient be seen in aerial images, building roofs have an important attri-

6 URISA Journal • Vol. 20, No. 1 • 2008


bute that distinguishes them from other spatial features. Building zero. A region-growing algorithm then is used to connect all
roofs often are homogenous objects that can be segmented from the pixels that do not belong to either the background or the
other image features. Homogeneity is a scale-dependent attribute, region. These pixels are described as holes and are attached to
but for small-scale images (1:4000), roof regions often appear the original region. Another problem also observed in the results
homogeneous. However, utility pipelines and ducts can disturb is the splitting of some roof patches into two or more regions.
the building roof homogeneity. Texture is another problem that To overcome this problem, an average intensity is computed for
can decrease the ability to segment building roofs. To account for each region and any two neighboring regions are merged if the
these problems, several modifications are proposed to the original difference between their average intensities is smaller than a given
split-and-merge algorithm. The algorithm presented differs from threshold, regardless of their sizes. Small regions, very dark regions,
the conventional split-and-merge algorithm in its capability to and very bright regions are eliminated. Thresholds for the region
join neighboring regions based on their intensity and size differ- merging and elimination are kept fixed relative, at this stage, to
ences and in its potential for detecting and filling region gaps. the intensity range of the images and the image size. The results
The segmentation process is implemented as follows. The of the segmentation process for one sample building are shown
image first is divided into smaller regions until a homogeneity in Figure 1. Figure 2 shows the results of segmenting the images
condition is satisfied. This is implemented by constructing a of five other buildings.
quadtree of the image, progressing down the tree, splitting as
necessary when inhomogeneities exist. Then a merging algorithm Region Classification
is implemented. The merging is carried out in two steps. Adjacent Buildings usually are elevated blobs in the DEMs. On the other
regions are first merged based on the differences between the mini- hand, building regions possess linear borders. Other features
mum and maximum intensities. Large regions then are merged such as trees also are elevated; however, they do not have linear
with their small neighbors if the differences in their intensities are borders. Roads and sidewalks have linear borders, but are not
smaller than a given threshold. Intensity thresholds for splitting elevated. Therefore, in this research the high elevation and border
and merging range between 10 and 15, while the size threshold linearity attributes are used to discriminate building roof regions
is kept fixed relative to the image size. from other regions using a neural network. Each image region is
One of the problems noticed after segmenting the image is assigned two attributes for the discrimination process. The first
the presence of holes inside the segmented regions. This can oc- attribute measures the linearity of the region boundaries, while
cur because of texture and/or utility features. Holes are detected the second attribute measures the percentage of the points in the
and removed as follows. First for each region its pixels are located region that are above a certain height.
and copied to a template image with a background intensity of Border linearity is measured using a modified version of
Hough transformation (Hough 1962). First, border points are
extracted and sorted so they traverse the border clockwise. For
each border point, the previous five points and the next five points
are found to form a local line at each point. The adjusted line
parameters (αa, Pa) and the quadratic form of the residuals for
the local line at each point are computed using the least-squares
estimation technique. The algorithm is implemented in three runs
at each point; the first run is when the point is in the middle of the
line; in the second run, the local line is shifted so that the point
is at the end. In the third run, the local line is shifted so that the
(a) (b) (c) specified point is the first point. If the minimum quadratic form
Figure 1. Split-and-merge segmentation results for the image of one value is small, the parameter space cell at the location of the local
building: (a) original image, (b) image after splitting, (c) final regions line parameters, i.e., αa, Pa, is increased by one. The parameter

(a) (b)
Figure 2. Split-and-merge segmentation results for the images of five buildings: (a) original images, (b) final regions

URISA Journal • Elaksher, Bethel 7


(a) (b)
Figure 3. The parameter space for a roof region (a) and a nonroof region (b)

(a) (b)
Figure 4. DEM (a) and DSM (b) for an area with several buildings, same vertical scale

space then is searched and analyzed to determine a measure for repeated iteratively until the DSM is extracted. The differences
border linearity. Border linearity is measured as the percentage of between the DEM and the DSM then are computed and used to
the number of points in the larger four cells to the total number of represent height information. The use of the height information
border points. Figure 3 shows a parameter space for a roof region in preference to the elevations makes the algorithm applicable for
and another parameter space for a nonroof region. both flat and slope terrains. Figure 4 shows the DEM and DSM
A digital elevation model (DEM) is used to quantify the of an area with several buildings.
height of each region. First, the digital surface model (DSM), Each point in the image then is assigned a height value by
i.e., representing bare ground, is extracted. Minimum filters are projecting the differences between the DEM and the DSM back
used to perform this task (Masaharu and Ohtsubo 2002, Wack to the image using the image registration information, the pixel
and Wimmer 2002). The filtering process detects and conse- location in the image, and the DEM elevation. For each image
quently removes points above the ground surface to recognize point a ray is generated, starting from the exposure station of the
high points in the data set. The minimum filter size should be camera and directed toward the point. The intersection between
large enough to include data points that are not noise. However, the ray and the DEM defines the location of the corresponding
iterative approaches could be used to avoid the effect of noise. DEM post. The height information at this location then is used
In this research, the size of the filter is 9x9 pixels. The filtering is as the height of the corresponding image point. The region height

8 URISA Journal • Vol. 20, No. 1 • 2008


measure is defined as the percentage of the number of points in Image Polygon Extraction
the region that are above a certain height to the total number of The two-dimensional modified Hough space is utilized in extract-
points in the same region. ing the borderlines for the roof regions. Given all points contrib-
The neural network implemented in this research is a feed- uting to a certain cell, a nonlinear least-squares estimation model
forward back-propagation network (see Figure 5). The network is used to adjust the line parameters of that cell. Lines then are
consists of three layers: an input layer, one hidden layer, and an grouped recursively until no more lines with similar parameters
output layer. The number of neurons in the first layer is two. are left and short lines are rejected. The next step is to convert the
The number of neurons in the second layer is selected to be ten. extracted lines to polygons via a rule-based system. The rules are
The number of neurons in the third, i.e., last, layer is one. The designed as complex as possible to cover a wide range of polygons.
output of this neuron is either one in case the region is a roof The mechanism works in three steps. The first step is to find all
region or zero in case the region is not a roof region. The activa- the possible intersections between the borderlines. However, if
tion function for all neurons in the first and second layers is the two lines are almost parallel, i.e., intersecting angle out of the
sigmoid functions (Principe et al. 1999). For the output neuron, range (30o—150o), the intersection point is not considered. The
the step function is chosen as the activation function. To study the next step is to generate a number of polygons from all recorded
performance of the neural network, a variety of training data sets intersections. Each combination of three and four intersection
are used with different sizes: 20, 50, 100, 200, and 400 samples, points is considered a polygon hypothesis. Hypotheses are ignored
including 2, 5, 10, 20, and 40 roof samples, respectively, while if the difference in area between the region and the hypothesized
the other samples are nonroof samples. The average detection polygon is large. The best polygon that represents the region is
rate and false alarm rate for each training data set is recorded and chosen from the remaining hypotheses using a template-matching
shown in Figure 6. Results show that increasing the size of the process. The template is chosen to be the region itself, while it is
training data set does not affect the detection rate significantly. matched across all polygon hypotheses. The hypothesis with the
However, increasing the size of the training data set does have a largest correlation and minimum number of vertices is chosen to
significant effect on the false alarm rate. be the best-fitting polygon. The extracted polygons for six sample
buildings are shown in Figure 7. This algorithm succeeded in
overcoming several limitations of the segmentation process, such
as partially occluded regions, overshooting and undershooting
borders, and incomplete regions. This is observed by comparing
Figures 1, 2, and 7.

Three-Dimensional Polygon
Extraction

Polygon Correspondence
After extracting the building roof polygons from the images, poly-
gon correspondence should be established. A new technique to
Figure 5. The implemented neural network find corresponding polygons based on their geometrical properties

(a) (b)
Figure 6. The detection rate (a) and the false alarm rate (b)

URISA Journal • Elaksher, Bethel 9


Figure 7. Extracted image polygons for six sample buildings

equations (Mikhail et al. 2001) can be written using the least-


squares estimation technique. After determining the object space
coordinates of the intersecting point, the quadratic form of the
residuals is computed. If the four polygons across all views are
corresponding, then the quadratic form should be small, otherwise
it would be large. Therefore the quadratic form for each polygon
correspondence set serves as its cost. A four-dimensional array
is constructed to store the corresponding cost values. The array
Figure 8. The roof vertices of one building, before reconstructing the dimension is the same as the number of images. Each axis in the
roof topology array represents the polygons of one image. The residual cost value
is stored at the corresponding location in the four-dimensional
array. The axis with the largest number of polygons is used as the
reference axis and for each polygon on this axis; its subspaces array
is used to find its corresponding polygons in the other images.
This subspace array is searched and its minimum value defines the
corresponded sets. The minimum value indexes for the subspace
array are used as the corresponding polygons in other images. If
a polygon corresponds to more than one set, the total residual
cost is used to solve such cases.

Computing the Three-Dimensional Polygon


Figure 9. The roof vertices of one building, after reconstructing the Coordinates
roof topology After finding the corresponding polygons, the three-dimensional
coordinates of each roof polygon center are computed; however,
the correspondence problem between the polygon vertices is not
solved. To solve this problem, another least-squares adjustment
is developed and presented in this section. All possible polygon model is implemented. The input observations are the image
corresponding combination sets are considered, and for each coordinates of the polygon vertices and the unknowns are the
combination a correspondence cost is computed. The computed object space coordinates for the three-dimensional polygon ver-
cost is used to determine the best corresponding polygon set. tices. To solve the correspondence relations between the vertices
First, the coordinates of each two-dimensional polygon center within a group of corresponding polygons, the following process
are computed from the coordinates of its vertices. All possible is implemented. For each polygon correspondence set, all vertex
polygon correspondence combinations are exhaustively enu- combinations are considered, and for each combination, the three-
merated and considered, and for each combination the polygon dimensional coordinates for the polygon vertices are computed.
centers are matched across all available views. Because more than Each set includes a number of subsets (three subsets for triangles
one pair of images is available, a function of the residuals of the and four for quadrilaterals); each subset includes a group of hy-
image coordinates can be calculated and used as the matching pothesized matching vertices. The quadratic form is computed
cost. For each combination of four polygons, four collinearity for each subset, then they are added to create the total residual

10 URISA Journal • Vol. 20, No. 1 • 2008


of the set. The quadratic forms of the sets then are compared and
the set with the minimum quadratic form is selected.

Roof Topology Reconstruction


After finding the corresponding vertices, the three-dimensional
coordinates for each vertex are computed; however, the building
topology is not yet constructed as shown in Figure 8. A geometri-
cally constrained least-squares model is implemented to refine the
locations of the polygon vertices and to construct the building
topology. The input observations are the image coordinates of the
polygon vertices; the unknowns are the object space coordinates
for the three-dimensional vertices. The aim of this step is to convert
groups of neighboring vertices into one vertex, adjust the elevations
of horizontal points, and reconstruct the correct relation between
adjacent polygons. The following constraints are used:
1. Nearby vertices should be grouped into one vertex: The resultant
vertices from the previous step need to be grouped, as shown
in Figure 8, for the extracted three-dimensional polygons may
not be contiguous. This results in having more than one point
in the position of a single roof vertex. If the distance between
any two or more vertices is smaller than a fixed threshold (0.50
meter), their coordinates are constrained to be equal.
2. The polygon vertices should be planer: It is assumed that
complex building roofs are built up from planar surfaces. The
aim of this constraint is to force the vertices of any polygon
to lie in the same plane. This constraint is used only if the
number of vertices in the polygon is larger than three.
3. Points that are almost in a horizontal plane are constrained
to have the same elevation: If a group of points has a small
difference in elevation (0.10 meter), they are constrained to
have the same (Z) coordinate.
4. Symmetric polygons should be constrained to have symmetric
parameters: If the computed parameters for any two planes Figure 10. The wireframes of the extracted buildings (repositioned)
indicate that the two planes are approximately symmetric,
the two planes are constrained to be symmetric.
accuracy of the three-dimensional buildings that can be extracted
Thresholds for applying the constraints are fixed for all build- using the presented system. To evaluate the accuracy of the
ings. The results of this step are shown in Figure 9. extracted buildings, the three-dimensional coordinates of the
extracted building vertices were extracted manually and compared
Building Topology Reconstruction with the automatically extracted coordinates. The average root-
After determining the topology of the roofs, the last step is to mean-square errors for the horizontal and vertical coordinates of
reconstruct the perimeter facets. This is achieved by the following all building vertices are 0.30 meter and 0.40 meter. Out of about
algorithm. First, border vertices are determined, using a point in 150 roof regions, 141 roof polygons were detected.
polygon algorithm, and sorted clockwise. A facet is generated for
any two successive border vertices. The two vertices are assumed to
be the upper points of the facet. The horizontal coordinates of the Discussions and
lower two points are taken exactly as the horizontal coordinates Conclusions
of the upper two points, while the elevations are automatically
measured from the nearest DSM post. This paper presents a new algorithm to generate high-quality
three-dimensional building information. Results show the great
Results improvement that the algorithm adds. The user only has to select
A sample of 30 buildings extracted using the presented algorithm an image patch for the building in the first image. The algorithm
is shown in Figure 10. The results show the completeness and utilizes radiometric and geometric properties of urban build-

URISA Journal • Elaksher, Bethel 11


ings. Image segmentations, neural networks, rule-based systems, West Lafayette, IN 47907-2051
photogrammetric mathematical models, and rigorous geometric bethel@ecn.purdue.edu
constraints are used. The paper shows that at the employed im-
age scale (1:4000), segmentation provides high-quality image References
regions that can be used in the building extraction process. The
DEM is only used to provide evidence about the height of the Avrahami, Y., Y. Raizman, and Y. Doytsher. 2004. Semiautomatic
regions. The DEM is generated from the same data set and the 3D mapping of buildings from medium scale (1:40,000) aerial
only requirement on its quality is that at least one DEM post photographs. The International Archives of Photogrammetry,
is inside each roof region. Regions are classified using border Remote Sensing, and Spatial Information Sciences, Istanbul,
linearity and region height. Classification results show that these Turkey, July 2004, XXXV (B3), 774-9.
attributes are adequate to discriminate roof regions. However, Brunn, A., and U. Weidner. 1997. Extracting buildings from
other attributes, such as color, could be used. Although the digital surface models, The International Archives of
rule-based system extracted either triangle or quadrilateral roof Photogrammetry, Remote Sensing, and Spatial Information
patches, the system could be used to extract any roof shape. The Sciences. Stuttgart, Germany, September 1997, XXXII
rigorous photogrammetric mathematical models succeeded in (3-4W2), 27-34.
finding conjugate polygons, simultaneously matching across Chein, L., and W. Hsu. 2000. Extraction of man-made buildings
several views, and implementing the constraint equations for to- in multispectral stereoscopic images. The International
pology reconstruction. Although several thresholds are used, their Archives of Photogrammetry, Remote Sensing, and Spatial
values were fixed for all buildings, except the intensity difference Information Sciences. Amsterdam, The Netherlands, July
threshold that ranged from 10 to 15. Future work will focus on 2000, XXXIII (B3/1), 169-76.
using color information in the segmentation. The model provided Fischer, A., T. Kolbe, F. Lang, A. Cremers, W. Förstner, L.
an average RMS of about 0.35 meter. With comparison to the Plümer, and V. Steinhage. 1998. Extracting buildings from
image scale, this represents an average accuracy of three pixels in aerial images using hierarchical aggregation in 2D and 3D.
the image coordinates for the vertices. The algorithm succeeds Computer Vision and Image Understanding 72, no. 2:
in extracting a wide range of urban building. The tested data set 185-203.
includes simple buildings with one rectangular roof, gabled roof Förstner, W. 1999. 3D-city models: Automatic and semiautomatic
buildings, multistore buildings with large relief, and a variety of acquisition methods. Photogrammetric Week, Stuttgart
complex buildings. (September 1999).
Henricsson, O., F. Bignone, W. Willuhn, F. Ade, O. Kubler, E.
About the Authors Baltsavias, S. Mason, and A. Grun. 1996. Project amobe:
Strategies, current status, and future work. The International
Ahmed F. Elaksher is an assistant professor at the engineering Archives of Photogrammetry, Remote Sensing, and Spatial
faculty at Cairo University. He earned his Ph.D. in August Information Sciences, Vienna, Austria, July 1996, XXXI
of 2002 from Purdue University. His areas of interest include (B3), 321-30.
object recognition, feature extraction, three-dimensional Hongjian, Y., and Z. Shiqiang. 2006. 3D building reconstruction
modeling from remote-sensing images, management of from aerial CCD image and sparse laser sample data. Optics
spatial information databases, and decision making using and Lasers in Engineering 44, no. 6: 555-66.
GIS, Internet, and wireless technologies. Horowitz, S. L., and T. Pavlidis. 1974. Picture segmentation
Corresponding Address: by a direct split and merge procedure. In Proceedings of
Department of Public Works 2nd International Conference on Pattern Recognition,
Faculty of Engineering Copenhagen, Denmark, August 1974, 424-33.
Cairo University Hough, P. V. C. 1962. Method and means for recognizing
Giza, Egypt complex patterns. U.S. Patent 3,069,654.
ahmedelaksher@yahoo.com Kim, A. 2001. Multi-view 3D object description with uncertain
reasoning and machine learning. Ph.D. Thesis, University
James Bethel is an associate professor at the School of Civil of Southern California, August 2001.
Engineering, Purdue University. His areas of interest include Lari, Z., and H. Ebadi. 2007. Automatic extraction of building
photogrammetry, remote sensing, data adjustment, and features from high resolution satellite images using artificial
digital-image processing. neural networks. In Proceedings of ISPRS Conference on
Information Extraction from SAR and Optical Data, with
Corresponding Address: Emphasis on Developing Countries, Istanbul, Turkey, May
Purdue University 2007.
School of Civil Engineering
550 Stadium Mall Drive

12 URISA Journal • Vol. 20, No. 1 • 2008


Lee, D. S., J. Shan, and J. S. Bethel. 2003. Class-guided building Shmid, C., and A. Zisserman. 2000. The geometry and matching
extraction from IKONOS imagery. Photogrammetric of lines and curves over multiple views. IJCV 40, no. 3:
Engineering and Remote Sensing 69, no. 2: 143-50. 199-233.
Lin, C., and R. Nevatia. 1998. Building detection and description Tarsha-Kurdi, F., T. Landes, P. Grussenmeyer, and E. Smigiel.
from a single intensity image. Computer Vision and Image 2006. New approach for automatic detection of buildings
Understanding, 72, no. 2: 101-21. in airborne laser scanner data using first echo only. The
Masaharu, H., and K. Ohtsubo. 2002. A filtering method International Archives of Photogrammetry, Remote
of airborne laser scanner data for complex terrain. The Sensing, and Spatial Information Sciences, Bonn, Germany,
International Archives of Photogrammetry, Remote Sensing, September 2006, XXXV (I/3) (on CD-ROM).
and Spatial Information Sciences, Graz, Austria, XXXIV Theng, L. B. 2006. Semi-automatic building extraction utilizing
(3B), 165-9. Quickbird imagery. Journal of Engineering Letters 13, no.
Mikhail, E., J. Bethel, and J. McGlone. 2001. Introduction 3 (November 2006), http://www/engineeringletters.com/
to modern photogrammetry. New York: John Wiley and issues_v13/is sue_3/.
Sons. Tse, R. O. C., C. M. Gold, and D. Kidner. 2006. A new approach
Morgan, M., and A. Habib. 2002. Interpolation of LIDAR and to urban modeling based on LIDAR. In Proceedings of
automatic building extraction. In Proceedings of the 2002 Winter School of Computer Sciences, the 14th International
ASPRS, Washington, D.C., April 2002 (on CD-ROM). Conference in Central Europe on Computer Graphics,
Muller, S., and D. W. Zaum. 2005. Robust building detection in Visualization and Computer Vision, Pilsen, Czech Republic,
aerial images. The International Archives of Photogrammetry, January-February 2006, 279-86.
Remote Sensing, and Spatial Information Sciences, XXXVI Wang, S. D., and Y. H. Tseng. 2004. Semi-automated CSG
(3/W24), Vienna, Austria, August 2005 (on CD-ROM). model-based building extraction from photogrammetric
Principe, J., N. Euliano, and W. Lefebvre. 1999. Neural and images. The International Archives of Photogrammetry,
adaptive systems: Fundamentals through simulation. New Remote Sensing, and Spatial Information Sciences, Istanbul,
York: John Wiley and Sons. Turkey, July 2004, XXXV (B3) (on CD-ROM).
Rottensteiner, F. 2000. Semiautomatic building recognition Wack, R., and A. Wimmer. 2002. Digital terrain models from
integrated in strict bundle block adjustment. The International airborne laser scanner data—a grid based approach. The
Archives of Photogrammetry, Remote Sensing, and Spatial International Archives of Photogrammetry, Remote Sensing,
Information Sciences, Amsterdam, The Netherlands, July and Spatial Information Sciences, Graz, Austria, XXXIV
2000, XXXIII (B2), 463-8. (3B), 293-6.
Samet, H. 1982. Neighbor finding techniques for images
represented by quadtrees. Computer Graphics and Image
Processing 18, no. 1: 37-57.

URISA Journal • Elaksher, Bethel 13


TM
Robust Principal Component Analysis and Geographically
Weighted Regression: Urbanization in the Twin Cities
Metropolitan Area of Minnesota
Debarchana Ghosh and Steven M. Manson

Abstract: In this paper, we present a hybrid approach, robust principal component geographically weighted regression (RP-
CGWR), in examining urbanization as a function of both extant urban land use and the effect of social and environmental
factors in the Twin Cities Metropolitan Area (TCMA) of Minnesota. We used remotely sensed data to treat urbanization via
the proxy of impervious surface. We then integrated two different methods, robust principal component analysis (RPCA) and
geographically weighted regression (GWR) to create an innovative approach to model urbanization. The RPCGWR results show
significant spatial heterogeneity in the relationships between proportion of impervious surface and the explanatory factors in
the TCMA. We link this heterogeneity to the “sprawling” nature of urban land use that has moved outward from the core Twin
Cities through to their suburbs and exurbs.

Keywords: Land use, urbanization, robust principal component principal component geographically weighted regression
analysis, geographically weighted regression (RPCGWR)—to examine both the location of urban land use
and the relative influence of socioeconomic, demographic, policy,
Introduction and environmental factors. We integrate two different methods,
We have long altered the land by clearing forests, farming, and robust principal component analysis (RPCA) and geographically
building settlements. This land change has serious social and weighted regression (GWR) to create a novel alternative to stan-
environmental impacts, many of which are increasingly evident dard statistical approaches. First, to reduce the dimensions and
in urban areas that now host the majority of the world’s popula- number of primary regressors, we applied principal component
tion. In the United States, urbanization is driven primarily by analysis (PCA) to the explanatory variables. To account for the
suburbanization or decentralized, low-density residential land influence of outliers in standard PCA, we conducted a robust
use, and creation of far-flung suburbs or exurbanization. While principal component analysis (RPCA) by employing a projection
suburbanization offers important benefits such as affordable pursuit approach. Second, to capture spatial heterogeneity in the
housing, it also has negative impacts on systems ranging from urban landscape, we conducted GWR on the robust principal
transportation to natural habitat to infrastructure efficiencies to components (RPCs). We compared the results of the RPCGWR
inner-city economies (Burchell et al. 1998, Daniels 1999, EPA with a standard global principal component regression (RPCGR)
2001). and used a series of visual and statistical comparisons to better
The magnitude and nature of urbanization impacts are tied understand how RPCGWR lends insight into the complex dy-
not only to the amount of land converted to urban use but also namics of urban land use.
to its spatial configuration and pattern (IGBP-IHDP 1995).
Dispersed urbanization, for example, creates infrastructure inef- Study Area and Background
ficiency by spreading out roads or sewer networks. Despite the Urbanization has profound implications for the environmental
importance of spatial patterning in determining impacts of urban- and socioeconomic sustainability of communities such as the Twin
ization, a good deal of urban research focuses on aggregate mea- Cities Metropolitan Area (TCMA) of Minnesota (see Figure 1).
sures such as commute time or population density (Galster et al. This 7,700 km2 seven-county area is the economic hub of a mul-
2001). Though this synoptic view is a critical avenue for research, tistate region. Home to 2.8 million people, it is forecasted to top
it may not capture the temporal and fine-scaled spatial patterns 3.5 million by 2020. It is also a major center of sprawl, the rapid
and processes of urbanization (Hasse and Lathrop 2003). expansion of low-density suburbs into formerly rural areas and the
A variety of approaches meet the need to examine and creation of urban, suburban, and exurb agglomerations buffered
model land use at fine spatial scales, and to these we add a new from others by undeveloped land. The metropolitan region also
one. Methodologies range from simple mathematical formulas has seen a marked increase in sprawl and associated aspects such
and gravity models to sophisticated spatiotemporal simulations as traffic congestion (CEE 1999, Schrank and Lomax 2004).
(Kaimowitz and Angelsen 1998, Lambin 1994, Parker et al. The TCMA is an ideal setting for examining land use. The
2003). In this paper, we present a hybrid approach—robust region exemplifies the spatial and temporal dynamics of urban-

URISA Journal • Ghosh, Manson 15


1964, Bockstael 1996). Such phenomena include proximity to
employment centers, infrastructure, and locations with aesthetic
or recreational value. For the TCMA, of particular importance is
access to key infrastructure including sewerage, primary highways,
and surface roads. Also critical is cost-distance to the core Twin
Cities (Minneapolis and St. Paul), distance to the nearest park,
water bodies, shopping centers, and urban agglomeration. We
included proximity to airplane noise as a nuisance factor that
discourages urban development. Euclidian distance to a feature
is denoted (D); for example, SEWERD is distance to the nearest
sewerage. We also calculated cost-distance surface to the nearest
Figure 1. Twin Cities Metropolitan study site and percentage of feature as a function of highways (D1) and surface roads (D2).
impervious surface
In this case, we calculated two different cost-distance surfaces that
vary according to whether highways or surface streets are used to
ization in the United States. It serves as the hub for a large geo-
arrive at the feature of interest. SHOPD1, for example, denotes
graphic area and stands in relative isolation from other large urban
the cost-distance (measured in time as a function of distance trav-
agglomerations, making it easier to extract land-use dynamics at
eled on the network) to the nearest shopping center via highways,
the metropolitan scale. It also is important to understand the
while SHOPD2 denotes cost-distance via surface streets.
role of the region’s distinctive policy setting in shaping land-use
On the other hand, related theories of absolute space consider
patterns because jurisdictions nationwide are wrestling with the
local neighborhood characteristics such as population density,
balance between local laissez-faire dynamics and tightly controlled
tax rates, existing land cover, lot size, and school district quality
regional development (Pendall et al. 2002).
(e.g., the Ricardian view of economic activity) (Bockstael 1996,
Irwin 2002). Absolute demographic and socioeconomic factors
Methodology that bear on urbanization in the TCMA include median income,
Urban researchers and policy makers have long used statistical nonwhite population, population density, school test scores, and
regression analysis to understand factors important to urbaniza- prevalence of subsidized school lunches as a measure of neighbor-
tion. This approach creates a mathematical model of the relation- hood characteristics (Bayoh et al. 2006, Soliani and Rossi 1992,
ship between some measure, termed the response variable, and a Vincent 2006). Key political and policy institutions include
series of explanatory variables. In the case of the TCMA, we can agricultural and natural areas protection programs, county gov-
assess the relationship between urbanization in a given location ernments, which control some development costs, and the Met-
(percentage of impervious surface) as a function of potential ropolitan Council’s Metropolitan Urban Services Area (MUSA),
environmental, socioeconomic, and demographic factors, and which enforces planned growth policies. Environmental factors
policy variables for that location (measured by the explanatory include soil quality, bedrock depth, elevation, and slope, which,
variables). We expanded on this general approach by using robust in turn, affect ease of construction, the potential for aesthetic
PCA (RPCA), which acts on the large number of explanatory views, and competition for land. Table 1 includes the variable
variables to create several key robust principal components (RPCs) name, description, and data sources for all the explanatory factors
that serve as composite variables (Li 1985). We then conducted included in the study.
geographically weighted regression (GWR) with the RPCs to
capture spatial heterogeneity in the relationship between urban
land use and the selected principal components.
Robust Principal
Component Analysis (RPCA)
We used an RPCA to reduce the number of explanatory variables
Data
examined in the GWR. There are two advantages in regressing the
We identified a broad slate of explanatory factors important to
response variable (impervious surface) against RPCs rather than
understanding urbanization in the TCMA through a theoretical
directly on the explanatory variables. First, explanatory variables
analysis of the land-use literature and consultation with experts in
often are highly correlated with one another (multicollinearity),
the community, in particular the staff of the Metropolitan Council
which may cause inaccurate estimations of regression coefficients
(see Table 1). The council is a comprehensive regional planning
or poorly behaved covariance matrices when estimating a stan-
framework that coordinates the land-use activities of the region’s
dard regression model. One solution to this problem, dropping
272 local units of government, including 188 townships.
variables, can be at odds with the need to keep theoretically valid
Theories of relative space focus on the broader spatial orga-
and distinct explanatory variables in the model, such as distance
nization of social and environmental factors that affect decision
to parks and distance to water, which are conceptually separate
making, particularly through returns to land that vary with their
but often highly correlated. Because the RPCs are uncorrelated,
distance to phenomena that, in turn, affect input costs or output
multicollinearity can be avoided by using the RPCs in place of
prices (e.g., bid-rent Alonso and Von Thünen circles) (Alonso

16 URISA Journal • Vol. 20, No. 1 • 2008


Table 1. Response and explanatory variables in the study the original explanatory variables. Second, extracting a subset of
VariablesDescription Source* RPCs for prediction reduces the dimensionality of the regressors.
Impervious surfaces in the TCMA MC In the case of the TCMA, ongoing collaboration with local officials
TCMAIMP (dependent variable) and researchers has identified a large number of potential explana-
Enrollment in the agricultural pro- MC tory variables (see Table 1). Thus, we used an RPCA to reduce
AGRIPROT tection program the dimensionality of the problem—the number of explanatory
BEDROCK Bedrock height MC
variables—and accommodate for multicollinearity.
COUNTY County (7 counties, dummy variables) NHGIS
Classical PCA is vulnerable to outlying observations. As
ELEV Elevation DNR
Highway, cost-distance to nearest MC even a single large outlier can heavily influence the parameter
HWYD1 via main freeways estimates of PCA, we used a method termed projection pursuit,
Highway, cost-distance via surface MC or the notion that while most projections (combinations of de-
HWYD2 streets rivatives of data) are of low-order complexity or largely normal
Median income by census block NHGIS Gaussian distribution, a few combinations or derivatives will
INCOME group offer far-from-Gaussian distributions or high-order complexity
Airport 65db noise contour, dis- MAC (Croux 1996, Li 1985).
MAC1995D tance to (1995) For given observations x1,…..,xn Є IRp, collected in the rows
Airport 65db noise contour, dis- MAC
of the data matrix X, a coefficient vector b Є IRp is defined for
MAC2006D tance to (2006)
Minneapolis, cost-distance to via MC one-dimensional projection of the data. We assume that the first
MPLSD1 main freeways k – 1 projection directions y1,……,yk – 1 (eigenvectors) ( k > 1)
Minneapolis, cost-distance to via MC have already been measured. For finding the kth eigenvalue, a
MPLSD2 surface streets projection matrix is defined as
MUSA Metropolitan Urban Services Area MC
Nonwhite population by census NHGIS (1)
NONWHITE block group
PARKD1 Park, cost-distance to nearest via DNR,
main freeways DOT for projection on the orthogonal complement of the space
Park, cost-distance to nearest via DNR, spanned by the first k – 1 eigenvectors (for k = 1 we can take P
PARKD2 surface streets DOT = I p). The kth eigenvector then is defined by maximizing the
k
Population density by census block NHGIS function b ® S (X P kb) under the conditions b T b.
POP group (persons/km2) To extract RPCs, we have to select a subset k < p of com-
PROTECTD Protected areas, distance to nearest MC
ponents through a three-part process. First, through sequential
SCHENG English test scores, by school district MC
Subsidized lunch programs, by MC selection, we include k RPCs that explain approximately 90
SCHLNCH school district (%) percent variation of the data set. Second, we select k RPCs with
SCHMATH Math test scores, by school district MC eigenvalues greater than one. Third, we analyze the “scree plot,”
SEWERD Sewerage, distance to nearest MC which graphs the eigenvalues (expressed as explained variance) by
Shopping center, cost-distance to MC each RPCs as a line diagram (detailed later in this paper).
SHOPD1 nearest via main freeways
Shopping center cost-distance to MC
SHOPD2 nearest via surface streets
Geographically Weighted
SLOPE Slope MC Regression of Robust
SOIL Soil types (3 types, dummy variables) DNR, Principal Components
USGS With the selected k RPCs, we conducted a geographically
St. Paul, cost-distance to via main MC
weighted regression (GWR) analysis of land use. GWR offers a
STPAULD1 freeways
St. Paul, cost-distance to via surface MC number of advantages over standard regression. A typical least-
STPAULD2 streets squares regression model of the form:
City, cost-distance to via main MC
TCD1 freeways (2)
City, cost-distance to nearest via MC
TCD2 surface streets is a “global” regression, which assumes that the relationship
WATERD Water body > 3 acres (distance to) DNR between the explanatory variables and the response variable is
*Metropolitan Airports Commission (MAC), Metropolitan Council
(MC), Minnesota Department of Natural Resources (DNR), Min- constant everywhere in the study area. In many situations, this is
nesota Department of Transport (DOT), National Historical GIS not necessarily true, especially with spatial varying variables. GWR
(NHGIS), Remote Sensing and Geospatial Analysis Laboratory (RS- extends the traditional global regression framework (equation 2)
GAL), and U.S. Geological Survey (USGS)
by allowing local rather than global parameters to be estimated.
In this case, the model in equation 2 is rewritten as:

URISA Journal • Ghosh, Manson 17


Results and Discussion
(3) As mentioned previously, we engaged in a three-step process.
We first extracted RPCs, and analyzed component loadings
where (ui, vi) denotes the coordinates of the ith point in space and clustering of initial explanatory variables in the component
and βk (ui,vi) is a realization of the continuous function βk (u,v) space. Second, we used the RPCs in a standard global regression
at point i. Equation 3 creates a continuous surface of estimated in what we term a robust principal component global regression
parameter values, and measurements of this surface are taken at (RPCGR), where we model the response variable, proportion of
certain points to denote the spatial variability of the surface. This impervious surface, against the selected RPCs. Third, we exam-
spatial variability is estimated through the geographical weight- ined differences between the results of RPCGR and a robust prin-
ing scheme, W(ui, vi), defined such that data points nearer to (ui, cipal component geographically weighted regression (RPCGWR).
vi) will be assigned higher weights in the model than data points We used the GWR 3.0 software package (Fotheringham 2002)
farther away. That is, and R statistical software (Robust PCA and Projection Pursuit,
pcaPP package) for statistical analysis and Arc GIS 9.1 (ESRI)
(3) for calibrating the RPCGWR model components and visualiza-
tion of the results.
where the bold type denotes a matrix that represents an
estimate of , and W(ui, vi) is an n by n matrix whose off-diagonal
elements are zero and whose diagonal elements denote the geo-
graphical weighting of each of the observed data for regression
point i (Fotheringham 2002). The resulting parameter estimates Table 2. Total variance explained by the robust principal components
then can be mapped to analyze local variations in the estimated
Principal Eigenvalue Variation Cumulative
parameter relationships. Various diagnostic measures further
Component (%) Variation
increase the analytical capability of GWR, such as the Akaike
(%)
Information Criterion (AIC), local standard errors, local measures
of influence, and local goodness of fit. As examined later on, the 1 7.32 74.80 74.80
parameter estimates also are tested for evidence of significant 2 1.10 13.07 87.87
spatial variation relative to the global model. Figure 2 summarizes 3 0.45 4.72 92.59
the steps involved in the methodology in a schematic diagram.
4 0.41 2.89 95.48
5 0.22 1.96 97.44
6 0.15 1.01 98.45
7 0.09 0.63 99.08

Figure 2. Schematic diagram showing methodological framework Figure 3. Scree plot for RPCA on TCMA variables

18 URISA Journal • Vol. 20, No. 1 • 2008


Robust Principal Component Analysis Results
RPCA using the projection pursuit approach extracted three un-
derlying dimensions from the 30 explanatory variables expected
to influence urban development in the TCMA (see Table 1).
Table 2 shows both the eigenvalue and the raw and cumulative
percentage of variance explained by the extracted RPCs that ac-
count for 99 percent of the total variation. The first three RPCs
account for 93 percent of the total variation. The first explains 75

Table 3. Explanatory variable loadings onto individual selected robust


principal components
Variable PC1 PC2 PC3
AGRIPROT 0.000 0.000 0.000
BEDROCK 0.000 0.000 -0.002
COUNTY 0.000 0.000 0.000
ELEV 0.000 0.000 0.000
HWYD1 0.222 0.098 -0.379
HWYD2 0.217 0.121 -0.384 Figure 4. Variables in three-dimensional component space

INCOME -0.070 0.976 0.058


MAC1995D 0.211 0.054 0.508 percent, the second 13 percent, and the third explains 5 percent
MAC2006D 0.209 0.052 0.52 of the variance (see Table 2). There is, therefore, a steep drop in
the percentage of explained variance after the first RPC.
MPLSD1 0.323 0.069 -0.255
This drop also is evident in a scree diagram, which plots
MPLSD2 0.488 0.038 -0.178 the eigenvalues (variances) of the RPCs on the y-axis against the
MUSA 0.000 0.000 0.000 RPC number on the x-axis (see Figure 3). The term scree refers
NONWHITE 0.001 0.002 0.002 to the fact that the explained-variance curve resembles the side of
a mountain with a scree, or rock debris, at the base. When read
PARKD1 0.056 -0.003 -0.107 left to right across the abscissa, this plot shows a clear separation
PARKD2 0.055 -0.007 -0.096 between RPCs with high-explained variance versus low-explained
POP -0.009 -0.011 -0.003 variance. The point of separation is termed the elbow for obvi-
ous reasons that nonetheless invite a justified charge of mixed
PROTECTD 0.000 0.000 0.000
metaphors.
SCHENG 0.000 0.000 0.000 In concordance with Table 2 and Figure 3, the first RPC
SCHLNCH 0.000 0.000 0.000 explains the large majority of variation, the second less so, and
the third a small amount. The elbow occurs at the third RPC,
SCHMATH 0.000 0.000 0.000
indicating the separation of the most important RPCs from less
SEWERD 0.127 -0.002 -0.083 important RPCs, namely the fourth onward. Thus, we retained
SHOPD1 0.266 0.032 -0.253 the first three RPCs as explanatory variables for further analy-
SHOPD2 0.312 -0.037 -0.157 sis.
The key opportunity, and challenge, of PCA is determin-
SLOPE 0.000 0.000 0.000 ing what the components actually mean in a real-world setting.
SOIL 0.000 0.000 0.000 Component loadings indicate the relative contribution of the
STPAULD1 0.34 0.029 -0.137 variables to each component (seeTable 3).
In addition to examining the degree of correspondence
STPAULD2 0.538 -0.064 0.178
between components and individual variables (Table 3), we also
TCD1 0.333 0.048 -0.21 can look for clustering in component space. Figure 4 is a three-
TCD2 0.418 -0.019 -0.049 dimensional graph that shows the position of explanatory variables
with high component loadings in the component space. The
WATERD 0.012 -0.016 -0.022
graph identifies how these variables relate to both the principal

URISA Journal • Ghosh, Manson 19


Figure 5. Cost-distance factor (RPC1)

Figure 7. Infrastructure factor (RPC3)

variables that load highly onto this component. Figure 5b shows


the influence of cost-distances to the Twin Cities and major shop-
ping centers. The cost-distance factor is least near the center or the
two central business districts (CBDs) of the TCMA and increases
gradually outward from the center to the suburbs and then to the
outer suburbs. Not surprisingly, the initial cost-distance variables
also show a similar spatial pattern (see Figure 5a).
The second component, RPC2, has median income by
Figure 6. Income factor (RPC2) block group (INCOME) with a very high component loading
of 0.976, almost double that of any other variable/component
loading combination. INCOME is positively related to RPC2,
components and other input variables. indicating that higher values of RPC2 are associated with higher
The first component, RPC1, has a number of variables with values of INCOME, leading us to term RPC2 as the “income
high loadings. The variables, in descending order, are cost-dis- factor.” Figure 6 demonstrates the strong correspondence in
tance surfaces by the second-order roads to St. Paul (STPAULD2), spatial variation between the initial variable, INCOME (Figure
Minneapolis (MPLSD2), and to the nearest of the two cities 6a), and the income-factor, PC2 (Figure 6b). Other indicators
(TCD2). These are followed by cost-distance surface by first-order that often (but not always) map onto socioeconomic status and
roads to St. Paul (STPAULD1), Minneapolis (MPLSD1), and to neighborhood characteristics, such as school lunch programs or
the nearest of the two cities by highways (TCD1). The seventh ethnicity, are almost completely subsumed by INCOME.
and eighth variables are the two cost-distances to the nearest The third component, RPC3, captures a more complicated
shopping center by highway and surface streets (SHOPD1 and situation than those associated with RPC1 and RPC2. Unlike
SHOPD2). These variables are positively related to RPC1, or, in the other components, RPC3 has both high positive and nega-
other words, observations with higher values of RPC1 also will tive loadings (see Figure 7). The two strongest positive loadings
indicate higher values of all the variables mentioned previously are past and present distance to the airport 65db noise contour
and vice versa. Because these variables measure cost-distances to (MAC1995D, MAC2006D), followed by a lower loading
key urban centers, markets, and infrastructure, we termed RPC1 on cost-distance to St. Paul via surface streets (STPAULD2).
as the cost-distance factor. The remaining variables have small negative loadings on PC3.
Figure 5 illustrates the spatial variation of the cost-distance These include cost-distances to the nearest highway (HWYD1
factor (RPC1) in the TCMA region along with the explanatory and HWYD2), cost-distances to Minneapolis (MPLSD1

20 URISA Journal • Vol. 20, No. 1 • 2008


Table 4. Global robust principal component regression (RPCGR)
analysis results
Parameter Coefficients Std. Error T-Value
Value
Intercept 16.033 0.566 28.346
PC1 -4.852 0.146 -33.119
PC2 -2.356 0.247 -9.530
PC3 1.601 0.416 3.847

Figure 8. Spatial variation in influence of cost-distance factor (RPC1)

and MPLSD2), cost-distances to the nearest shopping center


(SHOPD1 and SHOPD2), cost-distances to the nearest of the
two cities (TCD1), and cost-distance to the nearest park via arity underlying the global regression model. Theories of relative
surface streets (PARKD2). space, mentioned previously, contend that the intensity of urban
RPC3 is more difficult to interpret than the first two com- development declines with increasing distance from central cities,
ponents because of the small amount of variance explained by in this case, Minneapolis and St. Paul. While the global regres-
RPC3 and the mixed nature of explanatory variables contributing sion identified this relationship between urban development and
to it. Although it is not immediately obvious from the loadings distance from a central city, it may fail to identify local variations
on this component, we define PC3 as the “infrastructure factor,” in the power of this relationship. Thus, if the relationship be-
because as will be explored in the following section, this factor tween an explanatory variable and urban development is spatially
captures path-dependence on urban form exerted by exiting nonstationary, then the global multiple regression approach can
infrastructure. misspecify the actual relationship (Fotheringham 2002). One
approach for analyzing these local variations is RPCGWR.
Robust Principal Component Global Regression
(RPCGR) Robust Principal Component Geographically
We used the three RPCs as explanatory variables in a robust Weighted Regression
principal component global regression (RPCGR) to form a basis As noted previously, we term the combination of RPCA and
for comparison to robust principal component geographically GWR as robust principal component geographically weighted
weighted regression (RPCGWR). Table 4 shows the results of regression (RPCGWR). To assess the effectiveness of RPCGWR,
a standard multiple linear regression model (again, termed a we compared its performance to that of global regression analysis
global model to distinguish it from GWR). We estimated the with the same variables, proportion of impervious surface versus
model (and the RPCGWR that follows) with 7,000 observations the three RPCs that in turn condensed 30 explanatory variables
sampled across the TCMA. The RPCGR model is significant for shown in Table 1.
all components (R2 = 0.217, p > 0.01 for PC1-PC3). With GWR, the parameter estimation at any of the sample
The proportion of impervious surface in TCMA is nega- observations depends not only on the input data but also on the
tively related to the cost-distance factor (RPC1) and income kernel (model form) chosen and the kernel’s bandwidth (spatial
factor (RPC2) but positively related to the infrastructure factor extent of the sample). We used a Gaussian model because the
(RPC3). Regions with a higher proportion of impervious surface response variable is continuous. The bandwidth was optimized
have lower values of the cost-distance factor and the income fac- as a part of the GWR calibration using the AIC method, which
tor. In contrast, the association between RPC3 and impervious balances the complexity of the estimated model (defined by how
surface is positive; as PC3 increases, the percentage of impervious specific the bandwidth becomes) with the extent to which the
surface also increases. model fits the data (defined by explained variance). GWR uses
The global regression model explains only 22 percent of the a Monte Carlo simulation to test the following hypotheses: (1)
variance in the percentage of impervious surface as a function whether the data may be described by a GWR model rather than
of the RPCs, which indicates that the model does not account a stationary one and (2) whether individual regression coefficients
for all the factors influencing urban development in the TCMA. are stable over geographic space (Fotheringham 2002).
Beyond missing variables, however, the low explained variance A comparison of regression parameters illustrates that RP-
also can be attributed at least in part to the fact that the estimated CGWR outperforms RPCGR for this study. The AIC is reduced
parameters represent global averages of relationships between im- from 64,554.63 for the global regression model to 63,871.10 for
pervious surface and the RPCs that may exhibit spatial variation the RPCGWR model, where a lower AIC indicates a more ef-
(Fotheringham 2002). In other words, some of the unexplained ficient model. RPCGWR has an R2 of 0.49, which is reasonably
variance may be associated with the assumption of spatial station- high, especially compared to an R2 of 0.22 for RPCGR. RP-

URISA Journal • Ghosh, Manson 21


CGWR also has a much lower residual sum of squares (284,179 in which the range of values in Figure 8a constitute significant
versus 3,282,543). Finally, we can compare RPCGWR against spatial variation. Areas of particular significance are located in
RPCGWR via an ANOVA, where an F of 6.69 at the p < 0.01 Anoka and Washington counties in the northwest and Carver
level indicates that RPCGWR significantly improves on the global and Scott counties in the southwest (also see Figure 1). This said,
regression model. within these counties, there are areas where the inverse is true,
In spatial terms, the Monte Carlo test of the local estimates particularly northern Anoka County, central Carver County, and
for each of the three RPCs (cost-distance, income, and infrastruc- northwestern Washington County. These subareas show develop-
ture factors) indicates significant spatial variation for a bandwidth ment in the suburbs away from the CBD and Twin Cities and
of approximately three miles. The spatial pattern of the estimated confirm the “sprawling” nature of urbanization in the TCMA.
cost-distance factor (RPC1) is shown in Figure 8. The global relationship between the proportion of impervi-
Areas marked by higher absolute values indicate regions ous surface and the income factor (RPC2) is significantly positive.
where the explanatory variables with higher component loading Figure 9a shows that this positive relationship holds over most of
under RPC1 have a greater influence, while areas of lower values the study area because the majority of the local parameters also are
indicate where the explanatory variables are less influential. As positive. However, as we move away from the central region of the
noted previously, the global relationship between RPC1 and the TCMA to the peripheral regions, especially in the northwestern
proportion of impervious surface is negative, which suggests that corner of the Washington County (Figure 9 and Figure 1) and
urban development is more likely to occur closer (lower cost- the inner-city region of the Twin Cities of Minneapolis and St.
distance) to the major cites of St. Paul and Minneapolis, either Paul, the relationship becomes negative. This negative relation of
considered singly or in terms of whichever is closest to a given income factor and percentage of impervious surface highlights a
location (STPAULD1/2, MPLSD1/2, and TCD1/2, respec- typical inner-city scenario witnessed in almost all metropolitan
tively), the nearest shopping center (SHOPD1/2), or highway regions of the country. Growing population, cheap housing poli-
(HWYD1/2). cies, and the dense nature of transportation network are some of
RPCGWR shows that the contribution of RPC1 parameter the factors that changed the land to concrete urban impervious
in the regression equation varies over the study region, including surface but have simultaneously discouraged the reinvestment
a change in sign from negative to positive, which indicates that and redevelopment of older inner-city communities. This creates
this relationship is more complex than is suggested by the global a situation of “negative growth,” affecting the processes of land
regression results. We can assess the statistical significance of the use and environment around the metropolitan perimeter.
spatial variation by examining t-values at the observations. Values Figure 9b indicates that the negative relation between
falling beyond 1.96 (i.e., the 95 percent confidence interval) are percentage of impervious surface and the income factor in the
considered significant (Fotheringham 2002). According to this inner-city region of the Twin Cities and parts of northwestern
rubric, there are several areas in the TCMA where a significant Washington County are statistically significant.
positive relationship exits between impervious surfaces and the The spatial pattern of RPC3, the infrastructure factor, is
cost-distance factor. Figure 8b identifies the areas of the TCMA shown in Figure 10a. As noted previously in terms of the load-

Figure 9. Spatial variation in influence of income factor (RPC2)

22 URISA Journal • Vol. 20, No. 1 • 2008


ings on RPC3, there are strong positive association with both backyards than elsewhere in the TCMA and an attendant dense
the past and present distance to the airport’s 65db noise contour street grid and alley network.
(MAC1995D, MAC2006D) and weak positive associations The spatial patterning of RPC3, while driven in large part
with cost-distance to St. Paul via surface streets (STPAULD2). by areas of past development that map onto the spatial pattern
Conversely, there are weak negative associations between pro- of airport noise, also reflects negative loadings for several of the
portion of impervious surface and cost-distances to the nearest cost-distance measures. In particular, the majority of the areas
highway (HWYD1 and HWYD2), Minneapolis (MPLSD1 and with an apparently negative relationship with cost-distance fall in
MPLSD2), nearest shopping center (SHOPD1 and SHOPD2), largely rural areas that lie far from current infrastructure (highways
nearest of the two cities (TCD1), cost-distance to St. Paul via or markets) or in recent suburban developments that boast lower
freeways (STPAULD2), and cost-distance to the nearest park via amounts of imperviousness with their large lot sizes and irregular
surface streets (PARKD2). An examination of local parameters street grids. As seen in Figure 10a, these areas lie in the northern
from RPCGWR, however, shows a distinct spatial pattern in the and western parts of the TCMA.
changing nature of this relationship. PC3 is positively related to
impervious surface in the eastern part of Hennepin County, the Conclusion
entirety of Ramsey and Anoka counties, and the northern part This study has two types of conclusions, specific and applied.
of Dakota County (Figure 10a). These regions are characterized The specific conclusions are derived directly from the RPCGWR
by high values of airport noise (MAC1995D, MAC2006D) model and its interpretations in the context of the TCMA. These
and low values for cost-distances to infrastructure and markets conclusions contribute to the literature of spatial analysis and
(HWYD1/2, MPLSD1/2, SHOPD1/2, TCD1). modeling of urbanization in geography. On the other hand, the
We originally included airplane noise given anecdotal applied conclusions are the inputs to the policy community of the
evidence in the TCMA, supported by studies elsewhere (Nelson Metropolitan Council, the Minnesota Pollution Control Agency,
et al. 2004), that land development is less likely to increase in and other governmental agencies in the TCMA. The following
areas of high noise, limited indirectly through consumer prefer- paragraphs will expand on the specific conclusions first, followed
ences and directly through policy limits on land use. While the by the applied conclusions.
negative impact of airplane noise on urban development may be We introduce a hybrid approach, RPCGWR, to examine
true in many other study areas, the region around the airport on the relationships between urban development, as approximated
the border between Hennepin and Dakota Counties contained by impervious surface, and myriad explanatory social and envi-
some of the longest settled neighborhoods of the TCMA. These ronmental factors in the TCMA. This approach allows us to sift
areas and the adjacent suburbs were developed decades before through a large number of potential factors and identify several
the airport was built in 1962. These areas also contributed to key composite factors in the form of RPCs. RPCA, in addition to
further development of key infrastructure facilities, including rail, accounting for outliers in the original data set, is useful for reduc-
tram, and road networks and are home to the highest density of ing the complexity of explanatory variables, as seen in reducing
detached single-family dwellings, which, in turn, feature smaller the initial 30 explanatory variables to three robust components

Figure 10. Spatial variation in influence historical high-density factor (RPC3)

URISA Journal • Ghosh, Manson 23


that account for 93 percent of the total variation. planning future development. Satellite remote sensing provides
It is important to note, however, that RPCA does not have a cost-effective alternative for obtaining such information when
to divorce the components from the original factors, but, instead, the costs of traditional mapping approaches are increasing and
can serve as a way of aggregating and understanding relationships budgets are declining. Second, working with staff at the Metro-
among these factors. The first RPC was defined as the cost- politan Council and with the archives maintained by the council,
distance factor because it aggregates cost-distance measures for city we could identify likely scenarios of growth, tied to population
centers, the nearest highways, and shopping centers. The second and socioeconomic forecasts, paying particular attention to the
RPC collapses myriad socioeconomic and demographic factors role of policy instruments such as zoning, transportation, and
into a single measure, income. The third component tells a more infrastructure provision.
complicated story because, in addition to contributing a small
amount of explained variance, it captures the effect of historical
infrastructure development in heavily urban areas and relative Acknowledgments
paucity of infrastructure in other, more rural areas.
We then go one step further by using GWR to better under- This work is supported in part by the National Aeronautics and
stand the spatial variation in the RPCs as such and their relation- Space Administration New Investigator Program in Earth-Sun
ship with both urban development and the original explanatory System Science (NNX06AE85G), the University of Minnesota’s
factors. We demonstrate that, for the TCMA, RPCGWR out- College of Liberal Arts Graduate Research Partnership Program,
performed RPCGR as a tool for examining spatially varying and the McKnight Land-Grant Professorship Program. The
relationships. RPCGWR sheds light on how the importance of authors gratefully acknowledge the assistance of the editor and
socioeconomic and environmental factors can vary over space, anonymous reviewers. Responsibility for the opinions expressed
in essence adding context to theories of land use by highlighting herein is solely that of the authors.
areas where these factors play out differently. This approach also
identifies areas for further investigation. The spatial pattern of
the cost-distance component identified areas away from the Twin
Cities, for example, where significant positive relationships exist
About the Authors
between cost-distances and the proportion of impervious surface
Debarchana Ghosh is a Ph.D. candidate in the Department of
contrary to theoretical expectations and trends elsewhere in the
Geography, University of Minnesota. Her research interests
region. Similarly, even though the global relationship between the
include quantitative analysis, spatial statistics and modeling,
proportion of impervious surface and the income factor (RPC2) is
GIS, health, and environmental geography.
significantly negative, RPCGWR located patches outside the Twin
Cities where there is a significant positive relationship between Corresponding Address:
impervious surfaces and income. This spatial heterogeneity is University of Minnesota
linked to the differences in the nature of urban land use between Department of Geography
the core Twin Cities and their suburbs and exurbs. 414 Social Sciences
From the application point of view, this work is the first step 267 - 19th Avenue South
in developing trajectories of future land change for the Twin Cit- Minneapolis, MN 55455
ies. This aspect of the research is tied to the policy community Phone: (612) 625-6080
through ongoing consultation with experts from the Metropolitan Fax: (612) 624-1044
Council. The mission of this body is to develop, in coordina- ghos0033@umn.edu
tion with local communities, a comprehensive regional planning
framework that focuses on transportation and aviation, wastewater Dr. Steven M. Manson is an associate professor in the
treatment and water resources, regional parks, and regional devel- Department of Geography, University of Minnesota. He
opment and infrastructure. Its chief goal is to plan and coordinate combines environmental research, social science, and
efficient and sustainable growth of the metropolitan area. We are geographic information science to understand changing
exploring two key linkages to the Metropolitan Council. First, urban and rural landscapes in the United States and Mexico.
we have been working with the Metropolitan Council and the This work is part of his longer-term research on global
Minnesota Pollution Control Agency, for these agencies find the environmental change, decision making, and understanding
maps to be useful inputs to hydrology and runoff models and complex human-environment systems.

24 URISA Journal • Vol. 20, No. 1 • 2008


References IGBP-IHDP. 1995. Land-use and land-cover change: Science/
research plan. Stockholm: International Geosphere-
Alonso, W. 1964. Location and land use. Cambridge, MA: Biosphere Programme and International Human Dimensions
Harvard University Press. Programme.
Bayoh, I., E. G. Irwin, and T. Haab. 2006. Determinants of Irwin, E. 2002. The effects of open space on residential property
residential location choice: How important are local public values. Land Economics 78: 465-81.
goods in attracting homeowners to central city locations? Kaimowitz, D., and A. Angelsen. 1998. Economic models
Journal of Regional Science 46: 97-120. of tropical deforestation: A review. Jakarta: Centre for
Bockstael, N. E. 1996. Modeling economics and space: the International Forestry Research.
importance of a spatial perspective. American Journal of Lambin, E. F. 1994. Modelling deforestation processes: A review.
Agricultural Economics 78: 1168-80. Luxembourg: European Commission.
Burchell, R. W., N. A. Shad, D. Listokin, H. Phillips, A. Downs, Li, G., and Z. Chen. 1985. Projection-pursuit approach to robust
S. Seskin, J. S. Davis, T. Moore, D. Helton, and M. Gall. dispersion matrices and principal components: Primary
1998. Costs of sprawl: Revisited. Washington, D.C.: theory and Monte Carlo. Journal of the American Statitstical
National Academy Press. Association 80: 759-66.
CEE. 1999. Two roads diverge: Analyzing growth scenarios Nelson, A. C., R. J. Burby, E. Feser, C. J. Dawkins, E. E. Malizia,
for the Twin Cities. Minneapolis: Center for Energy and and R. Quercia. 2004. Urban containment and central city
Environment. revitalization. Journal of the American Planning Association
Croux, C., and A. Ruiz-Gazen. 1996. A fast algorithm for 70: 411-25.
robust principal components based on projection pursuit. Parker, D. C., S. M. Manson, M. Janssen, M. J. Hoffmann, and
In Computational statistics. Heidelberg: Physica-Verlag: P. J. Deadman. 2003. Multi-agent systems for the simulation
211-16. of land use and land cover change: a review. Annals of the
Daniels, T. 1999. When city and country collide: Managing Association of American Geographers 93: 316-40.
growth in the metropolitan fringe. Washington, D.C.: Pendall, R., J. Martin, and W. Fulton. 2002. Holding the line:
Island Press. Urban containment in the United States. Washington, D.C.:
EPA. 2001. Our built and natural environments. Washington, Brookings Institute.
D.C. Schrank, D., and T. Lomax. 2004. The 2004 urban mobility report.
Fotheringham, A. S., C. Brundson, and M. Charlton. 2002. College Station, TX: Texas Transportation Institute.
Geographically weighted regression—the analysis of spatially Soliani, L., and O. Rossi. 1992. Demographic factors and
varying relationship. West Sussex, England: John Wiley & land-use planning in the small islands of southern Europe.
Sons Ltd. Environmental Management 16: 603-11.
Galster, G., R. Hanson, M. R. Ratcliffe, H. Wolman, S. Coleman, Vincent, J. M. 2006. Public schools as public infrastructure—roles
and J. Freihage. 2001. Wrestling sprawl to the ground: for planning researchers. Journal of Planning Education and
Defining and measuring an elusive concept. Housing policy Research 25: 433-37.
debate 12: 681.
Hasse, J. E., and R. G. Lathrop. 2003. Land resource impact
indicators of urban sprawl. Applied Geography 23:
159-75.

URISA Journal • Ghosh, Manson 25


Where Are They? A Spatial Inquiry of Sex Offenders in
Brazos County
Praveen Maghelal, Miriam Olivares, Douglas Wunneburger, and Gustavo Roman

Abstract: The United States has several laws that restrict the movement of registered sex offenders. The majority of these laws
are spatial in nature. However, only a few studies have investigated the use of spatial technology (geographic information system)
to analyze the implications of these laws. This study uses GIS to map and identify the offenders who violate the laws policed by
federal agencies. Also, the perceived risk because of the sex offenders’ existence in the community is mapped to assist law-enforcement
agencies identify the potential suspects when sex crimes are reported. Spatial analysis revealed that more than 50 percent of the
offenders resided within the restriction distance of the places where children congregate. Immediate and lateral risk zones were
created around each offender to measure the risk each offender brings to the community in which he or she resides. Finally, this
study proposes the use of spatial technology to communicate sex-related crimes to increase the awareness of communities at risk
of sex-crime victimization.

Introduction their community. This notification system exists in all the states,
Several studies in the past decade have analyzed sexual abuse and makes it mandatory for the offenders to inform the respec-
on males and females under the age of 18 in the United States tive state authorities about their movements anywhere in the
(Tjaden and Thoennes 1998, Greenfield 1997). Finkelhor United States. This information then is made public to notify
(1994) informed that one in five females and one in seven males the communities of the offenders’ details. The Jacob Wetterling
are sexually abused by the age of 18. The fear, both personal act sets minimum standards by federal administration for states.
and altruistic, of becoming a victim of sexual abuse consciously Individual states, on the other hand, can impose more stringent
or semiconsciously exists in the community. This fear has been requirements on the offenders. In Texas, the Code of Criminal
rekindled even more with the recent unfortunate events of sex- Procedure, SB1054, Article 42.12, Section 13B (Texas Legislature
related crimes throughout the nation. Online, 78th Legislature) mandates the Child Safety Zone (CSZ)
In an attempt to redeem neighborhoods of these mishaps, for the state of Texas to be “within 1,000 feet of premises such
law-enforcement agencies have regulated various sex offender as school, day-care facility, playground, public or private youth
restriction statutes that can help manage the risk posed by sex center, public swimming pool, or video arcade facility, places
offenders. While numerous statutes have been in place for about where children generally gather.” Currently, the state of Texas
four decades now, the Jacob Wetterling Crimes Against Children stipulates anywhere from 200 feet to 1,000 feet for this zone,
and Sexually Violent Offenders Registration Program1 (42 U.S.C. which follows the drug-free zone restrictions used in the state.
14071 et seq.) of 1994 reshaped the way law enforcers managed This study investigates the locations of sex offenders’ residences
Registered Sex Offenders (RSOs) in the United States. This law with respect to the CSZ using a standard 1,000-foot buffer (as
required convicted sex offenders to register and notify their law mandated by the Texas legislature and currently under discussion
administrators of their movement. Information about offenders in the legislature2) around the child facilities on proximity to an
such as each offender’s name, age, gender, height, weight, race, RSO, area of risk owing to the RSO presence, and how such
and details of the offense are provided to the state authorities such information can be communicated to the general public.
as the State Department of Public Safety. The movement of RSOs within and between different states
After the death of Megan Kanka at the hands of a convicted with varying restriction laws makes it difficult for the offenders
sex offender living across the street in New Jersey, President and the supervising authorities to exactly determine the distance
Clinton signed an amendment to this law, requiring all states to between the residences of the offenders and the CSZ. However,
make the information about pedophiles and rapists available to current trends in modern technology such as using a geographic
the general public (Beck and Travis 2004, Engeler 2005). When information system (GIS) have made it feasible to closely super-
this law was signed in May of 1996, the local citizenry was and vise the mobility restrictions of the registered sex offenders. GIS
continues to be informed of the whereabouts of sex offenders in 2. The existing Texas legislature requires a distance of 1,000 feet.
1. Details of the Jacob Wetterling Crimes Against Children and However, the Board of Pardon and the judiciary system can decide
Sexually Violent Offenders Registration Program can be obtained this distance case by case. However, Martha Wong, Texas state
at the Cornell University Law School U.S. Code collection representative, has moved for an amendment that mandates that all
at http://www.law.cornell.edu/uscode/html/uscode42/usc_ offenders be subjected to a 1,000-foot distance throughout the state
sec_42_00014071----000-.html. of Texas. Details regarding the House Bill 1828 can be found on the
Texas legislature Website at http://www.capitol.state.tx.us/.

URISA Journal • Maghelal, Olivares, Wunneburger, Roman 27


provides a powerful tool to map these locations, efficiently update the number of offenders who resided within 1,000 feet from
the data, and frequently check for violators residing in the CSZ. such premises as schools, parks, and day-care centers. In several
This study uses the current restriction laws stipulated by the Texas cases, sex offenders were found to be living in close proximity to
legislature to inquire: (1) How many sex offenders reside within the premises where children congregate. Also, just about half of
the Child Safety Zone (1,000-foot buffer)? (2) how to identify the offenders (47 percent) were reported to be living within the
the known offenders in closest proximity to where a victim is 1,000-foot restriction area.
reported missing? and (3) how can this digital mapping system Although the impact of Megan’s Law, though spatial in na-
help notify and bring awareness to the local community? ture, has not been investigated spatially in relation to proximity
and crime (marauder), its effect has been reviewed for its mode
Background of notification by Thomas (2003). His study reported that the
Felson and Clark (1998), in their analysis, reported that all crimes methods of disseminating the community notification (Megan’s
are a result of available opportunities. The laws that have been Law) was through leaflets or flyers, community notification meet-
enacted to reduce sex crimes attempt to reduce these opportuni- ings, and other means such as the media, “need-to-know” basis,
ties as much as possible. Cohen and Felson (1979) stated that and marked-car licenses in various communities across the United
the increase in opportunities of crime is because of the presence States. While the exact distance traveled by an offender to commit
of three elements: (1) a suitable target, (2) absence of a guard- a sex crime remains to be investigated, Megan’s Law is based on
ian, and (3) a motivated offender. This is defined as the Routine the premise that an individual in immediate and close proximity
Activity Theory (RAT). Therefore, occurrence of sex crimes can to a sex offender is at risk of victimization and thus needs to be
be reduced or checked when a guardian is aware of the pres- notified about sex offender presence in the community. In Texas,
ence of a motivated offender who may attack a suitable victim. this act requires notification3 (Texas Criminal Procedures Code
Megan’s Law, in a sense, performs a similar task for society. The Annotated Section 62.201(a)) by mail to households within the
law-enforcement agencies inform the respective communities of zone of influence, defined in two levels depending on the location
the presence of a sex offender through the notification system to of the urban or rural setting of an RSO’s dwelling: immediate
increase the awareness of individuals living in close proximity to risk—within three city blocks (or approximately 0.33 miles) in
the offender. As a result, these acts help to avoid or reduce the urban or subdivided settings, and lateral risk—within one mile
chances of a motivated offender meeting a suitable target in space in rural settings (areas outside subdivisions).
and time. Extensive literature that analyzes proximity and sex Conversely, Terry Thomas (2001) reported that the Sex
crime are reported in Walker, Golden, and VanHouten (2001). Offender Act of 1997 in the United Kingdom had unforeseen
Similarly, to monitor the adjudicated offenders, agencies consequences. The implication of this law required setting up
such as the Parole Board and the Board of Pardon have devel- a mechanism of registration to help the police decide the risk
oped restrictions that control the mobility of the offenders in caused by a dangerous sex offender. This idea of registration of
the community. These restrictions have received wide attention sex offenders was based on three arguments: (1) to help the police
recently. identify the suspects after a crime, (2) to prevent crime, and (3)
Robertson (2000, 109) suggested that: to act as a deterrent (Home Office, 107). The registration laws
… an understanding of geographic trends of registered sex in United States serve the same purpose. However, residential
offenders, especially as they relate to schools and daycare restrictions for the sex offenders vary from state to state. In Illinois,
facilities, may help police narrow their suspect lists in open the least restrictive distance is 500 feet, while California restricts
cases, to those individuals contained within their registration the dwelling of sex offenders within a quarter mile of schools
database, living within a close proximity to the victim (Levenson and Cotter 2005). Texas law allows these restrictions
procurement site, who pose a high risk of recidivism. to vary for each offender, depending on the type of offense com-
mitted. Presently, the parole board assigns this distance based
Although RAT indicates that there is a relationship between on the individual’s offense. This distance can be anywhere up
proximity and repeated sex crimes, Levenson and Cotter (2005) to 1,000 feet. This makes it difficult for the law-enforcement
reported that there is lack of empirical research in this area. officials to keep a check on the movement of these offenders. To
Canter and Larkin (1993) proposed the commuter and marauder alleviate these difficulties, the Texas legislature has been forwarded
hypothesis based on the components of proximity and crime. a petition to make the restrictive distance of 1,000 feet uniform
They proposed that commuters travel to commit crimes in other throughout the state of Texas.
areas beyond their homes and the marauder criminals use the areas Nonetheless, the undisputed fact remains that the primary
around their homes to commit crimes. Crime and proximity in aim of the notification system is to increase the vigilantism of
relation to Child Safety Zone (commuters) has been investigated communities. One of the more recent methods of increasing
spatially by Walker et al. (2001). Their study analyzed the prox- vigilantism against crime is through Web-based technologies.
imity of sex offenders and the potential victims using GIS. They 3. Details regarding the notification system in Texas are available at
used the Arkansas Code of 1,000-foot buffer and identified the http://tlo2.tlc.state.tx.us/statutes/cr.toc.htm, titled “Article 62.201.
offenders who lived within this distance. They also investigated Additional public notice for individuals subject to civil commitment.”

28 URISA Journal • Vol. 20, No. 1 • 2008


Spatial technology, in addition to being used as a mapping tool
(e.g., Walker et al. 2001, Foote and Crum 1995), can be effectively
used as a communication tool using Web GIS. While Web GIS
has been used as a communication tool for some time now (e.g.,
Ramasubramanian 1995), its application to communicate sex
crimes has been increasingly advocated (e.g., Albrecht and Pingel
2005, Shyy, Stimson, Western, Murray, and Mazerolle 2005). It
is important to communicate information regarding sex crimes
because, as suggested by Sampson, Raudenbush, and Earls (1997),
it increases the “collective efficacy” of the community. They state
that an increase in collective efficacy of a community results in
lower expected levels of crime.
Therefore, this study provides a methodological approach to
identify and notify individuals about the sex offenders in their
community using GIS. Deriving from the RAT, sex crime, as a
result of proximity, can occur in two ways. In the absence of a Figure 1. Methodology used to develop the tool to locate sex
guardian by (1) proximity to probable or suitable victims and offenders
(2) proximity to motivated offenders. While a detailed literature
review by Walker et al. (2001) assesses the issues of proximity to
suitable victims using GIS, proximity to motivated offenders has This study required the geocoding of day-care facilities in
been neglected as a cause of sex crime. The following investiga- Brazos County obtained from the Department of Family and Pro-
tion will help (1) identify the percentage of offenders who violate tective Services, and schools and parks in Brazos County. Schools,
the residential restrictions if and when 1,000-foot restriction is parks, and day-care centers in Brazos County were geocoded
approved; (2) identify the areas of probable crime in close vicin- using parcel-level data. These are the locations where children
ity to the offenders’ homes; and (3) measure the effectiveness of generally gather and were used as basic themes to develop the
developing a Web-based GIS notification system that maps the CSZ. The address information of the RSOs was obtained from
offenders and the restriction zones. the Texas Department of Public Safety’s Sex Offender Database.
The spatial data of Brazos County parcels was used to geocode
(single-field (file)) the “USaddress” field with the address database
Methodology files of the day-care centers and the registered sex offenders in
Brazos County.
Study Area Matching interactively, the unmatched addresses of the of-
The area under study, Brazos County, Texas, has a population of fenders were searched and selected. About 12 of the 164 addresses
about 152,000 (U.S. Census 2000), comprised of the cities of of the sex offenders were either located out of the Brazos County
Bryan, College Station, and Wellborn. About 88 percent of the or could not be located in the Brazos parcels file and thus were
population of Brazos County resides in the twin cities of Bryan not used for further analysis. The layers with information on
and College Station. The Texas Department of Public Safety parks and schools in Brazos County were buffered for a distance
(TXDPS) lists 164 registered sex offenders in the zip codes of of 1,000 feet. These layers were appended and merged together
Brazos County. Also included on this list are the name of each to form the new dissolved layer of all the buffers that formed the
offender (including alias names), date of birth, gender, race, cur- Child Safety Zone. Playgrounds, public or private youth centers,
rent residential address, information pertinent to the offense, and and public swimming pools are a part of the schools in Bryan and
latest photograph with other information. Without a geographical College Station. Once the sex offender locations were geocoded,
system in place to track registered sex offenders, the cities of Col- spatial query was made to locate the offenders residing in the
lege Station and Bryan have not been able to check the violators CSZ (see Figure 1).
who reside within the Child Safety Zones for a long time. Thus,
it was necessary to provide the law-enforcement authorities with Locating Known Offenders
tools to help them locate such violators residing in the neighbor- Risk-assessment tools to predict risk have been investigated for
hoods within child safety zones. some time now (Hanson and Thornton 2000, Thornton et al.
2003). The Static 99 risk-assessment tool generates four categories
Spatial Inquiry into the Location of Sex Offenders of risk: low, low-medium, medium-high, and high, and has been
The spatial information for Brazos County was provided by the validated by Beech, Friendship, Erikson, and Hanson (2002)
city of Bryan Information Technology (IT) Department. The two and Thornton (2002). By using the Texas Case Classification
main themes created for this analysis were (1) the Child Safety and Risk-Assessment tool, community supervision officers have
Zone and (2) the location of residence of each offender. established three different levels of risks associated with each

URISA Journal • Maghelal, Olivares, Wunneburger, Roman 29


registered sex offender. The risk levels based on the nature of the
crime are high, medium, and low (Texas Department of Criminal
Justice Website). These risk levels are assigned by the Department
of Corrections, the Department of Social and Health Service, and
the Sentence Review Board. For high-risk offenders, the TXDPS
is required to send postcards to residents in the one-mile radius
of a nonsubdivided area and a three-block radius of a subdivided
neighborhood within seven days of release and ten days of move
of a sex offender to their neighborhood. This is because high-risk
offenders are considered the most probable convicts to reoffend.
Even though only the moving in of a high-risk offender requires
notification to the community, every offender induces a certain
level of perception of risk in the community where he or she lives.
Therefore, the Critical Risk Zones (CRZs) are classified based on
the risk level of an offender as high, moderate, and low and by
proximity as immediate and lateral risk.
Currently, law-enforcement agencies buffer the location
where a victim is reported missing and identify the offenders
within the buffer to check the possible reoffenders of the reported
crime (e.g., Hubbs 2003). This method, however productive, does
not allow the law-enforcement officials to identify the closest
offender with the highest level of risk. Including the dimension
of proximity and the level of risk in this search can help the of-
ficials search for the suspect beginning with the closest high-risk
offender to the farthest low-risk offender, possibly minimizing
the time to find the suspects most probable of committing the
Figure 2. Mapping of residential location of sex offenders in relation
reported crime.
to schools, parks, and day-care centers in Brazos County
This study utilized the standards required by Megan’s Law
as the baseline to geographically analyze the victim procurement
site. The Critical Risk Zone of each sex offender was based on
the distances specified in Megan’s Law. Two zones: immediate risk high-risk offenders. About 44.2 percent of the offenders were 15
zone, a three-block distance from the residence of the offender, to 30 years old when the crime was committed, and 35 percent
and lateral risk zone, a one-mile distance from the residence of the of them were 30 to 45 years old. More than 50 percent of the
registered sex offender, were created as the “area of influence” for victims were 15 to 25 years old, and more than 80 percent (131)
each offender. These were termed the “Critical Risk Zones.” of the victims were females.

Community Notification Spatial Inquiry into Location of Sex Offenders


The spatial mapping technique to locate sex offenders can be The offenders in the CSZ (77 of 164) were categorized based on
a useful community notification tool. Therefore, a Web-based their risk levels (see Figure 2). Thirty-eight were identified as low-
GIS interface was developed and launched at the city of Bryan risk offenders, 27 as moderate-risk offenders, and 12 as high-risk
police department Website. This Website can aid the individuals offenders. An investigation, similar to that conducted by Walker
of the community to access the spatial georeferenced information et al. (2001), revealed that six (50 percent) of these high-risk
regarding registered sex offenders and child safety zones. Public offenders were within a 1,000-foot proximity of at least one day
access to this service was monitored for number of hits to assess care, four were near schools, and 11 were in close proximity to
if individuals of the Brazos County used this service, resulting in parks in the cities of Bryan and College Station. Four offenders
improved collective efficacy of the community. were identified within 1,000 feet of at least one day care and one
park in Bryan. One high-risk offender, who had been charged
Result with indecency with females age 12 and 14 in 1982 and females
The descriptive analysis of the registered sex offenders residing in age 13 in 1990, resides within 1,000 feet of one day care, one
Brazos County, Texas, showed that 73 percent of the offenders park, and one school in Bryan.
were white. More than 10 percent of offenders have committed a The spatial query showed that an alarmingly high percent-
crime at least two or more times; about 5 percent of the offenders age (55.41 percent) of offenders resided within the CSZ. The
were females; and more than 10 percent of the offenders were proximity, as shown in Figure 2, of the offenders to the schools,

30 URISA Journal • Vol. 20, No. 1 • 2008


Figure 3. Critical risk zones (both immediate and lateral) for each offender based on each one’s risk

parks, and day-care centers in Bryan/College Station was not December 2004. Although the large increase in access to this
in adherence with the state restriction of 1,000 feet in Texas. Website can be attributed to the press release, the monitoring and
Although this percentage may vary with continuous moving in assessment of total number of hits on the Website in future can
or out of the offenders in Brazos County, the findings at this show if the individuals of the community accessed the Website
snapshot of time reveal that a high percentage of these offenders to constantly update themselves. Such assessment can indicate
reside within the CSZ. increased communication of sex crime–related information
through the Web-based GIS service. Except for December 2004
Locating Known Offenders (holiday season), there was an increase in access to the Bryan
The zones were classified based on the risk level and proximity, County Website in May of 2005 when the Web-based service
according to the following six divisions: (i) Low Immediate Risk was launched.
Zones, (ii) Low Lateral Risk Zones, (iii) Moderate Immediate
Risk Zones, (iv) Moderate Lateral Risk Zones, (v) High Immedi- Discussion
ate Risk Zones, and (vi) High Lateral Risk Zones (see Figure 3). Sex-crime analysis, like any other crime analysis, is associated with
These zones may or may not overlap for two or more offenders the notion of place with a geographical location. Occurrence of
based on the distance between each other. Using GIS, the loca- crime has a spatial dimension that has been explored since the
tion where the child was reported missing can be georeferenced. 1970s (Chainey and Ratcliffe 2005). Using GIS to analyze the
Upon identifying that location, a list of registered sex offenders sex-crime occurrence now is advocated more than ever (Grubesic,
who lie within the immediate risk zone and lateral risk zone can Mack, and Murray 2007). Thus, it is important to use spatial
be generated for the purpose of investigation. technology efficiently to analyze the occurrence of crime and as
This risk-level analysis provides a platform from where the a communication tool to disseminate sex crime–related informa-
authorities can identify the registered offender residing in the tion. This study used GIS to analyze the location of sex offenders
closest proximity of a reported victim, or provide some indication within the CSZ. Enforcement of a 1,000-foot buffer for CSZ
as to where to direct the investigations after a victim is reported would result in a high percentage of offenders being in violation
missing. of the law and require relocation by the authorities to avoid hav-
ing them in close proximity to children. However, irrespective
Community Notification of where the offenders reside, they bring some amount of risk to
The press release of the new Web-based GIS service was an- the community. GIS can help map that risk using existing law
nounced on May 5, 2005. Access to this Website was monitored as a framework to help locate potential offenders in proximity
and automatically recorded to measure the number and sources to the location where a sex crime was committed. Advancement
of hits (see Figure 4). The hits on the city of Bryan RSO Website in spatial technology allows dissemination of sex crime–related
shared about 50 percent of all the hits on the city of Bryan Website information to individuals of the community to increase their
immediately after the press release of the new Web-based tool. awareness about existing offenders in the neighborhood.
Also, May of 2005 reported three times the total hits (67,777) The high percentage (more than 55 percent) of violators liv-
compared to all other months except the holiday month of ing in the CSZ can be a concern for the local community. It has to
be noted that this high percentage was due to the fact that these

URISA Journal • Maghelal, Olivares, Wunneburger, Roman 31


offenders currently reside based on the restrictions that do not where children congregate, more than half of the offenders were
follow the 1,000-foot distance. Cross-verification of the ethnicity found to be in violation of the restriction distance mandated by
of offenders with their current photos on the Website revealed that the Texas legislature. Also, offenders were mapped for the risk
a high percentage of “White” offenders were reported because the they are perceived to bring to the community. This area under risk
general classification was categorized as Black or White. Hispanic- for each offender was mapped based on the notification system
looking individuals were classified as White as well. Nevertheless, and level of risk, allowing law enforcers to identify the offenders
at the discretion of the authorities, the violators can be notified with greatest risk and closest proximity to a reported sex crime.
to relocate and also suggested to locate outside the CSZ, yet still While spatial technology can be used to map a sex crime, it also
be accessible to their jobs. can help communicate this information visually to the individuals
The CRZ can help law-enforcement officers identify the of the community. This is evident from the analysis in this study
registered offenders in the closest proximity to a victim reported when the Web-based sex-offender service was monitored for the
missing or at the location of crime. This spatial service can be number of hits after its induction. The methodology proposed in
made available to the law-enforcement agencies to investigate a this article also can be an effective tool and more importantly cost-
reported crime. The categories of risk used here are not empiri- effective and time-effective in managing risk. With the availability
cally evolved, but have been used commonly by several states in of GIS, geocoding of offenders’ residences, and the availability of
the United States. themes such as parks and schools, sex crimes related to children
The Web-based GIS service was available for local com- can be efficiently managed. Therefore, the present study hopes
munity members to increase awareness among the community to encourage the use of GIS technology in investigating crime-
about registered sex offenders’ residence locations in reference to management strategies, especially related to sex crimes.
their homes, workplaces, as well as places they visit on a regular One primary limitation of this study is the use of a common
basis. The Website provides a sense of geographical reference to 1,000-foot buffer to create the CSZ. The restriction distance var-
law enforcers and to the local community. Increased access to ies between each state. In Brazos County, the restriction distance
the Website indicates increased public awareness and interest in
using Web-based mapping services as a communication tool.
This Web-based service helps individuals relate the location of
sex offenders with their residences and the paths their children
take to commute to school.

Conclusion
This study investigated the use of spatial technology to map and
communicate sex crime–related information. GIS was used to
map the locations of places where children congregate, such as
parks, schools, and day-care centers, and the residences of regis-
tered sex offenders in Brazos County, Texas. Based on the locations

Summary by Month
Daily Average Monthly Totals
Month
Hits Files Pages Visits Sites KBytes Visits Pages Files Hits

May 2005 67777 47263 15354 459 4939 14997006 8739 291740 898001 1287773
Apr 2005 24242 15410 5354 213 3219 11878090 6410 160623 462325 727261
Mar 2005 23776 14236 6270 226 3495 9754452 7033 194376 441340 737066
Feb 2005 22161 13352 5563 262 3456 10149913 7362 155774 373879 620522
Jan 2005 21700 13195 5621 356 4625 9163086 11064 174271 409075 672723
Dec 2004 58305 23736 27928 957 16659 18585619 29684 865796 735838 1807461
Nov 2004 15938 10612 3633 146 1814 6621826 4399 109006 318375 478166
Oct 2004 16401 11292 3709 144 2001 6963752 4478 114988 350074 508444
Sep 2004 15797 11035 3564 127 2062 7617430 3816 106937 331051 473912
Aug 2004 15989 10820 3642 127 1926 42513587 3942 112922 335437 495687
Jul 2004 15598 10979 3533 135 1866 7155400 4209 109529 340363 483545
Jun 2004 14743 10325 3455 120 1845 6884369 3623 103675 309772 442290

Totals 152284530 94759 2499637 5305530 8734850

Figure 4. Chart and table of total number of hits to access the sex offender Web service

32 URISA Journal • Vol. 20, No. 1 • 2008


varies with the individual offender. The significance of the distance and fear of victimization. Journal of Criminal Justice 32:
is more perceived than empirically tested. Therefore, a future 455-63.
scope of this study would include empirical examination of the Beech, A., C. Friendship, M. Erikson, and R. Hanson. 2002.
influence of restriction distance on the risk to the community us- The relationship between static and dynamic risk factors and
ing GIS. Nonetheless, this study demonstrated the use of GIS to reconviction in a sample of UK child abusers. Sexual Abuse:
efficiently map and communicate the information related to sex A Journal of Research and Treatment 14: 155-67.
crimes to the residents of the community and assist law enforcers Canter, D. V., and P. Larkin. 1993. The environmental range
to regulate sex crimes in their jurisdictions. of serial rapists. Journal of Environmental Psychology 13:
63-69.
About The Authors Chainey, S., and J. Ratcliffe. 2005. GIS and crime mapping.
Chichester: John Wiley and Sons.
Dr. Praveen Maghelal is an assistant professor in the Cohen, L. E., and M. Felson. 1979. Social change and crime rate
Department of Urban and Regional Planning at Florida trends: A routine activity approach. American Sociological
Atlantic University and has educational background in Review 44: 588-608.
civil engineering, architecture, and planning. His research Engeler, A. 2005. Is your child a target? The sex offender next
interest includes spatial planning, physical activity and built door. Good Housekeeping 240, no. 5: 192-97.
environment, and transportation planning. Finkelhor, D. 1994. Current information on the scope and nature
of child sexual abuse. Child Abuse and Neglect 4: 31-53.
Corresponding Address: Felson, M., and R. V. Clarke. 1998. Opportunity makes the
111 East Las Olas Boulevard thief: Practical theory for crime prevention. Police Research
Department of Urban and Regional Planning Series, Paper 98. Home Office, Policing and Reducing Crime
Florida Atlantic University Unit. Great Britain: Research, Development and Statistics
Fort Lauderdale, Florida 33301 Directorate.
Phone: (954) 726-5030 Foote, K., and S. Crum. 1995. Cartographic communication
Fax: (954) 762-5673 (November 17, 2007), http://www.colorado.edu/geography/
pmaghela@fau.edu gcraft/notes/cartocom/cartocom_f.html.
Miriam Olivares is a Ph.D. candidate in Urban and Regional Greenfeld, L. 1997. Sex offenses and offenders: An analysis of
Science at Texas A&M University, from where she earned a data on rape and sexual assault. Washington D.C.: U.S.
master’s degree in land development. She holds a bachelor’s Department of Justice, Bureau of Justice Statistics.
degree in architecture with emphasis on planning from Grubesic, T. H., E. Mack, and A. T. Murray. 2007. Geographic
Monterrey Tech, Mexico. Her research interest is in exclusion: Spatial analysis for evaluating the implications
sustainable development. Currently, she is working on her of Megan’s Law. Social Science Computer Review 25, no.
dissertation regarding sustainable communities and sex-crime 2: 143-62.
management. Hanson, R., and D. Thornton. 2000. Improving risk assessments
Dr. Douglas Wunneburger is a senior lecturer in the Department for sex offenders: A comparison of three actuarial scales. Law
of Landscape Architecture and Urban Planning at Texas and Human Behavior 24: 119-36.
A&M University. His primary research interests include the Hubbs, R. 2003. Mapping crime and community problems in
integration of spatial and information technology for studies Knoxville, Tennessee. In M. R. Leipnik and D. R. Albert,
in landscape ecology-based planning and management. eds., GIS in law enforcement: Implementation issues and
Gustavo Roman is the Director of Information Technology for case studies. London: Taylor & Francis.
the City of Bryan, Texas.  He holds master’s and bachelor’s Levenson, J. S., and L. P. Cotter. 2005. The impact of sex offender
degrees from Texas A&M University and has more than residence restrictions: 1000 feet from danger or one step
12 years of municipal government experience, including from absurd? International Journal of Offender Therapy and
seven-plus years in the implementation and management Comparative Criminology 49, no. 2: 168-78.
of GIS systems.  Ramasubramanian, L. 1995. Building communities: GIS and
participatory decision making. Journal of Urban Technology
3, no. 3: 67-79.
References Robertson, L. S. 2000. Assessment of the proximity of registered
sex offenders to schools and day cares: St. Louis City,
Albrecht, J., and J. Pingel. 2005. GIS as a communication process: Missouri, 1998. MO: Department of Administration of
Experiences from the Milwaukee COMPASS project. In Justice.
F. Wang, ed., Geographic information systems and crime Sampson, R., S. Raudenbush, and F. Earls. 1997. Neighborhoods
analysis. Hershey, PA: Idea Group Publishing, 1-23. and violent crime: A multilevel study of collective efficacy.
Beck, V. S., and L. F. Travis III. 2004. Sex offender notification Science 277: 918-24.

URISA Journal • Maghelal, Olivares, Wunneburger, Roman 33


Shyy, T-K., R. J. Stimson, J. Western, A. T. Murray, and L. Thomas, T. 2003. Sex offender community notification:
Mazerolle. 2005. WebGIS for mapping community crime experiences from America. The Howard Journal 42, no. 3:
rates: Approaches and challenges. In F. Wang, ed., Geographic 217-28.
information systems and crime analysis. Hershey, PA: Idea Thomas, T. 2001. Sex offenders, the home office, and the Sunday
Group Publishing, 14, 236-53. papers. Journal of Social Welfare and Family Law 23, no.
Texas Department of Family and Protective Services, Child Care 1: 103-8.
Licensing Division (November 20, 2004), http://www. Thornton, D., R. Mann, S. Webster, L. Blud, R. Travers, C.,
tdprs.state.tx.us/Child_Care/About_Child_Care_Licensing/ Friendship, and M. Erickson. 2003. Distinguishing and
default.asp. combining risks for sexual and violent recidivism. Annals of
Texas Department of Criminal Justice Website (January 10, the New York Academy of Sciences 989: 225-35.
2005), http://www.tdcj.state.tx.us/cjad/cjad-who-sa-tool. Thornton, D. 2002. Constructing and testing a framework for
htm. dynamic risk assessment. Sexual Abuse: A Journal of Research
Texas Department of Public Safety (November 20, 2004), http:// and Treatment 14: 139-53.
www.txdps.state.tx.us/. U.S. Census Bureau, http://www.census.gov.
Texas Legislature Online, 78th Legislature (November 20, Walker, J. T., J. W. Golden, and A. C. VanHouten. 2001. The
2004), http://www.tdh.state.tx.us/hcqs/plc/csotlaws. geographical link between sex offenders and potential
htm#legislation. victims: A routine activities approach. Justice Research and
Tjaden, P., and N. Thoennes. 1998. Prevalence, incidence, and Policy 3, no. 22: 15-33.
consequences of violence against women: Findings from the
national violence against women survey. Washington, D.C.:
U.S. Department of Justice, National Institute of Justice.

34 URISA Journal • Vol. 20, No. 1 • 2008


TOOLS AND METHODS FOR A TRANSPORTATION
HOUSEHOLD SURVEY
Martin Trépanier, Robert Chapleau, and Catherine Morency

Abstract: Nowadays, large transportation household surveys cannot be conducted without the help of powerful management
and support tools, and the information technologies are useful for preparing, conducting, and postanalyzing such surveys. In the
Greater Montreal Area (GMA), the 2003 household survey followed the general methodology that has been developed over the
past 20 years to integrate the finest software, databases, and methods. The tools making up the household survey information
system (HSIS) are based on the totally disaggregate approach and its object-oriented extension. This paper presents the background
and the fundamentals of the Montreal 2003 survey information system, and describes the way in which it has been assembled,
illustrating the functional and technical architectures that were used. It also emphasizes the transposition of the method to other
transportation survey activities and planning tools. The final discussion stresses the “winning” elements involved in conducting
a modern transportation household survey successfully.

Introduction Methods for Transport planning. With the advent of new tech-
Large household surveys always have presented a methodological nologies, combining spatial information systems and computa-
challenge for transportation planners and authorities. Conduct- tion capacities, travel surveys have become an integral part of the
ing a survey of more than 70,000 households is not a simple task continuing transportation planning process and assist many types
because of the sample size and the complexity of the survey itself. of transportation studies.
Every planner knows that transportation data are strongly related In the Transportation Research Board Millennium Paper of
to the spatial elements of a territory and to the transportation the Committee on Travel Surveys Methods, Griffiths et al. (2000)
network (roads and public transit), and that the survey tool must identify future directions for travel survey methods:
take these specificities into account. Today, even though intelligent • The improvement of the quality standards of travel surveys
transportation systems (ITSs) have provided new ways to collect through full and honest documentation of the survey process.
data, large transportation surveys still are needed. Data collected The need to document all stages of the survey process also
from these operations now are well integrated in the fields of appears as the most overriding conclusion of a conference
transportation planning, finance, and management. held in 1997 on raising the standards of travel surveys
This paper presents the information technologies that were (Richardson 2002).
used for the 2003 Greater Montreal Area Household Survey • The use of mixed-mode survey designs to meet the data
(Quebec, Canada). It also emphasizes the technological back- needs of the surveyor in ways that create the least burden
ground and architectures that were required to yield the best and the greatest flexibility for the respondents. The concept
results possible from the survey. Following a recounting of the of common cognitive space between an interviewer and a
history of the household survey in the Montreal area, the totally respondent was outlined by Brög (2000). The purpose of
disaggregate approach and transportation object-oriented model- survey tools is to maximize this common space to facilitate
ing, two key elements that helped support and develop the 2003 the exchange of information between the two agents and to
tools, are presented in the background section. The third part of lessen the respondent burden.
the paper, “Survey Information System Framework,” describes • A move toward a more continuous survey to provide more
the methodology that was used to prepare and synchronize the timely data in an economical manner, which also would
various software programs and databases. The “Implementation” develop and preserve technical and managerial skills in the
section is aimed at demonstrating the functions of the software conduct of complex surveys.
that was used for the survey. The conclusion reports some findings • The judicious use of new technologies to augment existing
on the 2003 experience in Montreal. survey techniques.

In this regard, computer-assisted telephone interviewing


Background (CATI) is one of the main fields of development regarding travel
In the past, travel surveys were conducted mainly by mail or
surveys. It allows interviewers to administer a survey question-
face-to-face interviews. They basically provided data for the de-
naire via telephone and capture responses electronically. CATI
velopment of aggregated travel forecasting models. Richardson
“employs interactive computing systems to assist interviewers
et al. (1995) propose a thorough description of classical Survey
and their supervisors in performing the basic data-collection

URISA Journal • Trépanier, Chapleau, Morency 35


tasks of telephone interview surveys” (Nicholls II 1988). It can reporting yields a lower trip rate than that of respondents
be viewed as a tool to facilitate or expedite telephone surveys, who are interviewed personally.” Badoe and Steuart (2002)
to enhance and control survey data quality, and to allow new also discuss the potential bias caused by interviewing by proxy
types of surveys. Jones and Polak (1992) point to the ability to respondents. Incidentally, the effects of proxy respondents on
combine the data collection and management functions as one trip rates are evaluated cyclically for the Montreal surveys.
of the key advantages shown by CATI. For recent discussions, • The sample is approximately 5 percent of the residing
the reader can refer to a report on Survey Automation Tools by population. This sample is drawn from a set of residential
the National Research Council (2003) or to Couper et al. (1998) phone numbers. Sampling strata are defined to monitor the
who discuss Computer Assisted Survey Information Collection construction of the final sample through the overall interview
(CASIC) methods. process (four months of survey).
As will be discussed in the following sections, the House- • The standard questionnaire is organized into three sections:
hold Survey Information System of the 2003 Montreal Survey households, people, and trips.
integrates tools and functions to address these issues. Automatic
documentation of the survey process, synchronous/asynchronous More details regarding the origin-destination surveys held
monitoring of interviews, transposition of the tools to other sur- in the Montreal area can be found on the Web site of the Met-
vey methods (postcoding of onboard surveys or self-completion ropolitan Information Centre on Urban Transportation (http://
of survey questionnaires through private access via Internet), www.cimtu.qc.ca/index.asp). The next survey in the Montreal
and integration with implemented planning and operational area will be conducted during the fall of 2008, with a proposed
tools are some of the features of the presented household survey budget of 1.8 millions CAD$ (Bergeron 2008).
information system. Through the years, this process has evolved both technically
and methodologically.
Household Surveys in the Greater Montreal Area
The history of origin-destination household surveys in Montreal Technical Evolution
begins in the 1960s, when the first large-scale survey in the re- Technically, the surveys have benefited from the evolution of
gion was conducted. Since 1970, eight large surveys have been computer technology. The first survey data were posttreated with
conducted, at five- to six-year intervals (see Table 1). The standard computer programs running during weekends on large comput-
survey method relies on the following principles: ers rented to sizable organizations such as the Montreal School
• The interviews are conducted by telephone, by agents Board. In 1982, data was validated using computer procedures
specially recruited and trained for this purpose. The that now form the basis for the well-known MADITUC (Modèle
telephone remains an efficient way to survey people, even D’analyse Désagrégée des Itinéraires de Transport Urbain Collec-
though there are problems of reach ability and nonresponse tif ) system and the totally disaggregate approach (both of which
with privacy protection systems and homes not equipped are defined in the following sections). In 1987, survey data was
with fixed phones (Westrick and Mount 2007, Link and coded, geocoded, and validated with the help of microcomputers.
Kresnow 2006). In 1993, a survey firm was contracted to conduct the survey, and
• All the trips made on the previous day are collected for data was captured by means of its own in-house software, based
every person residing at the contacted household; details on the VAX system. Because of data postvalidation and survey
regarding trip ends, times of departure, mode sequence, quality concerns, it was decided that computer-assisted interview-
and trip purposes are gathered. Surveys are trip-based and ing software (CATI) would be developed for future surveys. In
relate to a single week-day. Even if emerging issues regarding 1998 and 2003, a software suite was used, which combined the
the substitution of out-of-home activities by in-home best practices and procedures from past surveys. It also took ad-
activities are discussed in the literature, the metropolitan vantage of the evolution of computer technology (both hardware
steering committee on travel surveys sees no need to move and software) that had occurred during this period, especially the
toward an activity-based survey for the main purpose of the multi-tasking capabilities of Microsoft Windows.
origin-destination surveys is to precisely measure the use of
transportation networks. Thus, it appears more important to Methodological Evolution
preserve comparability between successive surveys to measure Most of the advances of the 1970–2000 period were method-
the evolution of trip patterns over the area. Moreover, totally ological. Many aspects of survey methodology have evolved
disaggregate functions allow the construction of activity since 1970:
patterns from individual travel behaviors. • Spatial zoning. Prior to 1987, the territory was divided into
• Generally, a unique respondent provides all the information several zones (the transit equivalent of traffic analysis zones, or
regarding the trips made by all the members of the household. TAZ), reflecting the general usage in transportation planning
Comparability issues also are dictating the continuity of this of synthesized and aggregate models for which little precision
methodological choice. As noted by Liss (2005) regarding is needed. This was also because of a lack of spatial search
the National Household Travel Survey (NHTS), “the Proxy engine capabilities in the survey tools. With the advent of the

36 URISA Journal • Vol. 20, No. 1 • 2008


Table 1. Comparative statistics for the past seven household surveys in Montreal
Year 1974 1978 1982 1987 1993 1998 2003
Total area 2,331 km2 2,331 km2 3,341 km2 3,350 km2 4,500 km2 5,300 km2 6,445 km2
Population 2,824 000 2,954,000 2,895,000 2,900,000 3,263,000 3,493,000 3,505,810

Sampling rate 4.8% 5.3% 7.0% 5.0% 4.7% 4.5% 4.8%

Surveyed 43,000 50,000 75,000 54,000 61,000 65,000 70,000


households
Surveyed trips 265,000 305,000 492,000 338,000 350,000 380,000 388,000

Zoning sys- 1,192 zones 1,264 zones 1,496 zones 70,000 PC 30,000 TG44,600 TG 77,800 TG
tem/ Geocodes 70,000 PC100,000 AR 160,000 AR
9,000 SN 89,000 PC 119,000 PC
40,000 IN34,000 SN 40,200 SN
191,000 IN 201,000 IN
AR: Address ranges, IN: Intersections, PC: Postal codes, SN: Street names, TG: Trip generators

totally disaggregate approach, zones have been abandoned at


the coding level for a much higher spatial level of resolution.
In 1987, the Canadian postal code (corresponding to block
faces in urban areas), then considered the best means of
location definition, was used. Now, every trip end is coded
at the X-Y coordinate level (in meters), which is the best
means available currently. In 1998 and 2003, every location
was stored and treated as well. For example, a trip generator
now can be identified under many names.
• Transit network definition. In the Greater Montreal Area,
household surveys tend to be oriented towards transit
planning usage and have been remarkably successful in
this field. Moreover, transit network data for analysis have
become more and more precise over the years. Early on, the
network was specially coded for the survey. Now, a more
Figure 1. Three-dimensional transit load profile of A.M. peak period,
“real” representation of the transit network is derived from Montreal 1998 household survey
operational data files obtained from transit authorities.
• Sampling and weighting (expansion). Significant changes
have been made to sampling methods over the years. In the
beginning, expansion was based only on people. Today, it holds, people, and trips are interleaved with looping; every trip
is categorized by both people and households, and different end needs to be geocoded; online transit trip declarations must
weights are given to people and households, depending on be validated; and the validity of trip chains within the household
their attributes (age and size, respectively). must be checked.
• Survey execution. Initially, a survey would be conducted by
employees of the transit authorities because of the absence of Totally Disaggregate Approach
CATI software and the complexity of the task. Since 1993, The totally disaggregate approach (TDA) was developed in the
a specialized survey firm has been mandated to do this. The 1980s in the Greater Montreal Area for the validation, process-
firm provides expertise in conducting surveys, and in staff ing, and modeling of large computer-assisted household origin-
and telephone infrastructure, but uses the CATI software destination surveys conducted by telephone interview. It was used
selected by the survey board committee. in particular to process transit usage declarations, but then was
extended to include other survey information. Typically, in 1998,
The question as to whether or not to use specialized CATI a telephone survey would involve more than 65,000 households
survey software has long been decided in the GMA. CATI pro- (5 percent overall sampling). To use such a quantity of data, even
vides the flexibility and the power that are needed in conducting a 1,500-zone system and its aggregate approach could not satisfy
such a complex survey. It is complex because questions on house- planners (Chapleau 1986), so a new method had to be developed

URISA Journal • Trépanier, Chapleau, Morency 37


Figure 2. Object model for 2003 household survey in Montreal

to store and process data on households, people, and trips. Set- Transportation Object-Oriented Modeling
ting aside its many features and special functionalities, the TDA Transportation object-oriented modeling (TOOM) is based on
is briefly defined here by its two essential elements: the use of transportation objects, which are special components
• Individual trip data processing throughout the transportation intended for the modeling, observation, planning, and analysis
analysis process, maintaining all trip characteristics (time, of a transportation system. For this purpose, these objects have a
purpose, modes, itinerary) with their associated person and variable state in time and space, and are characterized by special
household; properties and methods. A road link object, for example, has
• Use of X-Y coordinates, monuments, and place declarations common road properties (length, name, number), but also can
as the basic spatially referenced system for origin, destination, have time-varying properties (such as pavement condition).
residential, and intermodal junction locations for each trip Four metaclasses of transportation objects are involved in
and other spatialized objects in the system. dynamic and spatialized relations:
• Immobile (static) objects have fixed locations in time and
In terms of data completeness, the TDA does not use an space. Their roles are to describe the territory and serve as
origin-destination matrix, which would aggregate and dissolve transportation movement beacons. Some examples are the
information, but rather maintains origin-destination survey trip trip generator, postal code, census tract, and zone objects.
files containing information on trips, people, and households • Dynamic objects are the transportation actors. These objects
intact. The use of the most fully defined information improves “decide” and contribute to their movements. They represent a
the level of resolution of the system, while at the same time group of persons (household, person), a moving object (bus,
preserving any possible aggregation. As reported by O’Donnell car), or even moved objects (goods).
and Smith David (2000), possibilities are widened because the • Kinetic objects are the movement describers. Some examples
number of dimensions distinguished by the information system are the trip, transit link (simple kinetics), or the path and
is increased. The use of special analysis modules, combined with transit route (compound kinetics) objects.
the presence of an underlying GIS, provides useful tools to the • System objects are groups of embedded objects, with their
planner such as three-dimensional load profiles (see Figure 1). set of relations. They can be operational (transit network,
These load profiles help to calibrate the modules and validate the road network), informational (survey, census), or globally
results of the survey using ground counts and other observed data comprehensive (urban system).
over the transit network.

38 URISA Journal • Vol. 20, No. 1 • 2008


A transportation method is an “intelligent” sequence of proce- • Normalization of the postal code list. When possible, postal
dures used to manipulate and transform one or more transportation codes are linked to the road network to ensure a better
objects. It blends models with information, creating “infomodels” identification (always on the same multiple-alias basis).
to be reapplied to similar objects. It is important to mention that
transportation object-oriented modeling is not primarily aimed at The trip generator database is critical because most respon-
software design or database structure, and is not a database issue. dents give trip generators as trip ends and mistakes can easily be
First and foremost, TOOM is a “way of thinking” about the role made when choosing such locations in the database. Research
and specific use of every piece of information in the system. With projects by Trépanier et al. (2003) have identified important
adequate object diagrams, objects can be rapidly identified, along issues about trip generators, such as the fact that a “good” trip
with their properties and methods that are engaged in the analysis generator database must not be too narrowly defined because
(Chapleau et al. 1998). The software implementation can easily of possible mismatches between two places (for example, two
integrate these underlying concepts, but not all software languages franchise locations), but must contain all “major” generators.
are adapted to this methodology. TOOM was recently applied to To discover what these major generators are, data from previous
smart card data analysis (Morency et al. 2007). surveys were analyzed. In constructing the trip generator database,
In the object-model of the Montreal survey, there is an ob- it is important not to simply amass lists of companies provided
vious link between household, person, and trip objects, which by commercial data vendors, because these data have not been
constitute the core model of the interview (see Figure 2). But validated (they may contain double entries, spelling errors, deleted
derived objects, such as car, parking spaces, activity, and status, entries, or incomplete entries), usually are not well geocoded, and
also can be defined and analyzed with the help of the other objects, are not classified. For the 2003 survey, every trip generator was
even though they were not clearly declared in the survey. To bet- well characterized (class, exact location, named with aliases) and
ter understand the links between household surveys, TDA, and uniquely identified.
TOOM, please refer to Trépanier and Chapleau (2001). Each location in all the tables (civic addresses, intersections,
postal codes, and trip generators) is characterized and positioned at
Survey Information System the finest level of resolution, although variable definitions can be
used in surrounding regions. This ensures good geocoding during
Framework the survey. However, the CATI software also accepts locations that
Because the Montreal household survey is a short-term endeavor
are not so well defined, as is often the case in household surveys
(September to January), it must be well prepared at the begin-
because respondents do not know, or do not want to give, precise
ning; most of this preparation involves the assembly (“montage”)
information. For example, a street name alone can be given if the
of information systems, which requires mounting data structures
street is not too long, or a municipality name alone is acceptable
and collecting, normalizing, and storing data using a convenient
for places outside a territory, and so on.
software technology.
For the household survey, the GIS-T also integrates the best
possible definition of the transit network. An analytical transit
Assembly of the Geographical Information network (ATN), built up from the information provided by each
System for Transportation operator, also is required. When respondents describe transit trips,
Undoubtedly, there exists a need for a geographical information they give the sequence of routes taken. With the help of the ATN,
system for transportation (GIS-T) to support CATI during the the CATI software immediately validates the information, while
interview. The GIS-T is mainly used to geocode trip end and at the same time rejecting bad sequences and asking for precision.
junction locations, but also is called on to validate walking dis- Mistakes often are made when operators’ routes have the same
tances for transit access or to geocode places of work and home number. Also, some trips may include too much walking distance
locations. GIS-T is used because of its awareness of transportation to the stops, in which cases the CATI software flags the problems
specificities, which are different from those of classical GIS usage and asks for second validations.
(Trépanier et al. 2002).
A comprehensive road network database is developed first Assembly of the Household Survey Information
to ensure:
System
• Adequate identification of all streets within the region. This
The household survey information system (HSIS) gathers all
includes the various aliases (alternative street names) used
the information necessary to conduct the survey. The main data
by inhabitants and also considers the language differences
tables are:
between French and English.
• Households. This table contains all the information gathered
• Integrity of the list of civic numbers, which is based on street
about the households surveyed (respondent name, size, car
arc geometry and refers to the street-name database.
ownership).
• Automatic building of the intersection list from the geometry
• Persons. This table stores data on people, such as age, gender,
of the street network. (There is also a need to generate all
status (worker, retired, student, etc.), and possession of a
possible identification combinations.)
driver’s license.

URISA Journal • Trépanier, Chapleau, Morency 39


• Trips. This table describes each trip collected during the has to be transferred.
survey (purpose, time of departure, origin and destination The software was developed with Microsoft Visual Foxpro,
locations). It also contains the sequence of modes (car, using tools provided by the Windows technologies: Microsoft’s In-
transit routes, bike, foot, paratransit, school bus, taxi, etc.), ternet Information Server (IIS), the Microsoft Extended Markup
bridges crossed (if any), parking information, and freeways Language (MSXML) component model, and the Microsoft Office
traveled (if any). (MSO) component model. Applications also involved the use
• Locations. This table groups together all the described of Adobe Portable Document Files (PDF) and Microsoft Excel
locations in a single structure that identifies them by civic spreadsheets.
number, street, intersection, generator’s name, coordinates,
region, and categories. Implementation
• Calls. Each call made by an interviewer is stored in this table This section presents the features of the tools that were used in the
and is characterized by a status (completed, refused, busy Montreal 2003 survey. The intention here is not to focus on the
signal, voice mail, language problem, etc.). software itself (which is an in-house product and not commercially
• Sample. This table contains household home location data available) but rather on the various procedures employed, the
(civic number, street, municipality, postal code, telephone conduct of the survey itself, and some associated statistics.
number). To ensure spatial portioning of the sample, every Three software components were used in the 1998 and
household is geocoded at the finest level with the help of 2003 surveys. The technology evolved, but their principal roles
the civic address. Experience has shown that the postal code remained the same:
is not precise enough and is prone to error and a precise • CATI software used by interviewers (MADQUOI, Module
location is needed in the case of boundary streets between Questionneur Utilisé pour l’Obtention d’Information),
two sampling districts. • Real-time survey management software (MADASARE,
• Stratum. This table describes the group of sampled Module d’Application de Suivi et d’Analyse Rigoureux de
households within a given area of the territory, ensuring a l’Échantillon), and
uniform sampling rate during the survey. • Survey surveillance and statistics software (MADVIJIE,
• Batches. This table describes the group of sampled households Module de Validation Incontournable Journalière des
that must be chosen for a single survey day, ensuring Informations d’Enquête).
uniformity of interviews over time.
• Interviewers. Each interviewer is a user in the system, and his
or her user name serves to analyze his or her performance.
Computer-Assisted Telephone Interview
CATI is the core component of the survey software. CATI is
• Queries. Prestored queries are used by survey monitoring
one of the oldest computer-assisted interviewing methods used
staff and by transportation planners before, during, and
in travel surveys. Its primary role is to employ interactive com-
after the survey.
puting systems to assist interviewers in performing the basic
• GIS-T tables. These all are part of the HSIS (streets, addresses,
data-collection tasks of telephone interview surveys (Nicholls II
intersections, postal codes, generators, routes, road geometry,
1988, Wermuth et al. 2003). It guides the interviewer through the
transit geometry).
interview process—gathering, validating, and storing information.
Following are its main functions:
Technical Architecture • To display questions to the interviewer in suitable order,
The survey software suite works on Microsoft Windows and is according to the survey protocol. CATI also constructs the
installed on standard personal computers. In 2003, the survey necessary loops of questions when, for example, many people
floor was composed of 50 interviewer stations, five supervisor are interviewed for many trips. The answer to each question
stations (supervisors also could use interviewer stations when is stored with the entry, according to the authorized domain
needed), and a server station. All were equipped with Microsoft of answers (especially for locations). Then it quickly analyzes
Windows 2000 (workstation and server). The database architec- the answers and prompts the next question.
ture reflects the needs of survey operation. Large GIS-T tables • To geocode every location related to the interview (home
are placed directly on workstations to make CATI more powerful location, place of work, origins, destinations, junction points)
by accessing its own tables directly. In fact, these tables are not with a special interface supported by the GIS‑T database.
updated often during the survey, and so a centralized database CATI needs, and has, an “intelligent” way of dealing with
management system is not needed. locations. A location list is made up for each household and
Survey management tables containing the sample data and so a previously searched location can be easily reused. The
state of each household are stored on the server to ensure integrity. spatial logic of locations for a trip chain must be ensured: The
To facilitate data exchange, requests between workstations and origin of a trip is the destination of the preceding one.
the server are made with XML files, as these provide flexibility • To proceed to immediate answer validation in the case of
and, more importantly, variable data structures and length in cases spatially driven questions such as transit routes and bridges
where all the declaration information associated with a household taken. In the case of a transit route, the walking distance to

40 URISA Journal • Vol. 20, No. 1 • 2008


access the network is checked. Then the sequence of routes
is checked with the help of the transit network geometry.
• Finally, to proceed to overall interview validation. As needed,
the software checks the integrity of all answers using a
special procedure. This is usually performed following each
respondent’s answer and after the whole interview has been
completed. Warnings and error messages are displayed to the
interviewer. Because of productivity concerns, the interview
can be accepted even with such messages; final decisions are
made by supervisors.

In addition to the integrity of the interview itself, CATI, used


by all the interviewers, ensures the uniformity of the survey, for
it provides a single database for all locations, declarations, route
sequences, and so on, and facilitates postsurvey analysis. Figure 3
presents the trip declaration screen of the Montreal 2003 CATI
Figure 3. CATI software interface for trip declaration
software. Parts A and B show the household activity summary.
Locations and other circumstances of the declared trip are gath-
ered in the C section of the screen. The sequence of modes of
the trip is displayed in D for the current person. A detailed trip Meanwhile, the workstation stores a copy of each interview.
summary also is available in E. Finally, the F section is used to CATI also logs every entry made by the interviewer for later use.
display question and answers for the currently selected field in A model was estimated using these numerical logs to detect the
other sections of the screen. measurable variables that have a significant impact on interview
duration (Morency 2008, Chapleau 2003). In addition, the server
stores all raw XML files.
Real-Time Survey
Management Survey Statistics Management
To ensure real-time survey management, extended markup lan-
When conducting large household surveys, planners must be able
guage (XML) files were used for information exchange between
to follow survey activities on a day-to-day basis. Traditionally, this
workstations and the server. XML files were helpful because their
is accomplished through daily reports that are distributed among
data structure can change when needed. The server application
them. This generates large amounts of paper and is not always
controlled the distribution of the sample to the interviewer. This
suited to specific needs.
describes the process:
In 2003, the online survey statistics application provided
• From a workstation, CATI requests individual household
information on:
information as needed by an interviewer for an interview.
• Global productivity per day or per week, or of the whole
The request is written in an XML file.
survey (number of completed calls, trip rate per household,
• The server processes the XML file in sequential order to avoid
person, overall call status);
data collision and double sample distribution. Many criteria
• Productivity of each interviewer per day (completed calls,
are evaluated in the choice of a household: language (some
average interview duration, call status evolution);
interviewers can speak a foreign language), appointments
• Sample productivity (noncompleted calls per stratum, batch
made with respondents, batches that must be treated on a
progress, stratum household statistics, batch household
priority basis, and so on. Then it sends the information to
statistics);
CATI in the same XML file (header and workstation data
• Technical maintenance for the survey (list and types of errors,
are kept).
list of households to be completed, locations to be geocoded,
• CATI processes households one at a time. It completes
list of new trip generators, comments, error rates);
the XML file with the information that is gathered by the
• Daily maintenance (calls to a single household, interviewers
interviewer. This ensures data integrity, because the XML
on duty, list of current appointments);
file also is used for validation. Whether the interview is
• Reporting software management (list of reports, user accesses,
postponed, stopped, or completed, or an appointment is
log file).
made, the XML file is updated and sent back to the server.
The XML file also contains performance indicators for the
To generate reports, the application stores queries based on
interview, such as duration, number of errors, etc.
Structured Query Language (SQL), parameterized when needed
• The server receives the XML files from the workstations.
to input text, date, or number. Reports are available in four for-
It processes the data to transform and store them in the
mats, depending on their structure: paper-like reports in a PDF
centralized survey database.

URISA Journal • Trépanier, Chapleau, Morency 41


file, on-screen table-format reports, on-demand charts, or Excel Acknowledgments
spreadsheets.
The authors especially acknowledge the major contribution of
Conclusion Bruno Allard, research assistant, who implemented the software
Nowadays, it is our view that a large telephone household sur- and the methods. This work has been supported by the partners
vey in transportation cannot be conducted without the help of who conducted the 2003 survey: the Agence Métropolitaine de
powerful software components that will support all the functions Montréal (Montreal Transportation Agency), the Société de Trans-
surrounding this labor-intensive and costly activity. There is a port de Montréal (Montreal Transit Commission), the Société de
particular need for an adequate geographical information sys- Transport de la Ville de Laval (Laval Transit Commission), the
tem for both territorial and operational data. This GIS-T must Réseau de Transport de Longueuil (Longueuil Transportation
be comprehensive and its entities should be well identified for Network), and the Ministère des Transports du Québec (Quebec
searching for address, intersection, and postal code. Transit stop Ministry of Transportation).
locations and trip generators must be carefully selected to avoid
bias in location choice. In the Montreal region, this architecture About the Authors
has been dictated by the use of household survey data: transpor-
tation planning, transit network development, user behavior, Martin Trépanier, P. Eng., Ph.D., is a professor of industrial
transit financing, road network usage, trip generator analysis, and engineering at the Ecole Polytechnique de Montreal
so on. The same approach will be used for the survey that will be (Logistics and Transportation). His research is mainly related
conducted in the Greater Montreal Area in the Fall of 2008. to logistics, information systems, object-oriented modeling,
The contributions of such tools are multiple: GIS development, and Internet applications. He worked
• Better uniformity in interactions between interviewers and along with the MADITUC group on this research project.
respondents because the information is equally available
Corresponding Address :
among interviewers.
École Polytechnique de Montréal
• Immediate validation of most collected data, from single-
P.O. Box 6079
field values to complete transit itinerary.
Station Centre-Ville
• Autocorrection of sampling rates during survey. With the
Montreal, Quebec, H3C 3A7, Canada
help of the statistical tool, the sampling strategies can be
Phone: (514) 340-4711, #4911
modified anytime to reduce underreporting and spatial
Fax: (514) 340-4173
biases.
mtrepanier@polymtl.ca
• Improved interviewers monitoring and training. Rapid
Robert Chapleau, P. Eng., Ph.D., is a professor of civil
identification of failures and mistakes.
engineering (Transportation Planning) and founder-director
• Integrated call management: voice mails, appointments,
of the MADITUC group, Civil Engineering Department,
privacy concerns.
Ecole Polytechnique de Montreal. He developed the totally
• Improved overall quality of the household survey process,
disaggregate approach in transportation and is participating
from preparation to postvalidation. Better data helps
in several research projects in the Montreal area.
to minimize the need for technical resources, especially
Catherine Morency, P. Eng., Ph.D., is a professor of civil
transportation planners.
engineering (Transportation Planning) at the Ecole
Polytechnique de Montreal. Her research work covers
In the future, many elements will affect the way in which
household survey data analysis. She also is interested in urban
household surveys are conducted. The use of smart card payment
dynamics related to transportation. She is participating in
systems on a large scale will provide fresh data to update and
MADITUC research projects.
complete those obtained via telephone survey. The ever-growing
difficulties encountered in reaching people by telephone at home,
the increasing number of single-person households (these are hard References
to reach), and the growing use of cell phones that sometimes
replace home telephone service will challenge the traditional Badoe, D. A., and G. N. Steuart. 2002. Impact of interviewing
ways of conducting surveys and will prompt a rethinking of the by proxy in travel survey conducted by telephone. Journal
methods used. of Advanced Transportation 6, no. 1.
Bergeron, D. 2008. Enquêtes origine—destination métropolitaine:
bilan, expérimentation et choix pour l’avenir. 43e Congrès
de l’Association Québécoise du Transport et des Routes,
Québec.

42 URISA Journal • Vol. 20, No. 1 • 2008


Brög, W. 1997. Raising the standard! Transport survey quality Morency, C., M. Trépanier, and B. Agard. 2007. Measuring
and innovation. Keynote Paper for the 8th Meeting of transit use variability with smart-card data. Transport Policy
the International Association of Travel Behavior Research, 14, no. 3: 193-203.
Austin, Texas. Morency, C. 2008. Enhancing the travel survey process and
CERTU. 1998. L’enquête ménages déplacements méthode data using the CATI system. Transportation Planning and
standard. Collection du certu, Transport et mobilité. Technology 31, no. 2 (April 2008): 229-48.
Chapleau, R. 2003. Measuring the internal quality of a CATI National Research Council. 2003. Survey automation: Report
travel household survey. In Peter Stopher and Peter Jones, and workshop proceedings. Oversight Committee for the
Transport survey quality and innovation. Kruger, South Workshop on Survey Automation. In D. L. Cork, M. L.
Africa: Pergamon: 69-87. Cohen, R. Groves, and W. Kalsbeek, eds., Committee
Chapleau, R. 2000. Conducting telephone origin-destination on national statistics. Washington, D.C.: The National
household survey with an integrated informational approach: Academies Press.
Transport surveys: Raising the standard —Proceedings of Nicholls II, W. L. 1988. Computer-assisted telephone
an international conference on transport survey quality and interviewing: A general introduction. In R. M. Groves,
innovation, Grainau, Germany, Transportation Research P. P. Biemer, L. E. Lyberg, J. T. Massey, W. L. Nicholls II,
Circular EC-008. and J. Waksberg, eds., Telephone survey methodology. NY:
Chapleau, R. 1995. Symphonie d’usages des grandes enquêtes Wiley: 377-85.
origine-destination, en totalement désagrégé majeur. Opus O’Donnell, E., and J. Smith David. 2000. How information
Montréal 87 et 93, Huitièmes Entretiens du Centre Jacques- systems influence user decision: A research framework and
Cartier, Lyon, France. a literature review. International Journal of Accounting
Chapleau, R. 1986. Transit network analysis and evaluation Information Systems 1(2000): 178-203.
with a totally disaggregate approach. World Conference on Richardson, A. J. 2002. Current issues in travel and activity
Transportation Research, Vancouver. surveys. In H. Mahmassani, ed., In perpetual motion: travel
Chapleau, R., B. Allard, M. Trépanier, and C. Morency. 2001. behaviour research opportunities and application challenges.
Les logiciels d’enquête transport comme instruments Oxford UK: Pergamon: 341-57.
incontournables de planification analytique. Recherche, Richardson, A. J., E. S. Ampt, and A. H. Meyburg. 1995. Survey
Transport, Sécurité, Paris 70 (Janvier-Mars 2001): 59-77. methods for transport planning. Eucalyptus Press.
Chapleau, R., M. Trépanier, P. Lavigueur, and B. Allard. Trépanier, M., R. Chapleau, B. Allard, and C. Morency. 2003.
1997. Origin-destination survey data dissemination Trip generator relocation impact analysis methodology based
in a metropolitan context: A multimedia experience. on household surveys. Institute of Transportation Engineers
Transportation Research Record, Washington 1551: Journal (on the Web) Washington 73, no. 10.
26-36. Trépanier, M., R. Chapleau, and B. Allard. 2002. Geographic
Couper, M. P., R. P. Baker, J. Bethlehem, C. Z. F. Clark, J. Martin, information system for transportation operations: models
W. L. Nicholls II, and J. M. O’Reilly. 1998. Computer and specificity. Compte-rendus de la 30e conférence annuelle
assisted survey information collection. Wiley. de la Société canadienne de génie civil, Juin 2002, Montréal,
Griffiths, R., A. J. Richardson, and M. E. H. Lee-Gosselin. 559-68.
2000. Travel surveys, transportation in the new millennium. Trépanier, M., and R. Chapleau. 2001. Linking transit operational
Transportation Research Board, http://gulliver.trb.org/ data to road network with a transportation object-oriented
publications/millennium/00135.pdf. GIS. Journal of the Urban and Regional Information Systems
Jones, P., and J. Polak. 1992. Collecting complex household travel Association 13, no.2: 23-27.
data by computer. In selected readings in E. S. Ampt, A. Wermuth, M., C. Summer, and M. Kreitz. 2003. Impact of new
J. Richardson, and A. H. Meyburg, eds., Transport survey technologies in travel surveys. In P. Stopher and P. Jones.
methodology. Eucalyptus Press: 283-305. 2003. Transport survey quality and innovation. Kruger,
Link, M. W., and M. J. Kresnow. 2006. The future of random- South Africa: Pergamon: 455-81.
digit-dial surveys for injury prevention and violence research. Westrick, S. C., and J. K. Mount. 2007. Evaluating telephone
American Journal of Preventive Medicine 31, no. 5: 444- follow-up of a mail survey of community pharmacies.
50. Research in Social and Administrative Pharmacy 3, no. 2:
Liss, S. 2005. Planning for the next national household travel 160-82.
survey, Part 1. In Data for understanding our nation’s
travel, Transportation Research Circular, Number E-C071,
Transportation Research Board.

URISA Journal • Trépanier, Chapleau, Morency 43


Mapping Land-Use/Land-Cover Change in the
Olomouc Region, Czech Republic
Tomáš Václavík

Abstract: The Olomouc region in the Czech Republic has undergone significant changes in the past several decades, including
the change in political system of the country in 1989. Although the political and cultural transformation generally is recognized
as an important driver of land use (Ptáček 2000), few studies were conducted that would empirically assess and quantify land-
use/land-cover changes in the Czech Republic, especially in the context of the postsocialistic transformation (Fanta et al. 2004,
Zemek et al. 2005). This study presents an approach for identifying major land-use/land-cover changes in the Olomouc region,
applying remote-sensing techniques to compare data from multispectral satellite sensors acquired 12 years before and 12 years
after the revolution in 1989. The study closely covers specific trends in land-cover changes: changes in agricultural areas, forested
areas, and residential development. The results support initial assumptions that the land cover will reflect the changes in the hu-
man perception of landscape and natural resources, such as a smaller need for intensive agriculture, a shift to an environmentally
friendly management of forested areas, or increased development and suburbanization.

Introduction reaching the present 33 percent of the total vegetation cover in the
The Czech Republic currently is undergoing transformation from country (ÚHÚL 2006). Most of the forest is far from its natural
the centralized regime of a communist dictatorship towards a composition, for it was converted to monocultures of Norway
modern democratic state. Fanta et al. (2005) recognizes three main spruce (Picea abies), serving predominantly a productive function.
events in the last half century that had profound consequences for However, since the boom of environmental consciousness in the
the country and its land use. First, the communist coup d’état and 1990s, a distinctive tendency has grown towards alternative ap-
the following collectivization of land in the 1950s that introduced proaches in forest management considering the natural species
large-scale collective farming, especially intense in the Olomouc composition and potential vegetation (Neuhäuslová 1998).
region, which aimed at the maximum production of agricultural Residential development. As in other parts of Europe, the issue
commodities. Second, the abolition of the totalitarian political of suburbanization was well identified in the Czech Republic in
system in 1989, which was followed by the restitution of private the 1990s (Ptáček 1998; Jackson 2002). However, it is repre-
land ownership in the 1990s, the reintroduction of democracy sented by a relatively small extent of residential development in
and a market economy, and the development of market-driven vicinities of larger cities, and does not bear the typical traits and
forms of land use. Third, the preparation of the Czech Republic negative effects of the American-type large-scale suburban sprawl
for ingression into the European Union in 2004, including its as described by Václavík (2004).
complete association with the EU environmental and agricultural The main objective of this study is to analyze relevant remote-
policies, and its search for appropriate methods and forms of sensing data from 1976 and 2001 and to identify the locations,
land use. types, and trends of the main land-use and land-cover changes in
This research pays closer attention to specific trends in land- the past 25 years. Although the issue of land change is examined
use changes within the past 25 years: changes in agricultural areas, based on the background of political transformation of the coun-
forest areas, and residential development. These particular trends try, this article does not explicitly address the effect of political
can be described as followed. transitions on land-cover change. However, it was assumed that
Agricultural areas. Political transition in the Czech Republic the land cover will reflect some changes in the human perception
lead to marginalization of intensive agricultural areas, i.e., a pro- of landscape and natural resources, such as the decreased need
cess driven by a combination of socioeconomic and environmental for intensive agriculture, the shift to an environmentally friendly
factors caused by farming that ceased to be viable at many places, management of forested areas, or the increased development and
resulting in frequent abandonment of the agricultural land (Fanta suburbanization. The hypothesis is that the later satellite image
et al. 2004). Extensive areas of previously cultivated land in the of the Olomouc region study site will exhibit a smaller total area
country now are laying fallow or were converted to secondary of intensive agriculture and more meadows and pastures, fewer
grasslands—meadows and pastures. coniferous forests, and more mixed or deciduous tree cover, as
Forested areas. Since the time of their minimum extent at well as an overall higher residential development.
the end of the 18th century, forested areas have been increasing,

URISA Journal • Václavík 45


Methods the SPOT data for the study site was not freely available when it
was needed. Therefore, the Landsat Multispectral Scanner (MMS)
and Enhanced Thematic Mapper (ETM) scenes were acquired for
Study Site change detection analysis (see Table 1). The MSS data included
The study area chosen for this project is the Olomouc region in one scene (path 204, row 25) from May 8, 1976; the ETM+ data
the eastern Moravian part of the Czech Republic (see Figure 1). included two scenes (path 190, row 25, and path 190, row 26)
The study area of 5,012 km2 covers most of the Olomouc County from May 24, 2001. Described data sets were downloaded from
administration unit, one of the 14 administration units in the CR, the Global Land Cover Facility (GLCF) (http://glcf.umiacs.umd.
but the northeastern part overlaps to Moravskoslezsky County. edu/data/) through the Web interface and imported to IDRISI
The central part is formed by the wide alluvial plane of the up- geographic information system software using the GEOTIFF/
per stream of the Morava River, surrounded by the undulated TIFF conversion module.
hills of the Zabrezska and Drahanska uplands from the west and Ancillary sets of data were collected to support the land-
the Nizky Jesenik mountain range from the northeast, while change analysis. Two sets of scanned and georeferenced black-
the elevation ranges from 200 to 800 m a.s.l. The lowland areas and-white aerial photographs from 1970s and 1990 and a set
are highly urbanized, and include the major cities of Olomouc, of color orthophotographs from 2002 were obtained from the
Prerov, Zabreh, Sumperk, and others. Because of favorable climate Litovelské Pomoraví Protected Landscape Area Administration
and fertile soils, lowlands historically and currently represent the to serve as reference ground-truth data during the map classifica-
substantive agricultural areas in the Czech Republic. Despite its tion process. Vector data of the Czech Republic boundary and
intensive development, the core of the Olomouc region consists the Litovelské Pomoraví PLA area were obtained from the Czech
of the Litovelské Pomoraví Protected Landscape Area. This ex- Environmental Information Agency (CENIA) ArcIMS server
ceptional piece of natural landscape is formed by the naturally (http://geoportal.cenia.cz).
meandering Morava River and its several permanent and periodi-
cal branches with wetlands, meadows, and unique complexes of
floodplain forests, some of the few remnants in central Europe.
Image Processing and Classification
The other major forested habitats in the Olomouc region are
Acquired data sets were processed and examined in the Clark
located in the northeastern upland areas, and are predominantly
lab’s GIS software IDRISI 15.0, Andes edition. Figure 2 shows
composed of coniferous and mixed stands, which are used for
the steps of image processing and classification needed to achieve
timber production.
defined study objectives. After the satellite data were downloaded
from the Global Land Cover Facility and imported to IDRISI, it
Data Collection was assessed for image quality. While both ETM+ images did not
Because the study area is located in central Europe, the images exhibit any significant radiometric noise in the entire scene, the
from the high-resolution SPOT earth observation satellite would MSS image contained a fair amount of haziness in the northeast-
be the appropriate data source for the intended study. However, ern portion of the scene and subtle striping throughout the entire

Figure 1. Study area

Table 1. Acquired satellite images


Scene Num-ber Path/ Acquisition Date Sensor Format Spatial Resolution Bands
Row (m)
044-131 204/25 1976-05-08 Landsat MSS GEOTIFF 57x57 1-4
036-343 190/25 2001-05-24 Landsat ETM+ GEOTIFF 30x30 1 – 5, 7
036-344 190/26 2001-05-24 Landsat ETM+ GEOTIFF 30x30 1 – 5, 7

46 URISA Journal • Vol. 20, No. 1 • 2008


Figure 2. Work-flow diagram

area. As there were no meteorological data available for the time of resample module using parameters from the ETM image and map
MSS image acquisition, an absolute atmospheric correction could corners as ground control points was applied, producing a total
not be performed. Instead, the Principal Component Analysis root-mean-square error of 0.8 m, which is less than 0.5 pixels.
(PCA) was run, using standardized variance/covariance matrix The MSS 1976 and ETM 2001 images were classified using
and all four MSS bands as inputs. PCA created four principal the Maximum Likelihood supervised classification because most
component images in which the first two explained more than 98 of the land-cover mapping projects have applied either supervised
percent of the total variance and the remaining two components or unsupervised parametric classification algorithms to identify
contained most of the noise. The original four MSS bands were spectrally distinct groups of pixels (Smits et al. 1999). With
restored through an inverse PCA technique using the first two supervised classification, the spectral signatures of the known
components. land-cover categories are first developed, using digitized training
The study area of the Olomouc region is located in the sites. The software then uses a specific algorithm to assign all
overlap of the ETM images 036-343 and 036-344 from 2001. pixels in the image data set into defined land-cover classes (Jensen
A composite of the two overlapping images was created using 2004). The Maximum Likelihood classification is based on the
a mosaic technique by spatially orienting them and optionally probability density function that is associated with a particular
balancing the numeric characteristics of the image set based on training site signature. All pixels are assigned to the most likely
the overlapping areas. The average mosaic method was applied category based on an evaluation of the subsequent probability
to average the base image values with the adjusted overlap image that the pixel belongs to the signature (class) with the highest
values. In addition, the WINDOW module, extracting subimages probability of membership (Jensen 2004). Seven land-cover
from the set of original images, was utilized to isolate the desired categories were recognized in the Olomouc region: water, de-
extent of the study area. ciduous forest, coniferous forest, mixed forest, developed (urban)
The last step before actual image classification was to syn- areas, areas of (intensive) agriculture, and meadows (grassland).
chronize the spatial resolution of the images from both times. The Training sites were digitized based on the personal knowledge of
original resolution of the MSS image was 57x57 meters. For the the study area and ground-truth data of aerial photographs and
purpose of its comparison with the ETM image with resolution orthophotographs. Spectral signatures of individual land-cover
of 30x30 meters, the MSS image needed to be resampled. The classes were developed and assessed for their separability. Spectral

URISA Journal • Václavík 47


Figure 5. Gains and losses between 1976 and 2001 in km2

Figure 3. Land-cover map of 1976

Figure 6. Net change between 1976 and 2001 in km2

of qualitative data (Eastman 1995). It overlays two images and


calculates all their possible combinations of classes. In the case
Figure 4. Land-cover map of 2001 where images represent the same land-cover categories at different
times, persistence occurs where areas fall in the same land-cover
categories, and change occurs where a new category is created
(Eastman 1995). IDRISI Andes offers an efficient and easy-to-use
values of urban areas and agricultural areas with bare soil were
tool for rapid assessment of land-cover change and its implications
mixing, therefore their training sites had to be redefined, and also
based on cross-clasification principles. The Land Change Modeler
the texture analysis using Dominance index and kernel window
(LCM) for Ecological Sustainability allows a user to evaluate gains
of 5x5 pixels was conducted. Finally, the Maximum Likelihood
and losses in land-cover classes, land-cover persistence, and specific
classification was run with original bands and the texture image
transitions between selected categories. Using the classified land-
as inputs, producing two final land-cover maps of 1976 and 2001
cover maps from 1976 and 2001 as input parameters, this tool
that were compared.
was applied to identify the locations and magnitude of the major
An error matrix was constructed to estimate classification
land change, land persistence, and trends in transitions between
accuracy of produced land-cover maps. The error matrix provides
land-cover categories in the study area.
a basis for characterizing types of errors by cross-tabulating the
classified land-cover categories in sample locations against those
observed in ground reference data (Smits et al. 1999, Foody 2002). Results
A random sampling scheme was applied to define ground-truth Figures 3 and 4 represent the results of Maximum Likelihood
locations (n = 100) and the aerial photographs from 1970s and classification: land-cover maps depicting the situation in 1976
2002 were used to check for the “true” land-cover classes. The and 2001. The change analysis tool provides efficient statisti-
overall accuracy was calculated for both maps. This represents cal assessment of changes in individual land-cover categories.
the probability that any point on the land-cover map is assigned Its results in Figures 5 and 6 demonstrate that there have been
exactly the same category by the classifier, as the category that is significant changes in all land-cover/land-use categories between
identified in the ground-truth sites (Wulder and Franklin 2003). 1976 and 2001 with the exception of water, where the subtle
In addition, the producer’s and user’s accuracies that measure change can be caused by location errors in land-cover classifica-
omission and commission errors were estimated for individual tion. Concerning the net change, which represents the earlier
land-cover classes. area of a category with added gains and subtracted losses, three
A cross-classification procedure is one of the fundamental land-cover categories experienced major transitions. The total area
pairwise comparison techniques used to compare two images of meadows (grassland) increased by 942 km2, while the area of

48 URISA Journal • Vol. 20, No. 1 • 2008


Figure 7. Land-cover persistence and change

intensive agriculture decreased by 592 km2, as well as the area of


coniferous forests, which decreased by 603 km2. The category of
developed (urban) area also was affected by distinct change, with
a net gain of 127 km2.
A simplified cross-classification map (see Figure 7) represents
persistence in land-cover categories, areas where no change oc-
curred, and land-cover change, areas with any type of transitions
between categories (depicted in black). However, the land-change
and persistence map is difficult to visually interpret if the areas
of individual land-cover classes are not clustered, and also the
type of change is not represented in this map. Therefore, the
contribution to net change, i.e., the transition between specific
classes, was calculated to achieve the objectives of the study.
Data in Figure 8 represent the contribution to net change for
categories of meadows (grassland), developed (urban) area, and
mixed forest. They reveal that agricultural areas explain about
63 percent of the total increase in meadows, new development
occurred predominantly on former agricultural areas (more than
56 percent), and about 16 percent of previous coniferous forests
currently is identified as mixed forest. Figure 8. Contribution to net change in selected categories (km2)
Analysis of the error matrices revealed that the proportion
of agreement between land-cover categories in the classified map
and the ground-truth data was 77 percent for the 1976 period and identify the salient locations, types, and trends of the land-cover
81 percent for the 2001 period. The producer’s accuracy was the change in the past 25 years. The results support initial assump-
lowest for the agriculture class in both 1976 and 2001 maps (65 tions based on general knowledge of some of the land-use drivers
percent and 70 percent), as some of the agricultural areas were in different times. There have been significant losses in categories
classified as meadows or developed. The user’s accuracy was the of intensive agricultural areas and coniferous forest, and gains
lowest for the mixed forest class (60 percent) in the 1976 map in meadows and developed areas. From the former agricultural
and for the developed class (65 percent) in the 2001 map. Some areas, 23 percent became meadows and pastures, especially in the
sites with mixed forest were falsely identified as the coniferous or northeastern hilly part of the study site, and 3 percent was devel-
deciduous class in the reference data. Some developed sites were oped in the lowlands around the Litovelské Pomoraví Protected
falsely identified as agriculture or coniferous forest categories. Landscape Area. About 16 percent from the previous coniferous
forest in the eastern hilly part of the region currently was identi-
fied as mixed forest.
Discussion and Conclusions This study provides no empirical evidence of direct causality
This study applied remote-sensing techniques to classify satellite between discovered land-cover/land-use change in the study site
imagery of the Olomouc region, Czech Republic, from 1976 and political and cultural transformation of the country; however,
and 2001, from years before and after a major political change in the location and trends of observed land change suggest there
the country, and compared the two resulting land-cover maps to might be distinct correlation. Concerning the transition from a

URISA Journal • Václavík 49


category of intensive agriculture to a category of meadows, the Acknowledgments
major trend was observed in the northeastern uplands of the study
site. This observation is consistent with suggestions of Zemek et The author gratefully acknowledges Yelena Ogneva-Himmelberger
al. (2005) that the marginalization of agricultural areas occurs and John Rogan, professors from Clark University, for supervising
first at locations with unfavorable natural conditions, especially in this work and the Fulbright program for enabling the author’s
uplands where the agricultural production was previously forced training in the United States.
by an extensive use of fertilizers and pesticides. Concerning newly
developed areas, the major trend occurred especially in the central
lowland area of the study site around the Litovelské Pomoraví
Protected Landscape Area. This observation is consistent with
About the Author
the general suburbanization process in central Europe where
Tomas Vaclavik was a Fulbright exchange student in the program
new residential areas tend to be developed in the form of “satel-
of Geographic Information Sciences for Development and
lite” towns in the vicinity of existing cities and recreational areas
Environment at Clark University, Worcester, Massachusetts,
(Ptáček 1998). Regarding transition from coniferous tree cover
in the academic year 2006-2007. He earned both his
to mixed forest, this change was observed predominantly in the
bachelor’s and master’s degrees in Ecology and Environmental
northeastern hilly part of the study site, where the elevation and
Sciences at Palacky University in the Czech Republic. He
associated environmental conditions favor potential vegetation
recently has been working as a GIS analyst for the Agency for
of mixed and deciduous forest stands. This fact correlates with
Nature Conservation and Landscape Protection of the Czech
the general diversion in forest management in the past 15 years
Republic. Currently, he is pursuing his Ph.D. in Geography
from clear-cut practices and spruce and pine plantations to the
at the University of North Carolina at Charlotte, focusing
alternative use of native deciduous species of trees in the lower
on applications of GIS in ecological research and working
and middle altitudes of the country.
as a research assistant in the Center for Applied GIScience
Classification of multispectral satellite data and comparison
(CAGIS).
of land-cover maps are essential tools for assessing large-scale land-
cover/land-use changes. However, this research left considerable Correspondence Address:
room for future improvement. Visual comparison of classified Center for Applied Geographic Information Science (CAGIS)
maps with training sites as well as accuracy statistics calculated Department of Geography and Earth Sciences
from error matrices showed inaccuracies in the classification University of North Carolina at Charlotte
process. Spectral mixing was apparent between the classes of de- 9201 University City Boulevard
veloped and agriculture areas where barren soil was present and Charlotte, NC  28223
the texture analysis did not eliminate all of it. Also, all Landsat Phone: (704) 687-5963
scenes were acquired in the spring season, when certain types tova@email.cz
of crops, such as cereals, are in a phenological stage that exhibit tvaclavi@uncc.edu
similar spectral response as meadows and pastures. In addition, http://tova.euweb.cz
the MSS imagery from 1976 suffered from large amount of hazi-
ness and radiometric noise, which were not entirely removed by
the principal component analysis and could distinctively affect
image classification. An effort to collect better-quality remote-
sensing data, such as the ones from the SPOT sensor, should be
made to improve the overall accuracy of land-change assessment.
Finally, the Maximum Likelihood classification results might have
been improved if unequal prior probabilities of the land-cover
classes had been assigned. Alternatively, decision-tree classification
techniques that derive probabilities of land-cover classes from
the distribution in the training data (Rogan et al. 2002) may be
considered for future improvement of the analysis.

50 URISA Journal • Vol. 20, No. 1 • 2008


References Ptáček, P. 2000. Networking and local culture: Local community
transformation after 1989 on the example of Olomouc, Czech
Eastman, J. R. 1995. Change and time series analysis. Worcester, Republic. Acta Universitatis Palackianae—Geographica 36:
MA: United Nations Institute for Training and Research/ 59-64.
Clark Labs for Cartographic Technology and Geographic Rogan, J., J. Franklin, and D. A. Roberts. 2002. A comparison
Analysis, Clark University, 28. of methods for monitoring multitemporal vegetation
Fanta, J., K. Prach, and F. Zemek. 2004. Status of marginalisation change using thematic Mapper imagery. Remote Sensing of
in Czech Republic: Agriculture and land use. EUROLAN, Environment 80, no. 1: 143-56.
National Report—Czech Republic. Faculty of Biological Smits, P. C., S. G. Dellepiane, and R. A. Schowengerdt. 1999.
Sciences, University of South Bohemia, and Institute of Quality assessment of image classification algorithms for
Landscape Ecology—Academy of Sciences of the Czech land-cover mapping: a review and proposal for a cost-based
Republic, 29. approach. International Journal of Remote Sensing 20, no.
Fanta, J., F. Zemek, K. Prach, M. Heřman, and E. Boucníková. 8: 1461-86.
2005. Strengthening the multifunctional use of European ÚHÚL. 2006. Report on the state of forests and forestry in the
land: Coping with marginalization. EUROLAN, National Czech Republic 2006. Ústav pro hospodářskou úpravu lesa
Report—Czech Republic. Faculty of Biological Sciences, (Forest Management Institute), 128.
University of České Budějovice, Institute of System Biology Václavík, T. 2004. The use of GIS in ecological planning. (A case
and Ecology, Academy of Sciences of the Czech Republic, study of Mount Desert Island). Master’s thesis, Department
41. of Ecology and Environmental Sciences, Palacky University,
Foody, G. M. 2002. Status of land cover classification accuracy Olomouc, 82.
assessment. Remote Sensing of Environment 80: 185-201. Wulder, M. A., and S. E. Franklin. 2003. Remote sensing of forest
Jackson, J. 2002. Urban sprawl. Urbanismus a územní rozvoj environments. Concepts and case studies. Norwell, MA:
(Urbanism and urban development) 5, no. 6: 21-28. Kluwer Academic Publisher, 519.
Jensen, J. R. 2004. Introductory digital image processing. Upper Zemek, F., M. Heřman, Z. Mašková, and J. Květ. 2005.
Saddle River, NJ: Pearson Prentice Hall, 544. Multifunctional land use—a chance of resettling abandoned
Neuhäuslová, Z., et al. 1998. Map of potential natural vegetation landscapes? (A case study of the Zhůří territory, the Czech
of the Czech Republic. Prague: Academia, 341. Republic). Ecology 24, no. 1: 96-108.
Ptáček, P. 1998. Suburbanizace—měnící se tvář zázemí velkoměst
(Suburbanization—the changing face of metropolitan
hinterland). Geografické rozhledy (Geographical views) 7,
no. 5: 134-37.

URISA Journal • Václavík 51


Mapping the Future Success of Public Education
Donna L. Goldstein M. Ed

ABSTRACT: For better or worse, computers have revolutionized every aspect of our lives. As we quickly make the transition from
an industrial to an information age, computer literacy skills have become a basic necessity. Technology skills are now referred to
as the “Fourth R” in education, as coined by Michael Goodchild. To successfully learn and use GIS (Geographical Information
Systems) technology, one must incorporate the skills of reading, writing, and arithmetic. Understanding and utilizing a GIS
system requires a holistic combination of reading instructions, data, and maps; writing hypotheses, reports, and presentations;
and using arithmetic to understand queries and spatial analysis. Thus the 4th R as it relates to GIS is a new elevated skill that
incorporates the three original R’s in education. Teaching GIS may be just the boost our public educational system needs to
adequately prepare students for entrance into the emerging global society.

Mapping the Future Success thinking spatially and to provide them with the opportunity to
compete in an international society.
of Public Education
As we quickly move forward in the information age our nation’s
public school systems are placed precariously at a crossroad. While The Significance of GIS in
most school districts strive to incorporate new technologies, many Public School Curriculum
teachers simply don’t have the luxury of time to teach material that We live in a global society where competition for those skilled in
may not be directly aligned with state and federally mandated high geospatial technologies will only increase. Governments interna-
stakes tests, including new computer technology skills. There is tionally have paid close attention to the evolving technological
an urgent need to educate our students in technology, driven by shifts and view the development of technology skills as the foun-
the business and world economies. We must become a premier dation of the country’s future. So must we if we want our next
digital nation or face the consequences of not taking action. The generation to be competitive and viable in the global marketplace.
ramification of these students falling through the cracks strikes For the sake of our future and the future of our youth, we have
at the heart of public education and bankrupts all stakeholders: a responsibility to teach them how to effectively utilize the new
parents, teachers, school administrators, students, and society at and emerging geospatial technologies that are increasingly needed
large. Teaching GIS and geospatial technology may be just the in our new world.
boost our public educational system needs to adequately prepare Social and financial implications of GIS prompted the “U.S.
students for entrance into the emerging global society. Department of Labor to identify GIS as one of the three most
important emerging and evolving fields, along with nanotechnol-
Why GIS ogy and biotechnology, with over 900,000 additional jobs in the
Many young students today have not known a world without U.S. in geospatial technology expected from 2002 to 2012” (U.S.
computers; they are naturally curious and gravitate to the tech- Department of Labor Employment & Training Administration,
nology. The GIS system is a perfect vehicle to deliver necessary 2005). In addition “NASA says that 26% of their most highly
content and contextualize the lesson so that students will not only trained geotech staff is due to retire in the next decade, and the
be engaged but will be motivated to gain the knowledge presented. National Imagery and Mapping Agency is expected to need
Using this application in the classroom promotes critical thinking 7,000 people trained in GIS in the next three years” (Gewin and
skills for students in addition to honing their communication Virginia, 2004). With the aging of our many professionals in the
and presentation skills. current geospatial workforce, there are a number of organizations
In 2005, the National Research Council published a report seeking to recruit the next generation.
entitled “Learning to Think Spatially: GIS as a Support System What this means is that those of us in the geospatial industry
in the K-12 Curriculum” that identified the importance of have an obligation to our chosen field, to ensure its continued
promoting spatial thinking skills across curriculum subjects. As growth and success. By not preparing the next generation to take
indicated by the report, GIS has the potential to successfully the helm of this industry we turn our backs on providing for the
cultivate those skills. The overarching goal of this educational future sustainability of GIS and spatial technology.
initiative is to create the next generation of students skilled in

URISA Journal • Goldstein 53


One School District’s book, and assistance from administrative GIS staff, this individual
rose to the occasion and introduced her students to the future.
Response Their enthusiasm for GIS is evident, as students look forward to
Since the 1980’s the Palm Beach County School District’s GIS
the lab sessions and pummel questions as to what types of jobs
system was primarily used by staff on the operational or business
are available in GIS and where is it used. The program was so
side of the house for facility planning, enrollment projections
successful that the principal, Jack Thompson, has requested that
and transportation. During 2006 the Planning Director, Kristin
this teacher incorporate GIS into all of her classes next year.
Garrison and the GIS Coordinator recognized the tremendous
In addition to this recent success, plans are underway to
benefits GIS could offer students including enhanced learning
quickly expand GIS instruction throughout the school system.
techniques and potential career opportunities. This revelation
Projects in the pipeline include the development of Criminal
prompted communication with academic staff regarding imple-
Justice Career Academies and an in-house GIS Career Academy
menting GIS into the classroom curriculum. Enthusiasm for the
at a high school. Another exciting project involves a GIS/GPS
virtues of GIS in the educational environment quickly spread
science and social studies program for classroom instruction at
and a GIS Steering committee was assembled. The committee
six schools. This endeavor is a joint effort with resources from
involved participants across the spectrum of school district de-
both the School District and South Florida Water Management
partments including staff from Facilities, Transportation, Career
(SFWMD) and involves the development and delivery of a hands-
Academies, and Educational Technology, as well as social studies
on workshop for 27 teachers.
and science curriculum planners. The coordination of the business
and academic units of the organization provided a foundation to
methodically develop a plan. Participate in the Challenge
A GIS Educational Model (GEM) Charter was developed, a Skilled and experienced GIS professionals and those business
document that outlined the policy and structure for implement- and government agencies who utilize GIS can greatly enhance
ing GIS into the classroom. The charter also included how GIS the efforts to expand GIS into our public school system. There
directly addressed a number of school board-approved district are many unique ways to “pay it forward.” Contact your local
goals and policies. With a specific timeline developed to reach schools and see if there is an interest (or extol the virtues of GIS
various milestones, work began on acquiring a district-wide GIS to curriculum administrators and generate the excitement). You
site license from ESRI. The cost for this district-wide license was can offer to present or demonstrate GIS to a classroom of students
less than the price for the individual seats purchased for a select or find a teacher interested in using GIS and provide technical
group of operational staff and was the springboard needed to support and mentoring (GoTo meetings and web conferencing
push GIS into the classrooms. tools make this easier than ever). If you institute a program of
Recognizing that teachers hold the key that unlocks the po- internship for students, your ROI may be finding the GIS Tech
tential for implementation of any new program, the GIS Steering or GIS web programmer you’ve been looking for. Donate your
Committee provided a professional development workshop for replacement PCs or hardware to local schools. Your old computers
social studies teachers designed to spark their interest. Stipends are probably superior to what many schools are currently using.
and in-service points were included. A training room was config-
ured with the GIS software loaded on PCs for 30 teachers. The Future Possibilities
workshop included hands-on GIS lessons from the Mapping Our young people represent this nation’s greatest commodity and
Our World Lesson plan book (Malone et.al. 2005). The goals our future rests in their ability to succeed. We will reap the seeds
of this exercise were (a) to expose teachers to the benefits of we sow. If the U.S. in general and school districts in particular
incorporating GIS into their classroom curriculum (b) to have are remiss in providing opportunities for our teachers and stu-
them experience the ease of use, and (c) to prepare them to easily dents to learn new geospatial skills, we will be left behind in this
transfer what they learned directly to the classroom by following technological age. If we continue to focus on high stakes testing
the GIS lesson plans. at the cost of introducing new technologies, countries such as
China and India will surely pass us and the next generation will
A Success Story be poorly equipped to compete (Friedman, 2005). Administrators,
The barriers that Kerski points out in his article, “A National parents, community leaders and businesses must recognize that
Assessment of GIS in American High Schools, 2001,” are still we need to introduce new and improved methods for education.
prevalent today, seven years later. However, slow progress is be- Our future is at stake.
ing made in a number of schools. One such success story is Boca It’s time to answer the call of progress and chart a course for
Raton Middle School in Palm Beach County Florida. During the public education that elevates both the students and our industry.
summer of 2007 a teacher from this school attended the one day Perhaps the greatest benefit of including GIS in public education
workshop on GIS and took on the bold task of implementing GIS is societal: it is the legacy we leave for our next generation – the
into several of her classes this past year. Armed with little more opportunity to succeed and compete in the global society.
than a one-day training session, a Mapping our World lesson plan

54 URISA Journal • Vol. 20, No. 1 • 2008


URISA’s Role in K12 Education About the Author
Recognizing the imminent shortfall in trained GIS profession-
als to fill the upcoming void, URISA has embraced the topic of Donna Goldstein received her Masters in Educational Leadership,
introducing GIS in primary and secondary education. In fact and is currently attending Florida Atlantic University, pursu-
URISA now maintains a list of over 160 schools offering related ing a PhD in Educational Leadership, Adult and Community
programs. education. Ms. Goldstein was employed by Palm Beach County
In addition, at its 2007 Annual Conference the URISA Planning Zoning & Building from 1990 thru 2001 as the GIS
Program committee organized a special track for K-12 educators, Supervisor. In 2001 she accepted the position of GIS Coordina-
which included sessions to demonstrate the practical approaches tor with the Palm Beach County School District and is currently
for K12 teachers to implement GIS into their classroom ac- employed in that capacity.
tivities. While attendance was good the unfortunate truth is it’s
difficult for public school teachers to find the time and funding References:
to attend conferences, especially those workshops that are not
related to core academic topics covered in the high stakes testing Friedman, Thomas L., 2005). The World is Flat, A Brief History
requirements. of the Twenty-first Century, New York: Farrar, Straus and
The solution to this critical issue rests with GIS professionals Giroux, 257.
reaching out to the K12 school system. To address this URISA Gewin, Virginia (January 22 2004). Mapping opportunities.
also included a session to engage GIS practitioners to demonstrate Nature, International weekly journal of science opportunities.
how they can volunteer their services to promote the use of GIS 427, 376. Retrieved June 26, 2007 http://www.nature.com/
in local school programs. One such example of integrating the nature/journal/v427/n6972/full/nj6972-376a.html
business community in the GIS education of K12 students is the Goodchild, Michael (2006) The Fourth R? Rethinking GIS
non-profit group Hopeworks, a non-profit organization from Education. ESRI ArcNews. Retrieved April 24, 2008 from
Camden NJ whose mission is to educate intercity K12 youth. http://www.esri.com/news/arcnews/fall06articles/the-
This organization successfully promotes integration of business fourth-r.html
and private sector resources to help fulfill the goal of educating Kerski, Joseph (2001). A National Assessment of GIS in American
youth with marketable skills in the field of GIS (Staff 2006). High Schools. International Research in Geographical and
As URISA becomes more involved with this very timely Environmental Education. 10(1) Retrieved April 3, 2008
topic we will provide additional information on how businesses from http://www.multilingual-matters.net/irgee/010/0072/
can reach out to the community and reap mutually beneficial irgee0100072.pdf
rewards. Check out the K12 track at URISA’s 46th Annual Malone, Lyn, Palmer, Anita M., Voight, Christine L., Eileen
Conference & Exposition October 7-10, 2008 in New Orleans. Napoleon, and Feaster, Laura (2005). Mapping Our World:
And continue to look at URISA for ways to be part of the ArcGIS Desktop Edition, GIS Lessons for Educators. 3-47,
part of the solution. 215-239. Redlands Ca: ESRI Press
National Research Council’s Board on Earth Sciences and
Acknowledgements Resources of the Division of Earth and Life Studies. Report
in Brief. (2005). Learning to Think Spatially: GIS as a
The author would like to acknowledge the many contributions Support System in the K-12 Curriculum, Washington, D.C.:
provided by Kristin Garrison, AICP, over the past 17 years. In The National Academies Press, Retrieved August 21, 2007
her capacity as Executive Director of Planning Zoning and Build- from http://dels.nas.edu/dels/rpt_briefs/learning_to_think_
ing for Palm Beach County she supported and assisted with the spatially_final.pdf
emerging GIS system and contributed greatly to the success this Staff (2006) Hopeworks Founder Father Jeff Putthoff Encourages
countywide effort enjoys today. In addition as Planning Director Youth Development with Technology Training. ESRI ArcNews
for the Palm Beach County School District her vision, support, Retrieved April 23, 2008 from http://www.esri.com/news/
encouragement and unwavering belief in the value of GIS in the arcnews/fall06articles/hopeworks-founder.html
K-12 educational system has paved the way for this program to U.S. Department of Labor Employment & Training Administration
come to fruition. The author would also like to extend apprecia- (2005). High Growth Industry Profile. Retrieved July 5, 2007
tion to Dr. Lucy Guglielmino, without whom this article would from http://www.doleta.gov/BRG/Indprof/geospatial_
not have been developed. Many thanks for the editorial assistance profile.cfm
and perseverance in the production of this piece.

URISA Journal • Goldstein 55


Learn more about
GIS Professional Certification at
www.gisci.org
Extend your GIS Across the Enterprise.
Share the value of your work throughout your organization with Server GIS.

As a geographic information system could query accurate, up-to-date data


(GIS) professional, you receive countless without a lot of training. This increases “The server-based options
requests for maps and spatial informa- their productivity as well as yours. provided by ESRI allow the City of
tion. Answering special requests, while Mesa to deliver more enterprise
By making your maps, data, and
important, takes time and reduces GIS applications throughout
analyses readily available, you can
your productivity. our organization without high
help others reap the benefits of the GIS
Imagine being able to push your work that you do. You already know installation and training costs.”
maps, models, and tools out to the that spatial analysis and visualization Jason Bell
rest of the organization via focused, are important parts of daily decision IT Services Leader,
easy-to-use applications. Staff in making. Use Server GIS from ESRI to City of Mesa, Arizona

other departments and out in the field help others benefit from your work.

To learn how organizations are using ESRI server GIS


to improve productivity, visit www.esri.com/server.
In the United States: 1-866-447-3036
Outside the United States: +1-909-793-2853, extension 1-1235
On the Web: www.esri.com/international

Copyright © 2007 ESRI. All rights reserved. The ESRI globe logo, ESRI, and www.esri.com are trademarks, registered trademarks, or service marks of ESRI in the United States, the European Community, or certain other jurisdictions. Other companies and products mentioned herein may be trademarks or registered trademarks of their respective trademark owners.

Вам также может понравиться