Practical GIS

Analysis
About the Author
David L. Verbyla is an Associate Professor of GIS/Remote Sensing with the Department
of Forest Sciences, University of Alaska. He has a B.S. from Cook College, Rutgers Uni-
versity in Wildlife Management, an M.S. from Michigan State University in Park and
Recreation Resources, and a Ph.D. from Utah State University in Forest Resources with
minors in Applied Statistics and Soil Science. Dr. Verbyla has taught GIS workshops and
university courses at Utah State University, University of New Hampshire, University of
Idaho, University of Montana, and the University of Alaska.
Practical GIS
Analysis
David L. Verbyla
London and New York
First published 2002
by Taylor &Francis
11 New Fetter Lane, London EC4P 4EE
Simultaneously published in the USA and Canada
by Taylor &Francis
29 West 35th Street, New York, NY 10001
Taylor &Francis is an imprint of the Taylor &Francis Group
2002 Taylor &Francis
ArcView Graphical User Interface is the intellectual property of
ESRI and is used herein with permission.
Copyright 1997 ESRI. All rights reserved.
All rights reserved. No part of this book may be reprinted or
reproduced or utilised in any form or by any electronic,
mechanical, or other means, now known, or hereafter invented,
including photocopying and recording, or in any information
storage or retrieval system, without permission in writing
from the publisher.
Every effort has been made to ensure that the advice and
information in this book is true and accurate at the time of going
to press. However, neither the publisher nor the authors can
accept any legal responsibility or liability for any errors or
omissions that may be made. In the case of drug administration,
any medical procedure or the use of technical equipment
mentioned within this book, you are strongly advised to consult
the manufacturers guidelines.
British Library Cataloguing in Publishing Data
A catalogue record for this book is available from the
British Library
Library of Congress Cataloging-in-Publication Data
A catalog record for this title has been requested.
ISBN 0-415-28609-3
This edition published in the Taylor & Francis e-Library, 2003.
ISBN 0-203-21793-4 Master e-book ISBN
ISBN 0-203-27350-8 (Adobe eReader Format)
(Print Edition)
Contents
Preface ix
Chapter 1 GIS Data Models 1
Introduction 1
Point Themes 4
Line Themes 5
Network Themes 6
Dynamic Segmentation 7
Polygon Themes 7
Grid Themes 8
Image Themes 9
Tools for Managing GIS Features 9
Tools for Building Attribute Tables 10
Exercises 12
Chapter 2 GIS Tabular Analysis 15
Introduction 15
Selecting Table Records 17
Descriptive Statistics 19
Summarizing Tables 20
Other Commonly Used Tabular Tools 21
Indexing Attributes 22
Tabular Analysis Exercises 22
Chapter 3 Point Analysis 27
Introduction 27
Area Analysis Tools 28
Distance Analysis Tools 30
Attribute Analysis Tools 33
Point Analysis Exercises 36
Chapter 4 Line Analysis 43
Introduction 43
Dissolve 43
Countvertices 45
Buffer 45
Reselect 47
Intersect 48
Line Analysis Exercises 50
Chapter 5 Network Analysis 59
Introduction 59
Address Geocoding 60
Optimal Routing 61
Finding Closest Facilities 67
Resource Allocation 69
Turn Tables in Networks 72
Network Analysis Exercises 73
Chapter 6 Dynamic Segmentation 77
Introduction 77
Routes and Sections 78
Event Types 79
Creating Events 80
Dynamic Segmentation Exercises 93
Chapter 7 Polygon Analysis 103
Introduction 103
Dissolve 104
Eliminate 105
Clip 106
Split 107
Erase 108
Update 109
Spatial Joins: Intersect, Union, Identity 110
Polygon Analysis Exercises 111
Chapter 8 Grid Analysis 119
Introduction 119
Grid Arithmetic Operators 120
Selection Operations: TEST, SELECT, CON 121
Grouping Operations: RECLASS, REGIONGROUP 124
Distance Operations: EXPAND, EUCDISTANCE 126
Optimal Path Operations: COSTDISTANCE, COSTPATH 127
Topographic Operations: SLOPE, ASPECT 133
Grid Analysis Exercises 134
Chapter 9 Image Analysis 143
Introduction 143
Image Display 143
Linear Contrast Stretch 146
Image Rectification 149
The Affine Transformation 149
Rectification Model Errors 153
Image Resampling Options 155
Unsupervised Classification 155
Supervised Classification 158
Seed Pixel Training Fields 159
Image Classifiers 160
vi PRACTICAL GIS ANALYSIS
CONTENTS vii
Maximum Likelihood Classification 161
Classification Accuracy Assessment 162
Reference Data 162
Error Matrix 163
Producers and Consumers Accuracy 163
Image Analysis Exercises 164
Chapter 10 Vector Exercises 171
Introduction 171
Real Estate Application 175
Moose Habitat Analysis 177
Fire Hydrant Inspection Application 181
Forest Fire Application 182
Emergency Phone Application 182
Chapter 11 Grid Exercises 183
Introduction 183
Real Estate Application 184
Watershed Application 185
Waterfowl Habitat Analysis 186
Groundwater Application 187
Moose Habitat Analysis 188
Chapter 12 Saving Time in GIS Analysis 189
Introduction 189
Work While You Play 189
Avoid TelephoneTag 190
Painless Documentation 190
Assume Your GIS Lies 192
Less Can Be Better 196
GIS Tables Issues 197
Consider Abandoning Your GIS 197
Do Not Repeat Yourself 197
Indexing Features and Attributes 199
Appendix Exercise Solutions 201
Data Model Exercise Solutions 201
Tabular Analysis Exercise Solutions 205
Point Analysis Exercise Solutions 211
Line Analysis Exercise Solutions 220
Network Analysis Exercise Solutions 232
Dynamic Segmentation Exercise Solutions 240
Polygon Analysis Exercise Solutions 252
Grid Analysis Exercise Solutions 260
Image Analysis Exercise Solutions 268
Vector Exercise Solutions 278
Grid Exercise Solutions 285
Index 291
ix
Preface
Geographic Information System (GIS) analysis is basically spatial problem-solving. It can
be quite challenging at first, especially if you get bogged down in syntax and details. The
hardest part is to conceptually understand the problem, and the GIS tools available to
solve your problem. This book will help you understand conceptually how various GIS
tools work. So for now, forget about workspaces, libraries, spatial database engines,
dangling nodes, fuzzy tolerances, requests and instances. The syntax and details do vary
by GIS . . . yet conceptually the analysis tools used by GISs are pretty much the same.
Start by building a foundation of understanding about GIS tools at a conceptual level.
Test your understanding by trying the exercises at the end of each chapter.
This book is written so you can learn important concepts away from your GIS . . . at
the laundromat, airport, on a bus, or sitting under a tree in a park. My main GIS expe-
rience is with ESRIs ARC/INFO and ArcView. Therefore, some of the terms used in the
text are from these products. However, you do not have to be a user of these products to
learn conceptual GIS analysis from this book.
I have been teaching GIS at universities and workshops since 1989. Since then, my
students have endured earlier drafts of this book and have taught me many GIS pitfalls
and GIS applications that I would have never imagined. I thank these students for help-
ing me. I also thank the faculty and staff at the University of Alaska, where I started this
book, and also the folks at the Numerical Terradynamics Simulation Group (NTSG) Lab
at the University of Montana, where I finished this book while on sabbatical there. I am
grateful for their support, friendship, and kindness.
David Verbyla
Becker Ridge, Alaska
Once you have figured out that INTERSECT is the tool to use to solve your problem,
you can look up the specific usage of that tool in your GIS. By first conceptually solving
the problem, you can avoid becoming confused with thousands of GIS commands and
detailed options.
Once you know which tools to use, then you can concentrate on how to use them in
your GIS by reading the help manuals. For example, the conceptual solution is applied in
the ARC/INFO and Arcview GISs as follows:
1
Chapter 1
GIS Data Models
INTRODUCTION
This is a book about GIS analysis tools. The hard part of GIS analysis is figuring out
which tools to use to solve your GIS problem. Once you have learned conceptually how
these tools work, then you can simply look up the correct syntax for using the tools in
your specific GIS. Lets take an example. Imagine that you have a line theme of streams,
and a polygon theme of ownership. You want to determine the ownership for each
stream segment. Conceptually, this is a simple GIS problem to solve . . .
Application of conceptual solution in ArcView GIS:
2 PRACTICAL GIS ANALYSIS
GIS DATA MODELS 3
POINT THEMES
A point is a GIS feature that has no length or area. It has a specific X,Y coordinate and
attribute information associated with that location point.
For example, imagine that you have a point theme of waterfowl nests as follows:
4 PRACTICAL GIS ANALYSIS
Since the GIS stores these points in an X,Y coordinate system, it knows the location
of each nest and the spatial relationships among nests. You can assign an identification
number and all information about each nest point in an attribute table.
Nests Point Attribute Table
To avoid confusion when determining spatial relationships, each point feature must
have a unique identification number. To ensure this, the GIS assigns a unique Nests# to
each point. The user could make a mistake and assign the same Nests-ID to two or more
nests. The internal ID Nests# assigned by the GIS solves this potential problem.
By definition, a point has no area or perimeter. Keep in mind that X,Y coordinates of
points always have some positional error. Therefore, many GIS tools that analyze points
use a fixed area, or a distance within a search radius. For example, you might create a
buffer area of 1 km around an owl nest to determine the number of snags within 1 km
of the nest. Or, in another application, you might ask for the distance from a well site to
the nearest fuel tank located within a search radius of 2 km.
GIS DATA MODELS 5
Since the GIS stores each arc as a series of X,Y vertices, it can easily estimate the
length of each stream arc. The GIS computes the length of each stream arc in the same
units as your GIS coordinate system. And since each stream arc has a unique stream#,
the GIS can determine spatial relationships among arcs. As a user, you can store infor-
mation about each stream in the arc attribute table. Information could be quantities
(stream pH), categories (stream class), character strings (stream name), and dates
(month/day/year).
Each arc is composed of a series of X, Y coordinates called vertices. The beginning
Streams Arc Attribute Table
You will learn about GIS tools for point analysis in Chapter 3.
LINE THEMES
A line or arc is a GIS feature that has length but no width. The following is an example
line theme of streams.
and ending vertex of each arc is called a node. The beginning node of an arc is called the
from-node and is typically identified in an attribute table as the FNODE#. The ending
node of an arc is the to-node and is typically identified in an attribute table as the
TNODE#. The FNODE# and TNODE# describe the direction of an arc. Once the GIS
knows the direction of an arc, it knows which is the left side and which is the right
side of the arc.
6 PRACTICAL GIS ANALYSIS
You will learn about GIS tools for analysis of line themes in Chapter 4.
NETWORK THEMES
A network is a special type of line theme consisting of connected arcs such as streets,
utility lines, or stream networks. With network analysis tools you can find the fastest,
cheapest, shortest, or best path to get from one location to another along a network.
For example, the optimal path to get from Point A to Point B in the following network
costs 75 + 10 + 10 +75.
Other network applications include:
Address geocoding to estimate the location of street addresses.
Finding the closest resource facilities to any event location.
Allocation of resources from supply centers to customers within the network.
You will learn about network analysis in Chapter 5.
DYNAMIC SEGMENTATION
Sometimes important line information is not available in X, Y coordinates, but instead is
recorded as measurements along lines such as mileage along a road, meters along a tran-
sect, etc. This type of information can be translated into a GIS by using a technique
called Dynamic Segmentation. The technique allows for segmentation of arcs into sec-
tions without changing the arc-node structure of a line theme.
You will learn about Dynamic Segmentation Tools in Chapter 6.
POLYGON THEMES
A polygon is a GIS feature that has an area and a perimeter. The following example
polygon theme represents four lake polygons.
GIS DATA MODELS 7
Lakes Polygon Attribute Table
A polygon attribute table has a special record for an artificial polygon called the uni-
verse polygon. The universe polygon has an area that is the sum of the area of all poly-
gons in the theme. It is always assigned a negative sign because it is an artificial polygon
that is used by the GIS in computations. You will learn about tools for analysis of poly-
gons in Chapter 7.
GRID THEMES
Grids are grid cells with a fixed number of rows and columns that have several tables as-
sociated with them. If the grid contains integer values, typically it is assumed that these
values represent categories and a value attribute table is associated with the grid as fol-
lows:
8 PRACTICAL GIS ANALYSIS
If the grid contains floating-point values, it is assumed that these values represent
quantities rather than categories and a statistics table is associated with the grid. For ex-
ample, imagine that we had a grid of elevation values as follows:
Grids that are common in GIS include digital elevation and land cover grids. You will
learn about tools for grid analysis in Chapter 8.
IMAGE THEMES
Images are special grids typically derived from some remote sensing device like a satellite
sensor, a digital camera, or a desktop scanner. Examples of images commonly used in re-
mote sensing include digital orthos, satellite imagery, and scanned maps.
GIS DATA MODELS 9
Example satellite image of Yukon River floodplain, Alaska.
You will learn more about image analysis tools in Chapter 9.
TOOLS FOR MANAGING GIS FEATURES
There are several generic tools that we will start with in this chapter that are applicable
to managing points, lines, polygons, grids, and images. They are as follows:
LISTList the contents of any GIS table.
COPYMakes a new copy theme from any point, line, polygon, or grid theme.
APPENDAppends 2 or more point, line, or polygon themes.
KILLDeletes a user-specified point, line, polygon, or grid theme.
RENAMERenames a user-specified point, line, polygon, or grid theme.
DESCRIBETells the user information about a point, line, polygon, grid, or image
theme.
Information produced by DESCRIBE operation:
10 PRACTICAL GIS ANALYSIS
TOOLS FOR BUILDING ATTRIBUTE TABLES
Sometimes you will obtain a GIS theme that does not have an attribute table associated
with it. The following generic tools can be used for creating tables associated with your
themes.
BUILD
Builds an attribute table for a point, line, or polygon theme.
As an example, imagine that you have the following theme.
At this point, the GIS knows that you have arcs#1, 2, and 3. You need to use BUILD
to tell the GIS to determine the polygons and their spatial relationships (called topol-
ogy). After building, you would have the following:
DESCRIBE Theme
BUILDVAT
Builds a value attribute table for an integer grid theme.
As an example, imagine that you have the following grid theme.
Grid
LIST Grid.VAT
file not found
BUILDVAT Grid
LIST Grid.VAT
BUILDSTA
Builds a statistics table for a grid theme.
As an example, imagine that you have the following grid theme.
Grid
LIST Grid.STA
file not found
BUIDSTA Grid
LIST Grid.VAT
GIS DATA MODELS 11
LIST Theme.PAT
EXERCISES
1) You have the following line theme:
Fill in the following information from using the DESCRIBE operation.
2) You have the following points. Point 100 is a well that is 200 feet deep, on parcel
356, while point 101 is a well that is 300 feet deep on parcel 387. Fill in the correct val-
ues for the point attribute table associated with this theme.
12 PRACTICAL GIS ANALYSIS
3) You have the following lines. Arc#1 has a length of 100 meters.
Fill in the correct values for the arc attribute table associated with this theme.
4) You have the following polygons. These parcels have the following tax informa-
tion:
Fill in the correct values for the polygon attribute table associated with this theme.
GIS DATA MODELS 13
5) You have the following grid of slope classes. Build a Value Attribute Table for your
grid.
6) You have the following grid of soil pH. Build a Statistics Table for your grid.
14 PRACTICAL GIS ANALYSIS
Chapter 2
GIS Tabular Analysis
INTRODUCTION
GIS Information about spatial features is typically stored in tables using a database man-
agement system. Typically the databases are stored as spreadsheets with each row or
record corresponding to one feature such as a point, line, or polygon. Each column in
the table corresponds to a feature attribute. The table columns are typically called fields
or items.
Each column in a table typically has the following characteristics:
1) Item Name. The item name is simply the name of the table column.
2) Item Type. The item types most commonly used are binary integer (B), floating
point (F), character (C), and date (D). Examples of binary integer items include
categorical attributes such as soil texture class, vegetation class, or road surface
type. Examples of floating point items include quantitative values such soil pH,
tree diameter, or road length. Examples of character items include names such
as soil order, plant genus/species, or street name.
3) Item Width. This refers to the number of bytes required to store each item. The
most basic storage unit for computers is a Bit (or Binary Digit). A bit has two
possible states, either a 0 or 1. Eight bits together make up a Byte.
With one byte, you could represent any integer ranging from 0 to 255:
For example:
What integer values would the following bytes represent?
The first byte would represent 2
7
+ 2
1
= 128 + 2 = 130. The second byte would rep-
resent 2
6
+ 2
4
+ 2
0
= 64 + 4 + 1 = 69.
15
The Character type uses a scheme of coding called the American Standard Code for
Information Interchange or ASCII to code each character using one byte per character.
The ASCII character set is as follows (0 through 31, 128 through 255 are special char-
acters such as bell, carriage return, line feed, escape, control, and so on).
What do the following three bytes represent as ASCII characters?
The first byte has a decimal equivalent of 64+4+2+1 = 71, which has an ASCII code
of G. The second byte has a decimal equivalent of 64+8+1 = 73, which has an ASCII
code of I. And the last byte has a decimal equivalent of 64+16+2+1 = 83, which has an
ASCII code of S.
Character items require one byte per character since the ASCII coding scheme is typi-
cally used to store characters. For example, if you want to store up to 10 characters in
an item called NAME, it would require an item width of 10 bytes.
Since date items are stored as YYYYMMDD, the item width for a date type is 8 bytes
(one byte for each value in the date).
You can store binary integers as either 2-byte or 4-byte integers.
What would be the range of possible values for a 2-byte (16-bit) integer?
16 PRACTICAL GIS ANALYSIS
The range of values for a 4-byte (32-bit) integer attribute would be +/- 2,147,483,647
Floating point items are stored using either 4- or 8-bytes. A 4-byte width is single-pre-
cision real (7 digits of precision), and 8 bytes is double precision (14 digits of precision).
4. Output Width. This refers to the number of number of characters used to dis-
play the item value. For example, you would need an output width of 13 to list
the attribute values of Quaking Aspen (a character attribute value), or
3.14159265359 (a floating point attribute value).
5. No. of Decimals. The number of digits to the right of the decimal place for a
floating point item.
As an example, imagine that we have a floating point item called PI (3.1415926536)
which is stored as follows. The value displayed will depend on how you originally de-
fined the item width, type, and output parameters:
Notice that in the second example, *** tells the user that the output format is not ap-
propriate for the attribute value. In this example, 3.145926536 cannot be displayed with
5 digits to the right of the decimal and an output width of only 3 characters. Notice also
that the fourth example stores PI with correct precision since the attribute has an item
width of 8 bytes. The third example has an item width of 4 bytes which allows for 7 sig-
nificant digits . . . thus the value displayed beyond 3.141592 . . . (beyond the 7
th
signif-
icant digit) is incorrect. In the last example, the value of 3.14 . . . is stored as a binary in-
teger and therefore all numeric information beyond the decimal point is lost.
SELECTING TABLE RECORDS
To select records, you must first select the table that contains the information you are in-
terested in. Once you have selected your table, you can specify records to select using the
following query tools and a logical expression. Logical expressions are questions com-
posed of components called operands, operators, and connectors.
Operands are items or values such as an item name (for examle NEST-ID), a numeric
value (for example, 3.14), or a character value (for example, Bald Eagle)
Operators allow you to ask questions regarding your operands.
For example, NEST-ID > 100
The following are common GIS expression operators:
GIS TABULAR ANALYSIS 17
Connectors are used to connect simple logical expressions to compound logical ex-
pressions. For example: soil_type in {1,5,7} and texture cn Silt Loam.
The following are common GIS expression connectors:
The following query commands are typically available to build your record selection
with.
RESELECTallows you to reduce your selected set of records by issuing selection cri-
teria using a logical expression.
ASELECTallows you to add records to your selected set of records.
NSELECTreplaces the currently selected records with those not selected.
As an example, imagine you have a table of soil attributes as follows:
RESELECT DRAINAGE GT 1 AND DEPTH IN {15,25,99} would select the following
records:
18 PRACTICAL GIS ANALYSIS
NSELECT
ASELECT TEXTURE CN Silt would first select the records not currently selected, and
then add to the selection set all the records that contain Silt as a texture attribute
. . . thus the following records would be selected. Notice that Silty Loam is included
since the characters Silt are in that attribute.
GIS TABULAR ANALYSIS 19
DESCRIPTIVE STATISTICS
STATISTICSComputes descriptive statistics of user-specified items from selected
records. The descriptive statistics include frequency, mean, sum, maximum, minimum,
standard deviation.
Imagine that you have the following arc attribute table of hiking trails:
What values would you get for the following STATISTICS queries?
ASELECT
STATISTICS WILD_CLASS
SUM LENGTH
First all records are selected, then statistics are to be summarized by the Wild_class at-
tribute. Finally the sum of the attribute Length is requested. The following statistics are
returned:
ASELECT
RESELECT LENGTH LT 5 AND DIFFICULTY LE 2
STATISTICS WILD_CLASS
MIN LENGTH
First all records are selected, then records that have length less than 5 and difficulty
less than or equal to 2 are selected (5 records). Then these 5 records are summarized by
Wild class and the minimum length is computed for each Wild class value:
SUMMARIZING TABLES
FREQUENCYThis GIS tool produces a list of the unique attribute values and their
frequency for your selected records.
The FREQUENCY program asks you for two parameters:
1) Which item(s) do you want to be analyzed in terms of unique attribute values? (the
FREQUENCY ITEM).
2) Which item(s) do you want totals of in your output frequency table? (the SUM-
MARY ITEM).
As a simple example, the following table is generated after requesting Wild_class as the
frequency item and Length as the summary item:
20 PRACTICAL GIS ANALYSIS
Notice that the table contains the same information we got by using the STATISTICS
command.
FREQUENCY is different than STATISTICS, in that FREQUENCY can summarize
by many combinations of attributes while STATISTICS can summarize only by a single
attribute.
For example, in the next example we summarize by both Difficulty and Wild_Class:
OTHER COMMONLY USED TABULAR TOOLS
VIEWING TABLES
DIRLists the tables available in your current workspace.
ITEMSLists the item definitions (name, type, input/output width, etc.) for your table.
LISTLists the information contained in your table.
MANAGING TABLES
Modifying tables:
The following are commonly used to modify table columns and rows:
ALTERUsed to alter the item characteristics such as item name or output width.
CALCULATEAssigns new values to an item in all selected records, using an arithmetic
expression or string. For example, CALCULATE HECTARES = ACRES / 2.471
REDEFINEUsed to create new items that share column space with existing items. One
example would be to redefine a new item called AREA_CODE, from an existing item
called PHONE_NUMBER.
SORTAllows you to sort selected records by specified table item(s).
UPDATEallows you to interactively type in new values for selected record items.
Adding items, records, and tables:
ADDAllows you to add records or rows to your table.
ADDITEMis used to add new items or columns to your table.
Deleting items, records, and tables:
DROPITEMDeletes any specified items from your table.
PURGEDeletes the selected records from your table.
KILLDeletes any user-specified tables.
Exporting tables:
COPYCopies an existing table to a new table.
UNLOADWrites selected table information to ASCII text file.
SAVEWrites selected table information to binary INFO file.
MERGING TABLES
Tables can be merged if there is a key item or field that is in common with the tables.
GIS TABULAR ANALYSIS 21
FREQUENCY ITEMS: DIFFICULTY,WILD_CLASS
SUMMARY ITEMS: LENGTH
JOINITEMPermanently merges two tables.
RELATE Temporarily merges two or more tables.
Imagine that you have a huge soils database of 20,000 soil polygons with the following
polygon attribute table. You could use the RELATE command to temporarily link the
soils polygons to the soil texture look-up table:
The advantage of linking to a look-up table instead of storing TNAME values as
polygon attributes is that it is much easier to maintain a small look-up table instead of
20,000 polygon attribute records. For example, if we choose to store the texture name in
the polygon attribute table, incorrect attribute values such as Silt Loam, Silty Clay
Loam, Sandy Silt would be more difficult to find and correct compared to the linked
look-up table approach.
INDEXING ATTRIBUTES
Imagine that you just picked a gallon of blueberries and wanted to find all the blueberry
recipes in a cookbook. You could start on page one and search the entire cookbook,
page by page in a sequential manner. However, it would be much more efficient to look
for the attribute value Blueberry in your cookbook index. In a similar manner, items
can be much more efficiently searched if they have been indexed . . .
INDEXITEMCreates an attribute index to increase query speed for that item.
TABULAR ANALYSIS EXERCISES
1) You have an attribute table about soil polygons. You run the statistics program to cre-
ate a new table summarizing the area of soil polygons by texture class. The output table
is as follows:
22 PRACTICAL GIS ANALYSIS
Why is the Frequency for texture class 3 not displayed in the table?
How can you solve the problem so that Frequency for texture class 3 is displayed in
the table?
Which records would be selected in the following expression:
Tables: SELECT stream.aat
Tables: RESELECT Ownership in {1,3} AND Trout_count GT 200
Which records would be selected in the following expression:
Tables: SELECT stream.aat
Tables: RESELECT Ownership in {1,3} OR Trout_count GT 200
3) You have a street arc attribute table containing two attributes: Speed_Limit which
is the maximum allowable speed in miles per hour and Length which is the length of
each arc in meters. You add another attribute column called Time . How would you cal-
culate the time in minutes it would take to travel across each arc at the maximum speed
limit? There are 5280 feet in a mile and 3.281 feet in a meter. You start by adding two
new columns: FT_PER_MIN , the speed limit expressed in feet per minute and
Length_FT , the arc length in feet.
CALCULATE FT_PER_MIN = _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _
CALCULATE Length_FT = _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _
CALCULATE TIME = _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _
4) Correct the following logical expression:
RESELECT SPECIES CN KING SALMON OR CN SOCKEYE SALMON
5) Correct the following calculation:
CALCULATE ACRES = HECTARES X 2.471
6) Correct the following calculation:
CALCULATE ACRES = AREA / 43,560
7) Correct the following logical expressions:
RESELECT VEGCODE = 1
CALCULATE SHADECOLOR = 27
RESELECT VEGCODE = 2
CALCULATE SHADECOLOR = 35
RESELECT VEGCODE = 3
CALCULATE SHADECOLOR = 67
8) You have selected a FOREST.PAT polygon attribute table. You create a new at-
tribute called SITE_CLASS based on an existing SITE_INDEX attribute. SITE_CLASS
of 1 would be any polygon with SITE_INDEX less than 50, SITE_CLASS of 2 would be
any polygon with SITE_INDEX between 50 and 75, and SITE_CLASS of 3 would be
any polygon with a SITE_INDEX greater than 75.
GIS TABULAR ANALYSIS 23
2) You have the following arc attribute table:
Fill in the appropriate TABLES commands to do the following:
_____________________ /***Add a new attribute column called Site_class
_____________________ /****Select the forest polygon attribute table
_____________________ /***Select all records with site_index less than 50
_____________________ /***Fill in the Site_class attribute with a value of 1
_____________________ /***Select all records in the table
_____________________ /***Select all records with site_index between 50 and 75
_____________________ /***Fill in the Site_class attribute with a value of 2
_____________________ /***Select all records in the table
_____________________ /***Select all records with site_index greater than 75
_____________________ /***Fill in the Site_class attribute with a value of 3
_____________________ /***Select all records in the table
_____________________ /***Select the universe polygon record
_____________________ /***Fill in the Site_class attribute with a value of 0
9) You have a point attribute table of waterfowl nests containing the following items:
You want to produce a table with the following information:
What would you use for frequency and summary items to generate this information?
Frequency item(s) _______________________________ Summary item(s) _____________________________
10) There is a proposal to purchase some land for an experimental forest research
site. Your job is to produce a table listing the hectares and percent of area for each veg-
etation class in this area. You have a vegetation polygon attribute table and another
table of vegetation names as follows:
24 PRACTICAL GIS ANALYSIS
Fill in the appropriate tools in the following flowchart that would produce a sorted
table of total hectares and percent of area for each vegetation class.
GIS TABULAR ANALYSIS 25
Chapter 3
Point Analysis
INTRODUCTION
A point is a GIS X,Y location that has no area or perimeter. Example point themes in-
clude point locations from research plots, radio-collared animals, wells, individual trees,
fire hydrants, or utility poles.
The following GIS tools can be used in managing points:
ADDXYAdd each point X,Y coordinate as new items in the point attribute table.
NODEPOINTCreate a new point theme from nodes in a line or polygon theme.
ARCPOINTCreate a new point theme from arc vertices or polygon labels in a
line or polygon theme.
POINTNODETransfer attributes from points to the nearest node in a line or
polygon theme.
Points can be analyzed using the following GIS tools:
AREA ANALYSIS TOOLS:
THIESSENCreate polygons of proximity from points.
BUFFERGenerate buffer areas of a user-specified distance around points.
DISTANCE ANALYSIS TOOLS:
NEARCompute the nearest distance from points to features in a second point,
line, or polygon theme.
POINTDISTANCECompute all point distances between two point themes.
ATTRIBUTE ANALYSIS TOOLS:
RESELECTCreate a new point theme by selecting points using a logical expres-
sion.
INTERSECTTransfer polygon attributes to a point theme.
27
AREA ANALYSIS TOOLS
THIESSEN
Converts a point theme into a theme of proximal (or Thiessen) polygons
Thiessen polygons can be used to apportion a point theme into polygons known as
Thiessen or Voronoi polygons. Each polygon will represent only one points region. Each
polygon has the unique property that any location within that polygon is closer to the
polygons source point than to any other points. Lets do the following example.
Step 1) Connect the points 1 to 2, 2 to 3, 3 to 4, and 4 to 1 with dashed lines.
Step 2) Draw the midpoint of each line.
Step 3) Draw a solid line at each midpoint perpendicular to your dashed lines.
Each solid line represents the distance halfway between two points. Draw
these lines until they merge to create polygons. Erase your dashed lines and
points. Build your polygon attribute table.
Input Point Theme
28 PRACTICAL GIS ANALYSIS
Output Polygon Theme
BUFFER
Creates a buffer polygon of user-specified distance around each point.
You have two basic options with buffering points.
1) You can buffer all points in your theme by a fixed distance.
2) Or you can use a look-up table to buffer all points in your theme by a variable dis-
tance, depending upon each points attribute value.
Imagine that you have a look-up table called ZONE.LUT and a point attribute table
called WELLS.PAT as follows:
POINT ANALYSIS 29
The following theme results from using TYPE and ZONE.LUT with BUFFER to cre-
ate variable buffers:
Notice that the output table has a special attribute to indicate whether each polygon
is inside a buffer (100) or outside a buffer (1). Also notice that the area of each buffer
polygon is not equal to PI * radius
2
. For example, for wells 1 and 3, the buffer radius is
10. PI * 10
2 =
314.15926. The GIS has to use many straight arcs to approximate a
buffer circle, and thus the area will always be less than a true circular buffer.
Buf_Well Polygon Attribute Table
Imagine that you change your look-up table as follows
The following theme results from using TYPE and ZONE.LUT with BUFFER to cre-
ate variable buffers:
Notice that if your buffer sizes are relatively large, the number of resulting buffer
polygons can be less than the number of original points.
DISTANCE ANALYSIS TOOLS
NEAR
Computes the distance from each point in a point theme to the nearest feature in a
point, line, or polygon theme.
For a line or polygon theme, you can specify whether NEAR looks for only the nearest
arc or polygon node (NODE option)or whether nodes and vertices (LINE option) are
both searched for. The user also has the option of requesting that the X,Y coordinates of
each nearest point be stored in the point attribute table.
Assume the two open dots are points in a point theme. Imagine you ran NEAR with
the NODE option using the point theme and the following line theme. The arrows indi-
cate the nearest points selected by the NEAR program:
30 PRACTICAL GIS ANALYSIS
Buf_Wells2 Polygon Attribute Table
Imagine you ran NEAR with the LINE option using the point theme and the follow-
ing line theme; the arrows represent the nearest points selected by the NEAR program:
POINT ANALYSIS 31
Note: There are two flaws in the NEAR program:
Flaw#1: Imagine you assigned a search radius of 100 meters in the following example.
Moose#11 is farther away from any willow polygon than your search radius, and it is there-
fore assigned a distance of zero.
POINTDISTANCE
Creates a new table that contains the distances between the points in one
theme to ALL points in a second theme that are within the specified search
radius.
A distance of zero means that either the two points being compared have exactly the
same X,Y coordinates, or the distance is greater than your search radius.
You have two point themes : juvenile mallards and adult mallards:
= Adults

= Juveniles
What would the distance between Adult# 43 and Juvenile#13 be output if you run
POINTDISTANCE with a search radius of 200? Assume each square represents a dis-
tance of 10 units.
32 PRACTICAL GIS ANALYSIS
Flaw#2: Imagine you assigned a search radius of 1000 meters in the following example.
The flaw here is that the GIS computes the distance to the nearest arc, even though the
moose is inside a willow polygon.
In the prior example, Adult# 43 has X,Y coordinates of 70,20 and Juvenile#13 has
X,Y coordinates of 110,40. We can use the Pythagorean theorem to compute the
straight-line distance between these two locations:
Since the GIS knows the X,Y coordinates of each point, it can easily compute A and
B portions of the right triangle: A= 110 70 = 40 and B= 40 20 = 20
C = SQRT (A
2
+ B
2
) = SQRT (40
2
+ 20
2
) = 44.7
POINT ANALYSIS 33
ATTRIBUTE ANALYSIS TOOLS
RESELECT
Creates a new point theme by selecting points from existing point theme using user-
specified logical expressions.
Reselect can be used for point, line, and polygon features to select the features you want
by asking one or more logical expressions. The following examples use this point theme:
34 PRACTICAL GIS ANALYSIS
INTERSECT
For points that intersect with polygons, produce an output point theme containing
attributes from both the input point and polygon theme.
Think of INTERSECT as transferring polygon information to points if the points lie in-
side the polygons.
SOILS theme
POINT ANALYSIS 35
SLOPE_CLASS theme
The output theme point attribute table resulting from INTERSECT SOILS and
SLOPE_CLASS is as follows:
POINT ANALYSIS EXERCISES
1) You have a point theme of manhole covers and a line theme of streets. You want to
determine the street name for each manhole cover.
Fill in the following flowchart with the appropriate GIS tools to solve your problem:
36 PRACTICAL GIS ANALYSIS
2) You have a point theme of wells to sample groundwater. You want to calculate the
mean distance for each well to all other wells. Fill in the following flowchart to solve
your problem:
3) You have a point theme of lightning strike locations and a polygon theme of eleva-
tion classes. Your goal is to produce a table like the following:
Fill in the following flowchart with the appropriate GIS tools to solve your problem:
POINT ANALYSIS 37
4) You have five GPS survey monuments set up in the Tanana Flats area. Draw five
Thiessen polygons associated with these GPS monuments:
38 PRACTICAL GIS ANALYSIS
5) You have a point theme of bufflehead nest box locations along the Chena River.
Outline the GIS analysis tools that you would use to select all nest box points that are at
least 1 km away from any other nest box.
6) You have a point theme in meters of snowshoe hare locations. You want to know
how many of these locations are within 50 meters of a willow polygon . . . .
POINT ANALYSIS 39
How many of the ten hare locations are within 50 meters of a willow polygon?
7) You have a point coverage of radio-collared moose. The point attributes include
Moose-ID, sex, and age. You also have a line coverage of rivers. You want to produce a
table as follows:
Fill in the following flowchart to produce the above table:
8) You have a point theme of Polar Bear locations stored in longitude/latitude. You
want to create a new theme of bear locations north of 70 degrees and between 170 and
100 degrees of latitude. Fill in the following flowchart to produce this new theme.
40 PRACTICAL GIS ANALYSIS
9) You have a theme of endangered plant species locations and a theme of randomly
located points. You want to know what percent of plant points are within clay loam soil
polygons compared to the percent of random points in clay loam polygons. Fill the fol-
lowing flowchart with the appropriate GIS tools to solve this problem
POINT ANALYSIS 41
10) You have a theme of endangered plant species locations. You want the mean,
minimum, and maximum distance of the plant locations to the nearest clay_loam poly-
gon. Fill the following flowchart with the appropriate GIS tools to solve this problem.
42 PRACTICAL GIS ANALYSIS
Chapter 4
Line Analysis
INTRODUCTION
A line is a GIS feature that has length but no width. Typical line themes include arcs
from roads, streams, hiking trails, pipelines, or contours.
The following GIS tools are commonly used in analyzing lines:
DISSOLVEMerges adjacent arcs (by dissolving the node), if they have the same
value for a user-specified arc attribute.
COUNTVERTICESAdd an item in the Arc Attribute Table called VERTICES
and fill the item with the total number of vertices (including beginning and ending
node) for each arc.
BUFFERGenerate buffer polygons of a user-specified distance around arcs.
RESELECTCreate a new line theme by selecting arcs using a logical expression.
INTERSECTTransfer the polygon attributes to a line theme.
DISSOLVE
Merges adjacent arcs if they have the same attribute value.
The new Cover-ID of the merged arc will be the lowest Cover-ID of the combined arcs.
An #ALL option tells the GIS to merge the adjacent arcs, only if all attributes beyond the
Arc-ID are exactly the same.
The following line theme is dissolved based on the item TYPE:
43
The #ALL option tells DISSOLVE to merge the adjacent arcs only if all attributes be-
yond the Arc-ID are exactly the same.
44 PRACTICAL GIS ANALYSIS
COUNTVERTICES
Counts the number of points in each arc.
The item called VERTICES is added to the arc attribute table as the total number of
points (nodes and vertices) for each arc.
LINE ANALYSIS 45
BUFFER
Create a buffer polygon around each arc.
When buffering a line theme, you can specify the arc ends to be buffered using a square
(FLAT) or semicircle (ROUND) shape. You can also specify whether the buffer should
be on both sides of each arc (FULL) , on the topological left side (LEFT) , or on the
topological right side (RIGHT) . Like buffering points, you can specify a fixed buffer dis-
tance, or by using a look-up table, variable buffers.
Here is an example theme called ROADS and an associated look-up table:
Imagine that you request the following buffers:
The first step is to figure out the left and right side of each arc based on the location
of the from and to nodes:
46 PRACTICAL GIS ANALYSIS
The following polygon themes would be output after buffering:
LINE ANALYSIS 47
RESELECT
Creates a new line theme by selecting arcs based on user-defined logical expres-
sion(s).
In the following example, we use RESELECT to create an output theme of short, clear-
water streams.
INTERSECT
Creates a new line theme by intersecting arcs with a polygon theme and then
adding the polygon attributes to each arc.
Imagine the following example:
48 PRACTICAL GIS ANALYSIS
The following theme results from intersecting these two themes:
LINE ANALYSIS 49
Note that you can not intersect a point and line theme . . . you must intersect either
points or lines with a polygon theme. That is because by definition, both points and lines
have no area and therefore point in line or line in point intersection is impossible while
point in polygon or line in polygon intersection is possible.
50 PRACTICAL GIS ANALYSIS
LINE ANALYSIS EXERCISES
1) You have the following line theme of pipes. You run DISSOLVE using Pipe_Class
as the dissolve item. What will your output theme look like?
2) You have a line theme of roads and a point theme of cabin locations. You want to
find all cabins that are within 1 mile of a road. Your GIS coordinate system is in meters.
Fill in the following flowchart to solve the problem:
LINE ANALYSIS 51
3) You have a line theme of streams with an arc attribute called King_count repre-
senting the count of king salmon observed along each arc. For each stream, determine
the total density of king salmon per mile. Your GIS coordinate system is in meters.
52 PRACTICAL GIS ANALYSIS
4) You have a line theme where each line represents the border between two vegeta-
tion types as follows:
From searching the polygon attribute table, you know there is one stand of Sweet
Gum and one stand of Pin Oak in the theme. What is the length of the border between
the Sweet Gum and Pin Oak stand? Fill in the appropriate TABLES tools to solve the
problem:
LINE ANALYSIS 53
5) You have a line theme of roads and a point theme of auto accidents. Each arc has
a highway code attribute. You want to determine all the accidents that occurred on the
Parks Highway (highway code= 2) as a new point theme.
Fill in the following flowchart to solve your GIS problem:
6) You have a polygon theme of parcels and a line theme representing a utility right of
way. Each parcel polygon has the owners phone number and address stored in the poly-
gon attribute table. You want to generate a text file of owner phone numbers and ad-
dresses for all parcels within the right of way.
Fill in the following flowchart to solve your GIS problem:
54 PRACTICAL GIS ANALYSIS
8) You have a line theme of contour lines with an attribute of elevation. Elevations
are recorded at contour intervals of 10 meters, ranging from 10 to 980 meters. You want
to assign a dashed line to minor contour elevations that are tens of meters (for example:
10,20,30,40, etc. ) and a solid line to major contour elevations that are hundreds of me-
ters (for example 100,200,300).
Fill in the following flowchart to solve your GIS problem:
LINE ANALYSIS 55
7) You have a line theme of gas pipelines and a polygon theme of ownership. You
want to produce a table showing for each ownership, the total length of pipeline by pipe
diameter class and by pipe age class.
Fill in the following flowchart to solve your GIS problem:
9) You have a line theme of streets and a point theme of fire hydrant locations. You
want to find all areas where the distance between fire hydrants is greater than 1 km and
the hydrants are within 10 meters of a street.
Fill in the following flowchart to solve the problem:
56 PRACTICAL GIS ANALYSIS
10) A grizzly bear biologist wants you to use a line theme of streams and a point
theme of bears to produce the following three tables. The GIS coordinate system is in
meters. Your job is to figure out the total area within 1 km of each stream and the total
number of bears within these 1 km areas
Fill in the appropriate tools to solve the problem on the next page:
LINE ANALYSIS 57
Chapter 5
Network Analysis
INTRODUCTION
Network analysis is a special type of line analysis involving a set of interconnected lines.
Typical networks include themes such as roads, streams, hiking trails, and pipelines.
Network Analysis can be used to answer at least four types of questions:
1) Address Geocoding. Address geocoding is the process of taking addresses and esti-
mating their locations in your GIS coordinate system. This is done by relating an
address database to your GIS theme. Example applications include displaying a
house address on a GIS street view for delivery of some product, generating driv-
ing directions to a given address, or displaying customer locations in a GIS view
from a list of customer addresses.
2) Optimal Routing. Optimal routing is the process of delineating the best route to
get from one location to one or more locations. The best route could be the
shortest, the quickest, or the most esthetic, depending on the GIS users preference
for defining best. Example applications include determining the quickest way to
get from a fire station to a fire location, determining the shortest route to divert
water lines given a closed junction, or determining the most economic delivery
route to several stops.
3) Finding Closest Facilities. This is a special type of optimal routing problem where
you are trying to find the closest points to a given location. Typically the points
are called facilities and the given location is called an event location. Example ap-
plications include determining which two fire stations would have the best re-
sponse time to a reported fire event, determining the best ambulance station to re-
spond to an accident report event, or finding ten houses for sale that are closest to
a day care center.
4) Resource Allocation. Resource allocation is the allocation of resources from sup-
ply centers to customers on a network. Example applications include a sand and
gravel center supplying road sanding after a snowstorm, a school supplying seats
to students, a well supplying water to an irrigation network, etc. In resource allo-
cation, typically resources are allocated across the network until either resources
are exhausted, or demand across the network is satisfied.
59
ADDRESS GEOCODING
Address geocoding requires a table of addresses and theme that contains attributes that
can be used to match to the table of addresses. Address geocoding most often involves a
line theme like streets, but can also be used with point (e.g., light pole IDs), or polygon
(e.g., parcels) themes.
Building addresses are most commonly used with street themes in address geocoding.
Typical street attributes might include prefix (e.g., East, North, Old), street_name (e.g.,
5
th
, Maple, Broadway), type (e.g., street, avenue, way, boulevard ) and suffix (e.g., NW).
Each street arc typically contains address information such as the beginning and end-
ing addresses on the left and right side of the street. How does the GIS know which side
of the street is right and which side is left? Each arc has a beginning node and an ending
node. Topological left and right sides are inferred from these nodes. For example:
The to- and from- nodes allow the GIS to delineate the following right and left sides
of the streets:
The GIS typically would have address information for the beginning and ending of
each arc. For example,
60 PRACTICAL GIS ANALYSIS
Imagine that you got a report of a broken gas line in a building at 2906 East Main
Street. The GIS can estimate the location, first by determining which arc contains the
Prefix/Name/Type/Suffix. Once the GIS determines that this is Arc 102, the 2906 address
can be estimated using linear interpolation. Since it is an even-number address, it is on
the right side of the arc. The address ranges from 3200 to 1800, (a range of 1400) along
the right side of the arc. The length of the arc is 2200 meters. The address of 2906 can
be geocoded as
In the previous example, the address perfectly matched the address attributes of an
arc. What would happen if the address did not perfectly match? For example, it might be
reported as 2906 E. Main Street, or 2906 East Main, or 2906 E. Main St. Most GIS
allow the user to specify parameters such as spelling sensitivitiy and minimum match
score. A simple example might be, start with a perfect score of 100 and then deduct 25
points for address Prefix/Name/Type/Suffix that does not match any street address at-
tributes. In the above example, 2906 E. Main Street would score 75, 2906 East Main
would score 75, and 2906 E. Main St. would score 50. In real GIS applications, the
matching score rules are much more sophisticated. The GIS user can specify the mini-
mum acceptable matching score. The higher the minimum matching score, the higher the
risk of failure to match all addresses entered for geocoding. The lower the minimum
matching score, the higher the risk for some incorrect address geocoding.
OPTIMAL ROUTING
Optimal routing is the process of delineating the best route (or path) to get from one lo-
cation to one or more locations. It is not usually feasible to test all possible paths that
exist in a network. Instead, a pathfinding algorithm is used. Perhaps the most commonly
used pathfinding algorithm is the Dijkstra algorithm, first published by E. W. Dijkstra.
This algorithm is explained below with a simple example.
Imagine that we have the following network of hiking trails. We want to find the
quickest path to get from node 1 (trailhead) to node 7 (tent site). The time to hike each
trail varies depending upon trail conditions. Trails 3,5,and 6 are relatively quick hikes.
NETWORK ANALYSIS 61
We build two tables, one of nodes that have already been processed, and one of adja-
cent nodes to process. We start at the trailhead, node#1:
62 PRACTICAL GIS ANALYSIS
We then pick the adjacent node with the least cumulative cost (hiking to node#4 will
take 10 minutes), and add that node to the processed nodes list.
. . . and scan the nodes adjacent to your latest processed node (4) . . . and add them to
the Adjacent Nodes list. There are 3 nodes adjacent to node#4: 1, 3, and 7. Node#1 is
already in the processed nodes list, so it is excluded from the adjacent nodes list. The
cumulative cost for node#3 is the cost of hiking to node#4 (10 minutes), plus the cost of
hiking from node#4 to node#3 (10 minutes). The cumulative cost to get to node#7 is 10
minutes + 30 minutes.
We then pick the adjacent node with the least cumulative cost (node#2 with a cost of
15), and add it to the processed nodes list.
We then scan the nodes adjacent to your latest processed node (2) . . . and add them
to the Adjacent Nodes list. The nodes adjacent to node#2 are nodes 1,3, and 5. Node#1
is already in the processed nodes list, so it is excluded from the adjacent nodes list. The
cumulative cost for node 3 is the cost to hike to node#2 (15 minutes) plus the cost to
hike from node#2 to node#3 (30 minutes). The cumulative cost to get to node#5 is the
time to hike to node#2 (15 minutes) plus the time to hike from node#2 to node#5 (10
minutes).
We then pick the adjacent node with the least cumulative cost, and add it to the
processed nodes list.
Next, scan the nodes adjacent to your latest processed node (3) . . . and add them to
the Adjacent Nodes list. The nodes adjacent to node#3 include 1,2,4,5,6,and 7. Nodes
NETWORK ANALYSIS 63
1,2,3, and 4 are already in the processed nodes list and are therefore excluded from the
adjacent nodes list. The cumulative cost to hike to node#3 is 20 minutes. To go all the
way to node#5 will take 20 + 40 = 60 minutes. To go to node# 3 and then node#6 will
take 20 + 30 = 50 minutes. And to hike to node#3 and then node#7 will take 20 + 20 =
40 minutes.
We then pick the adjacent node with the least cumulative cost, and add it to the
processed nodes list.
Scan the nodes adjacent to your latest processed node (5) . . . and add them to the Ad-
jacent Nodes list. The nodes adjacent to node#5 include 2,3, and 6, Nodes 2 and 3 are
already in the processed nodes list and therefore are excluded from the adjacent nodes
list. The cumulative cost to hike to node 6 is 25 minutes to hike to node#5 and then 5
minutes to hike to node#6, totaling 30 minutes.
We then pick the adjacent node with the least cumulative cost, and add it to the
processed nodes list.
64 PRACTICAL GIS ANALYSIS
Scan the nodes adjacent to your latest processed node (6) . . . and add them to the Ad-
jacent Nodes list. The nodes adjacent to node 6 include nodes 3,5, and 7. Nodes 3 and 5
are already in the processed nodes list and are therefore excluded from the adjacent
nodes list. The cumulative cost to hike to node#7 through node#6 is 5 + 30 = 35
minutes.
We then pick the adjacent node with the least cumulative cost and add that to the
processed nodes list.
All seven nodes are now in the processed nodes list, therefore the processing stops.
The quickest route to get to node#7 will take 35 minutes and from the processed nodes
list, it is
Node#7 Node#6 Node#5 Node#2 Node#1
We can ask the GIS to display this optimal route in our hiking trail network view.
NETWORK ANALYSIS 65
We could also ask the GIS to give us directions of the optimal route:
66 PRACTICAL GIS ANALYSIS
Once the nodes have been processed, all the optimal routes to node#1 from any node
are solved. However, there may also be more than one optimal route that is not reflected
in the analysis. For example, from the processed nodes list, the optimal route to hike
from node#3 is Node#3 Node#4 Node#1 which would take 20 minutes. The di-
rect Node#1 Node#3 route would also take 20 minutes.
In this simple example, there are relatively few possible routes, so as a check you
could try all possible routes. This is generally not possible with complex networks typi-
cal of real-life applications.
Possible Routes from Trailhead to Tent Site
FINDING CLOSEST FACILITIES
This is a special type of optimal routing problem where you are trying to find the closest
points to a given location. Typically the points are called facilities and the given location
is called an event location. Example applications include determining which two fire
stations would have the best response time to a reported fire event, determining the best
ambulance station to respond to an accident report event, or finding ten houses for sale
that are closest to a day care center.
As example, imagine that a fire is reported 1610 East Willow Street and we want to
determine the 2 fire stations that would have the best response time to that event loca-
tion. In this simple example, we have a streets theme with slow streets that take 3
minutes to travel, and faster streets that take 1 minute to travel.
When a fire address is reported, it is geocoded to a street location and then the GIS
can compute the optimal path from each fire station to the fire location.
NETWORK ANALYSIS 67
Typically the location of facilities are fixed and exist as a GIS point theme. Event lo-
cations can be entered into the network in several ways. In dispatching, typically some-
one calls in an address of the event and it is located by address geocoding. In other ap-
plications, the event location may be an existing GIS point theme. For example, a realtor
may have a point theme of day care centers, and a point theme of houses currently for
sale. She then selects one day care center from a neighborhood a client is interested in,
and then runs network analysis to determine the ten closest houses that are for sale.
Event locations can also be entered by graphically creating an event point theme in the
GIS view.
68 PRACTICAL GIS ANALYSIS
RESOURCE ALLOCATION
Resource allocation or service area delineation is the allocation of some resource from
supply centers to users in the network. Examples include a sand and gravel center sup-
plying plowing and sanding of roads after a snowstorm, a school supplying seats to stu-
dents, a well supplying water to an irrigation network, etc. In resource allocation, typi-
cally resources are allocated across the network until either resources are exhausted, or
demand across the network is satisfied.
Imagine that a new recycling center is being proposed. Your job is to analyze two al-
ternative locations. What is the area within a 1 km drive of each proposed location?
NETWORK ANALYSIS 69
The first step is to start at each service center and go out along the network until the
limit of 1km is met. This is the service network of each proposed recycle center.
70 PRACTICAL GIS ANALYSIS
In some applications, you might want a polygon theme of service areas corresponding
to your service networks. An example might be where you want to select all houses from
a point theme that are within each service area, so that you can generate mailing labels
from each house address. Service areas or service polygons are delineated by connecting
the outside nodes of each service network as follows:
In some applications, the resource centers may have a supply constraint. For example,
schools can only hold a maximum number of students, a reservoir can only hold some
maximum quantity of water, and so on. In our simple recycling center example, imagine
that each center has the maximum capacity of processing 6,000 units per week. And as-
sume that each household generates 10 units of recyclable waste per week. If our streets
theme has an attribute containing the number of households within each arc, we could
model this capacity constraint. In this simple example, we will assume 50 households
per 250 m of arc in neighborhood A, and 10 households per 250 m of arc in neighbor-
hood B. We could allocate along the network up to 1 km from each center, until the cen-
ter capacity has been reached.
NETWORK ANALYSIS 71
Neighborhood A is more densely populated and the center reaches capacity at 600
households (or 600 X 10 = 6,000 units per week). Neighborhood B is less densely pop-
ulated and reaches the 1 km distance limit while 160 households supply 1,600 units per
week and therefore the capacity is not met for allocation of households to this facility.
TURN TABLES IN NETWORKS
So far we have dealt with relatively simple networks. Complexities such as stop signs, U-
turns, one-way streets, red lights, and so on, can easily be incorporated in networks. One
way of incorporating such complexities is by using turn tables.
A turn table can be used to assign impedance values at nodes. Left turns are assigned
positive angles, while right turns are assigned negative angles. For example, a stop sign
or red light could be modeled as follows:
72 PRACTICAL GIS ANALYSIS
In the above example, to go straight from arc 66 to arc 67 would take 30 seconds. A
left turn from arc 66 to 68 would take 60 seconds while a right turn would take 20 sec-
onds. A U-turn on arc 66 is not allowed and therefore is assigned a negative impedance
value.
If a turn is not represented in a turntable, it will simply have no travel cost associated
with it. Turntables can also be used to model impossible turns such as a freeway over-
pass at nodes 101 and 102.
NETWORK ANALYSIS EXERCISES
1) You have a report of a fire at 1721 First Avenue West. Draw the location of that
address on the following streets theme.
NETWORK ANALYSIS 73
74 PRACTICAL GIS ANALYSIS
2) Given the following network theme, determine the path that is the quickest to get
from point A to point B.
NETWORK ANALYSIS 75
3) You are managing wilderness cabins on a network of backcountry ski trails. The
average skiing time per 100 meters of trail for various slopes is as follows:
76 PRACTICAL GIS ANALYSIS
Delineate service network that represents all trails that are within a half hour from
the Colorado Creek Cabin.
Chapter 6
Dynamic Segmentation
INTRODUCTION
Dynamic Segmentation is a special type of line analysis. Arcs are segmented dynamically
and temporarily for the analysis. Dynamic segmentation is especially useful in situations
when arcs are used for many different applications. For example, imagine that you have
an arc from a streets theme:
The street maintenance department wants to segment the street based on pavement
type.
The police department wants to segment the street based on speed limits.
And the fire department wants to segment the street based on hydrant locations.
You can see that the arc becomes quite segmented and complex with these three ap-
plications . . .
An alternative is to maintain the original streets theme and use dynamic segmentation
for each application. Dynamic segmentation has another advantage: it does not require
X,Y coordinates of events that are used to segment an arc. You can use events such as
accidents with distance from an intersection, interpretive stops with highway mile mark-
ers, and so on. This type of linear data is called route-measure formatted data.
77
ROUTES AND SECTIONS
A route is a collection of arc segments or sections which have attributes. A route does
not necessarily have to start or end at a node. The following might be an example of a
bus route.
78 PRACTICAL GIS ANALYSIS
BUS ROUTE
BUS ROUTE #1
BUS ROUTE #2
BUS ROUTE #3
And a collection of routes is called a route system. The following is an example of a
route system composed of three bus routes.
Each route is composed of one or more sections. For example, assume you have a
road route 19 that has three sections: dirt from mile 0 to mile 1.5, gravel from mile 1.5
to 3.5, and paved from mile 3.5 to 10. This could be represented as a route with the fol-
lowing section attributes:
ARC #5 ARC #6
where the From- and To- measure attributes are the miles at the start and end of each
section. Corresponding to these attributes are the From- and To- positions which are the
percentage of the arc. For example, Arc#5 is 4.5 miles long. Section#1 is from the start
of Arc#5 (0) to 33.33 percent (1.5 / 4.5) of Arc#5 at mile 1.5.
EVENT TYPES
An event describes a location or a section along a route. Examples from transportation
include traffic accident event locations, road surface types for each road section, or
speed limit along each road section. There are three types of events, depending on how
measures are used to locate the events. The three event types are point, continuous, and
linear.
Point events are events that occur at a precise point location on a route. Examples in-
clude the location of fire hydrants on a street route or the location of stops along a na-
ture trail route. An example event table for nature trail stops might be as follows:
Continuous events represent the locations where a change occurs along a route. Ex-
amples include the location where the speed limit changes on a street route or the loca-
tion where the class of white water changes along a river route. An example event table
for a white-water river might be as follows:
From the start of the route until distance 10, the river is class 3. From distance 10 to
distance 16, the river is class 2, and so on. Typically the last record of the continuous
event table represents the end of the route. In the above example, the river is class 1 from
distance 35 to distance 50.
Linear events represent the from- and to- locations along a route. Therefore they can
be discontinuous. Examples include sections of a street route that were paved in 1999,
or stretches of a stream route that contain spawning sockeye salmon. An example linear
event table for a stream route might be as follows:
DYNAMIC SEGMENTATION 79
All events must have at least two attribute fields. First, there must be a key field that
identifies which route each event belongs to. Second, there must be a measure field that
identifies the location of the event along the route.
CREATING EVENTS
There are three basic ways to create events, 1) tabular entry, 2) creating point events
from a point theme, and 3) creating linear or continuous events from a polygon theme.
Creating events by tabular entry can be done either by using the GIS tables utilities or
an external database management system. The basic process is to create an event table,
define the appropriate columns, and then enter the correct data. There will be one record
for each event. Once an event table has been created, the key field and measure field are
used to link the event table to a route system. This is typically done using the
EVENTSOURCE tool. The following example establishes the event source relationship
between the rapids.tbl event table and the river route system.
80 PRACTICAL GIS ANALYSIS
To establish the event source, you would specify the relationships linking the event
table to the route system. For example, you would specify the event source type is con-
tinuous, the event key field is Route-ID, while the route key field is Routelink#, and the
event to-measure field is To_Distance. Once the event source has been established, the
GIS can compute the locations of the events. For example,
DYNAMIC SEGMENTATION 81
CLASS 1
CLASS 2
CLASS 3
1 2
ROUTE #1
3 4
You can create point, linear, or continuous events by creating event tables. You can
also create a point event table from a point theme. You could do this using the AD-
DROUTEMEASURE tool. You need to specify the name of your point theme, the name
of your route theme and route system, the name of your output event table. You can also
specify a search radius if your points are not perfectly aligned with your route system. As
an example, imagine that we have a route system for irrigation and a point theme of gate
locations and when they were last inspected.
We enter the ADDROUTEMEASURE command and specify GATES as the name of
our point theme, IRRIGATION as the name of our route theme and route system, and
cleaned_gates.tbl as the name of our output event table. The output event table contains
the measures in the route system for each gate. You could also link the date of cleaning
from the original gates point attribute table.
82 PRACTICAL GIS ANALYSIS
You can also create event tables from a polygon theme. As an example, imagine that
you have a line theme of hiking trails and a polygon theme of forest types. You want to
create an event table delineating the forest types along the hiking trails. You could do
this using the POLYGONEVENTS tool. You need to specify the name of your polygon
theme, the name of your route theme and route system.
The resulting event table would contain all the information from the polygon theme
for each route section:
DYNAMIC SEGMENTATION 83
ANALYSIS TOOLS
Many of the line analysis tools can be used with routes or sections after you have con-
verted the selected routes or sections to arcs. Important tools that you can use with
routes/sections and events are listed in the following table:
RESELECT
Creates a route system by selecting routes or sections specified by the user
Imagine that we have a line theme of streams as follows:
84 PRACTICAL GIS ANALYSIS
RESELECT ROUTE-ID EQ 2
What would the output theme look like? In this example, route 2 is selected for the out-
put theme . . .
DYNAMIC SEGMENTATION 85
ROUTEARC
Convert each route into one arc in the output theme
Imagine that we have a line theme of streams as follows:
What would the output theme generated by ROUTEARC look like? Route 1 would
be output as a single arc . . .
86 PRACTICAL GIS ANALYSIS
SECTIONARC
Convert each section into one arc in the output theme
Imagine that we have a line theme of streams as follows:
What would the output theme generated by SECTIONARC look like? Each section of
route 1 would be output as an arc . . .
DYNAMIC SEGMENTATION 87
ROUTESTATS
Computes descriptive statistics for routes and sections
The following table describes the types of descriptive statistics that are available.
Imagine that we have a line theme of streams as follows:
88 PRACTICAL GIS ANALYSIS
What would the output tables generated by ROUTESTATS contain, if you requested the
following descriptive statistics?
RouteStats: numsections RouteStats: arclength SectionStats: measurelength
In this example, there is one route. The route statistics requested are the total number
of sections in route#1 (3 sections) and the total length of the route in GIS coordinate
units (295 + 210 + 65 = 570). The section statistics requested is the length of each sec-
tion in measure units. Section 1 is 517 units long, section 2 is 885 517 = 368, and sec-
tion 3 is 1000 885 = 115 measure units long.
EVENTPOINT
Convert point events into a new point theme
Imagine an example of stops along a nature trail . . .
DYNAMIC SEGMENTATION 89
The output point theme would be as follows:
90 PRACTICAL GIS ANALYSIS
EVENTARC
Convert linear or continuous events into a new line theme
As an example, we have an event table of rapid class along a route corresponding to a
float down a river.
OVERLAYEVENTS
Combine two or more event tables
As an example, imagine that you have two event tables: one of bird observations, and
one of vegetation types along a transect.
DYNAMIC SEGMENTATION 91
There are two possible options to use with OVERLAYEVENTS: intersect or union.
Intersect will return only the events that occur at the same location in all event tables.
Overlay will return all the events from any location in any event table.
DISSOLVEEVENTS
Combines adjacent records in linear event tables if they have the same value
You already learned about DISSOLVE in the chapter on line analysis. DIS-
SOLVEEVENTS is analogous and can be used to simplify event tables.
As an example, we could use Veg_code as the dissolve item from the following event
table, and produce the following output event table by using DISSOLVEEVENTS.
92 PRACTICAL GIS ANALYSIS
DYNAMIC SEGMENTATION EXERCISES
1) You have a route system of irrigation canals as follows. Circle the first section of
route#1 and the second section of route#3.
DYNAMIC SEGMENTATION 93
2) You have the following point theme of auto accidents and network of streets.
94 PRACTICAL GIS ANALYSIS
You enter the ADDROUTEMEASURE command and specify Accidents as the name
of your point theme, Streets as the name of your route theme and route system, and
acc_events.tbl as the name of our output event table. You specify a search radius of 100
meters. Fill in your output event table:
3) You have a route system of streams as follows.
DYNAMIC SEGMENTATION 95
Circle the stream section that has the highest density of Sockeye Salmon.
Circle the stream section that has the highest density of King Salmon.
4) You have a route system of streams as follows. What would be the output if you
run RESELECT on your streams with the following expression:
96 PRACTICAL GIS ANALYSIS
5) You have the following route system. Draw what the output theme would look like
if you applied the ROUTEARC command to this route system.
DYNAMIC SEGMENTATION 97
6) You have the following route system. Draw what the output theme would look
like if you applied the SECTIONARC command to this route system.
98 PRACTICAL GIS ANALYSIS
7) You have the following bus route system:
DYNAMIC SEGMENTATION 99
What would be the results if you run ROUTESTATS on your route system with the
following options?
RouteStats: ARCLENGTH
RouteStats: MEASURELENGTH
8) You have the following point and linear events along a street route system. Draw
the output themes that would result from using EVENTPOINT and EVENTARC.
100 PRACTICAL GIS ANALYSIS
9) You have two event tables of king salmon counts along a stream route. Fill in the
output table from executing OVERLAYEVENTS with these two event tables.
DYNAMIC SEGMENTATION 101
Output Point Theme:
Output Line Theme:
10) You have the following route system and irrigation event table. Fill in the output
event table resulting from using DISSOLVEEVENTS with Pipe_Type as the dissolve
item.
102 PRACTICAL GIS ANALYSIS
Chapter 7
Polygon Analysis
INTRODUCTION
A polygon is a GIS feature that has an area and perimeter. Typical polygon themes may
include themes such as country or state boundaries, township sections, vegetation poly-
gons, soil mapping units, lakes, parcels, or watersheds.
The following are some GIS tools that you can use in analyzing polygons:
GIS tools already covered in point and line analysis chapters:
COUNTVERTICESCount the number of vertices (including nodes) in each poly-
gon and add the count as a polygon attribute called VERTICES.
BUFFERGenerate buffer polygons of a user-specified distance around polygons.
RESELECTCreate a new polygon theme by selecting polygons using a logical ex-
pression.
DISSOLVEMerge adjacent polygons if they have the same value for a user-speci-
fied polygon attribute.
New tools:
ELIMINATE Merges selected polygons with neighboring polygons that have the
largest shared border between them. Used to remove sliver polygons.
CLIPUse a polygon theme as a cookie-cutter to cut out features from a point,
line, or polygon theme. A cut and copy operation to produce a single output theme.
SPLITA cut and copy operation that uses each polygon as a separate cookie-
cutter to produce many output themes.
ERASEEliminate polygons in a theme from the area covered by polygons in a
second theme. A cut operation.
UPDATEReplace polygons in a theme with polygons from a second theme. A cut
and paste operation.
INTERSECT, UNION, IDENTITYTransfer the polygon attributes to a second
polygon theme (spatial join).
103
DISSOLVE
Merges adjacent polygons if they have the same attribute value.
The output polygons will only contain the item used for the dissolving and no additional
attributes. If #ALL is used as the dissolve item, then original polygon attributes will be
output but User-IDs will be altered. The #ALL option is typically used following the
edge-matching of tiles or map sheets.
In the following example, MapClass is used as the dissolve item
104 PRACTICAL GIS ANALYSIS
ELIMINATE
Merges selected polygons with neighboring polygons that have a shared border be-
tween them.
ELIMINATE is often used to remove sliver polygons. You can control merging of se-
lected polygons by using the BORDER or AREA keywords. The default, BORDER
merges a selected polygon with a neighboring unselected polygon by dropping an arc
and retaining the longest shared border. AREA drops an arc and retains the polygon
with the largest area.
In the following example, we use ELIMINATE with RESELECT AREA LT 100 AND
AREA GT 0 on the following theme and the BORDER option:
POLYGON ANALYSIS 105
CLIP
Use a polygon theme as a cookie cutter to cut out features from a point, line, or
polygon theme. A cut and copy operation to produce a single output theme. You
can clip point, line, or polygon themes.
Imagine that you have a theme containing all lands owned by the state Department of
Natural Resources. You want to create a new theme of vegetation polygons within these
properties. So you clip VEGPOLYS with DNR_LANDS:
106 PRACTICAL GIS ANALYSIS
SPLIT
A cut and copy operation that uses each polygon as a separate cookie cutter to
produce many output themes. You can split point, line, or polygon themes.
Imagine that you have a polygon theme of watershed boundaries and a wells point
theme. You want to produce separate point themes of wells inside each watershed 1 and
watershed 2.
SPLIT WELLS WATERSHEDS WATERSHED-ID POINT
Enter the 1st theme: WELLS_SHED1
Enter item value: 1
Enter the 2nd theme: WELLS_SHED2
Enter item value: 2
Enter the 3rd theme: end
POLYGON ANALYSIS 107
ERASE
Eliminate polygons in a theme from the area covered by polygons in a second
theme. A cut operation that works on point, line, or polygon themes.
Imagine that you have a line theme of winter skiing trails and a polygon theme of wet-
lands. We create a new theme called HIKE_TRAILS representing the trails above water
in the summertime.
108 PRACTICAL GIS ANALYSIS
UPDATE
Replace polygons in a theme with polygons from a second theme. A cut and paste
operation. The keyword KEEPBORDER or DROPBORDER specifies whether or
not the outside border of the update theme will be kept after it is inserted into the
input theme.
Imagine that you have a polygon theme of vegetation and polygon theme of planted
spruce seedlings. You want to update your vegetation theme.
POLYGON ANALYSIS 109
SPATIAL JOINS: INTERSECT, UNION, IDENTITY
Transfer the polygon attributes to a second polygon theme (spatial join).
You have already learned about the INTERSECT tool for intersecting points and
lines with polygon themes.
INTERSECT creates a new theme by overlaying a point, line, or polygon theme with
an intersecting polygon theme. The output theme contains only the features inside the in-
tersecting polygons.
110 PRACTICAL GIS ANALYSIS
UNION creates a new theme by overlaying two polygon themes. The output theme
contains the combined polygons and attributes of both themes. Note that only polygon
themes can be combined using UNION.
IDENTITY creates a new theme by overlaying a point, line, or polygon theme with
an intersecting polygon theme. The output theme contains only the all original point,
line, or polygons as well as the attributes transferred by the intersecting polygon theme.
POLYGON ANALYSIS EXERCISES
1) Polygons are assumed to be pure in their content and discrete and absolute in their
boundary locations. In other words, they are assumed to be homogeneous with sharp,
distinct boundaries. This is rarely the case. Rank the following polygon themes in terms
of their likely purity and boundary sharpness.
2) You have polygon themes of a wildfire burn from 1963 and a burn from the year
2000. Fill in the following flowchart to determine the total number of hectares from the
1963 burn that burned in 2000.
POLYGON ANALYSIS 111
3) You have a line theme of rivers and streams and a polygon theme of three water-
shed areas (1= Rock Creek Watershed, 2=Clear Creek Watershed, 3=Willow Creek Wa-
tershed). For each watershed, you want to know the density of rivers and streams in me-
ters per hectare of the watershed. Your GIS coordinate system is in meters.
Fill in the following flowchart to solve your GIS problem.
112 PRACTICAL GIS ANALYSIS
4)You have a point theme of mallard nest locations with three age classes of nesting
mallards. You also have a polygon theme of land cover types (1= tussock, 2=low shrub,
3=high shrub, 4=barren). Your GIS coordinate system is in meters.
You want to estimate the density of nests in each land cover type by mallard age class
as follows:
Fill in the following flowchart to solve your GIS problem.
POLYGON ANALYSIS 113
5) You have a polygon theme of vegetation types (vegpolys) and a polygon theme of
watershed areas (watersheds). You want a table showing the area of each watersed and
the area of black spruce in each watershed.
114 PRACTICAL GIS ANALYSIS
6) You have three polygon themes: SOIL, VEG, and TINPOLYS. SOIL contains soil
information, VEG contains vegetation information, and TINPOLYS are triangles where
the slope and aspect of each triangle has been estimated.
You are checking for possible errors in the vegetation theme. You want to find if there
are any aspen polygons that occur over poorly drained soils or on northerly facing slopes
that are over 10 percent in slope gradient. These would likely be errors in vegetation
mapping because aspen typically grows on relatively dry, warm sites.
Fill in the following flowchart to solve your GIS problem.
POLYGON ANALYSIS 115
7) You have three candidate routes for a new power line. The cost of building the line
varies by amount of land owned privately, by soil type to build on, and by amount of
wetlands to cross and as well as total distance. You want to know this type of informa-
tion for the three routes. Fill in the following flowchart with the most appropriate GIS
tools:
116 PRACTICAL GIS ANALYSIS
8) You have a pretimber harvest polygon theme of alder clumps and a post-timber
harvest point theme of planted spruce seedlings. You want to find the mean height
growth of the spruce seedlings in areas that were in or within 10 meters of pre-harvest
alder versus areas that were at least 10 meters away from the pre-harvest alder clumps.
Fill in the following flowchart with the most appropriate GIS tools:
POLYGON ANALYSIS 117
9) You have polygon themes of lakes. One theme is from early summer and the other
theme is from late summer. You want to create a new theme of the best lakes for water-
fowl habitat . . . lakes that had at least 1 hectare of exposed mudflats in late summer
due to drawdown of the lake water level. Fill in the following flowchart to solve your
GIS problem.
118 PRACTICAL GIS ANALYSIS
10) You have a point theme of well locations, a point theme of gas station locations,
a soils polygon theme, and a parcels polygon theme containing the address of the owner
of each parcel.
Outline what tools you would use to develop a text file of all well owners that have
wells on sand, loamy sand, or sandy loam soils within a kilometer of a gas station.
Chapter 8
Grid Analysis
INTRODUCTION
GRID AND VECTOR GIS WORLDS
Until now, you have worked mainly in the vector world of points, lines, and polygons.
These are discrete features represented by X,Y coordinates, and each feature has a
unique identification number. Grids are different because they may represent a continu-
ous surface instead of discrete features. In general, the grid GIS data model has several
advantages relative to the vector GIS data model of points, lines, and polygons:
Grids allow for faster analysis, especially for any overlay type analysis
Grids allow for modeling continuous surfaces such as wildfire or oil spill spreads,
or determining the optimal path through a surface.
Grid is the natural model for many data sources in natural resources such as digital
elevation models, scanned maps, and land cover data derived from satellite data.
Grids allow you to represent the world as a fuzzier system. For example, in the
vector world, vegetation polygons are often assigned a vegetation type from visual
interpretation of aerial photographs. Typically, relatively large polygons are delin-
eated because of an arbitrary minimum mapping unit criteria used during the air
photo interpretation. With land cover grids, the minimum mapping unit is the size
of the grid cell, and therefore heterogeneous areas can be better represented. In the
vector world, polygons are assigned as one homogeneous land cover class. In the
grid world, the same area may be several land cover classes, and the probability of
correct classification of each grid cell can also be stored.
The vector model of points, lines, and polygons does have some advantages over the
grid model:
Since features are stored as X, Y coordinates, it usually takes less disk space to
store points, lines, and polygons relative to grids that have the same precision of
spatial detail.
The vector model allows for modeling of linear networks and the incorporation of
measured events via dynamic segmentation.
The vector model is a natural model for many data sources derived from GPS sur-
veying, or for utility, urban, and transportation data.
119
The vector model allows you to represent the spatial location of features with high
precision relative to grid representations.
Your choice of vector versus grid analysis will depend largely on your data sources
and applications. In this chapter, you will learn that many of the grid operations are
analogous to vector operations.
There are potentially hundreds of grid operations. In this chapter, you will learn about
some common grid operations, especially applicable for natural resource applications.
These operations include grid arithmetic operations, selection operations, grouping opera-
tions, distance operations, optimal path operations, and common topographic operations.
GRID ARITHMETIC OPERATORS
Grids can be added, subtracted, multiplied, and divided. If an output cell value is un-
defined, it is assigned a value of NODATA. For example, imagine that you have two
grids that you want to divide with:
GRID3 = GRID1 / GRID2
The two input grids are integer type, therefore the output grid will also be integer
type. For example, cell(row 2, column 1) is output as
1
2 or 0.5 and this is output as an
integer value of 0. The cells that are undefined because they were divided by zero are as-
signed to NODATA. Notice that zero is a valid grid cell value while NODATA indicates
no valid data for that cell.
In order to have an arithmetic expression return fractions, at least one of the input
grids must be floating point type. You can use the FLOAT function to convert integer
grids to floating-point grids. For example:
GRID4 = FLOAT(GRID1) / FLOAT(GRID2)
120 PRACTICAL GIS ANALYSIS
GRID ANALYSIS 121
Keep in mind that NODATA cells always remain NODATA cells in arithmetic
operations.
You can change NODATA to any valid value by using the CON function which you
will learn about soon.
SELECTION OPERATIONS: TEST, SELECT, CON
TEST
Returns a 1 if the cell meets some logical expression criteria, otherwise the output
cell is assigned a zero.
As an example, imagine that you have a grid of land cover types with 1= water, 2= conif-
erous forest, 3= broadleaf forest, 4= mixed forest, and 5=marsh. You want to create a
grid where 0 represents nonforest and 1 represents forest.
Forest = TEST(Land_Cover, Value NE 1 and Value NE 5)
SELECT
Returns the original cell value if the cell meets some logical expression criteria, oth-
erwise the output cell is assigned as NODATA.
As an example, imagine that you have a grid of elevation values with 255 representing
fill pixels along the edge of the grid. If you asked for the mean elevation value of the
grid, zeros would be included in the calculation. You want to use the SELECT function
to convert the 255s to NODATA.
Elev2 = SELECT ( ELEVATION, Value NE 255 )
122 PRACTICAL GIS ANALYSIS
CON
Tests for a user-specified logical expression and returns user-specified values.
The CON function stands for conditional test. The user specifies what happens if the test
is true or false. For example you could create a HIGH_ELEV grid as follows:
HIGH_ELEV = CON ( ELEV > 1000, 1, 0 )
If the elevation value is greater than 1000, the output cell receives a 1, and if it is less
than or equal to 100, an output cell receives a zero.
If you do not indicate the output value for a false test, NODATA is assigned auto-
matically. For example, you could create a HIGH_ELEV grid as follows:
HIGH_ELEV = CON ( ELEV > 1000, ELEV )
If the elevation value is greater than 1000, the output cell receives the elevation value,
and if it is less than or equal to 100, an output cell receives a NODATA .
You can convert NODATA cells to some other value by using the CON(ISNULL . . .
as follows:
GRID ANALYSIS 123
FILLZEROS = CON( ISNULL( HIGH_ELEV ), 0, HIGH_ELEV )
The condition tested asks: Is the grid cell value of HIGH_ELEV NODATA or null? If
it is, a zero value is returned, and if it is not, the original cell value of HIGH_ELEV is re-
turned:
GROUPING OPERATIONS: RECLASS, REGIONGROUP
RECLASS
Changes cell values based on the rules in a remap table.
Lets start with a simple example. You have a single line remap table as follows: 2 3 : 1
The remap table specifies a rule that cells with values of 2 or 3 are reclassed to a value
of 1 in the output grid. A value of 1 or 4 would be assigned NODATA if you have a NO-
DATA flag set on. If you have a NODATA flag set to off, the original values of 1 or 4
would be output, while 2 or 3s would be output as 1 values.
The RECLASS function is often used to classify quantitative data into categories. As
an example, you could use the RECLASS function to create elevation classes from a grid
of elevation values. Elevations of 100 to 300, be reclassed to cell values of 1. Elevations
above 300 and less than or equal to 400 will be reclassed to cell values of 2. And eleva-
tions above 400 and less than or equal to 500 will be reclassed to cell values of 3.
124 PRACTICAL GIS ANALYSIS
REGIONGROUP
Groups cells of the same value that touch each other.
REGIONGROUP allows you to aggregate cells that have the same value into groups. As
an example, you might have a vegetation grid and you are interested in small patches of
a particular class. REGIONGROUP starts at the upper left corner cell and proceeds left
to right assigning group numbers based on cells that touch and have the same cell values.
GRID ANALYSIS 125
Where the Value is the group number, and the Link is the original value before the
grouping occurred. For example, group 5 originally had values of 4 and were cells that
touched.
DISTANCE OPERATIONS: EXPAND, EUCDISTANCE
EXPAND
Expands user-specified cell values (analogous to BUFFER).
The user specifies the number of cells to expand and the cell values that are candidates
for expansion. For example, EXPAND( Grid, 4, list, 0,1 ) would select any cells that
have values of 0 or 1 and expand them out by 4 cells. When two values compete to ex-
pand into the same area, the conflict is resolved based on the value of the majority of
surrounding cells. Cells can also expand into NODATA cells.
As an example, imagine that you have a grid of lakes and you want to buffer the lakes
by one grid cell by using the EXPAND function.
EUCDISTANCE
Calculates the distance to the closest non-NODATA cell for each cell.
The distance calculation is from the center of one cell to the center of the closest non-
NODATA cell. The distance is in grid cell units, rather than the GIS coordinate system
units. For example,
126 PRACTICAL GIS ANALYSIS
As an example, imagine that you have a grid of lakes and you compute the closest dis-
tance to any lake for each grid cell.
GRID ANALYSIS 127
OPTIMAL PATH OPERATIONS: COSTDISTANCE, COSTPATH
You learned about optimal paths in the Network Analysis chapter. With grids, you can
estimate the optimal path from source cells to destination cells in an analogous manner.
COSTDISTANCE
Estimates the minimum accumulative cost to resource cells.
The COSTDISTANCE function computes the minimum accumulative cost for a grid and
outputs this cost grid, a grid of directions associated with the minimum cost, and a grid
allocating the source cells. This information is required to compute the optimal path
using the COSTPATH function.
Imagine that you have two resources, coded with values 1 and 2. You also have a
grid representing the cost of crossing each grid cell. This cost might be travel time, road
construction cost, fuel consumption cost, or cost in terms of aesthetics. You want to
compute the minimum accumulative cost for all cells in your area.
What is the least costly path to reach a 1 or a 2 in the source grid? The cost is zero if
you are already sitting in a source grid cell . . . assign these to the output grid.
The calculation of accumulative costs is modeled based on a node-link model of
graph theory. The center of each cell is represented by nodes and all possible costs of
movement from one cell to another are represented by links. For example, our grid is
represented as follows.
128 PRACTICAL GIS ANALYSIS
The cost to move horizontally or vertically from one cell to another is the total cost of
the two cells divided by 2 cells. For example, to move from cell (1,1) to cell (2,1) would
cost you (1+3) / 2 cells = 2.0.
The cost to move diagonally from one cell to another is the total cost of the two cells
divided by 2 cells times the diagonal distance (square root of 2 = 1.4142). For example,
to move from cell (4,3) to cell (3,2) would cost (2+5)/2 * 1.4142 = 4.9497
The algorithm computes the lowest possible accumulative cost for all cells neighbor-
ing the most recently assigned output cells. For example, to get to a resource from cell
(4,3) you could move diagonally to cell (3,2) and it would cost you 4.9497. Or you
could move from cell (4,3) up to cell (4,2) and then left to cell (3,2) and it would cost
you (5+3)/2 + ( 3+2)/2 or 6.5. Or you could move from cell (4,3) left to cell (3,3) and
then up to cell (3,2) and it would cost you (5+7)/2 + (7+2)/2 = 10.5. The lowest of the
possible moves is the diagonal move which costs you 4.9.
The minimum accumulative costs of all neighboring cells is assigned to the output
grid. In this example, it is 1.5 at cell (2,6). Then the accumulative cost of the new neigh-
boring cells is computed. For cell (3,6) the accumulative cost is the cost of moving from
cell (3,6) to cell (2,6) to cell (1,6) = (2+2) / 2 + 1.5 = 3.5. For cell (3,5) the accumulative
cost is the cost of moving from to cell (2,6) to cell (1,6) = 1.4142( 5+2)/2 + 1.5 = 4.9497
+ 1.5 = 6.4497.
GRID ANALYSIS 129
The minimum accumulative costs of all neighboring cells is assigned to the output
grid. In this example, it is 2.0 at cell (1,1). Then the accumulative cost of the new neigh-
boring cells is computed. In this case there are no new costs to compute.
The minimum accumulative costs of all neighboring cells is assigned to the output
grid. In this example, it is 2.5 at cell (4,2). Then the accumulative cost of the new neigh-
boring cells is computed. For example, for cell (5,1) the accumulative cost is the cost of
moving from cell (5,1) to cell (4,2) to cell (3,2), which is 1.4142( 3+3/2) + 2.5 = 6.7426.
For cell (5,2) the accumulative cost is the cost of moving from cell (5,2) to cell (4,2) to
cell (3,2), which is (7+3)/2 + 2.5 = 7.5. For cell (5,3) the accumulative cost is the cost of
moving from cell (5,3) to cell (4,2) to cell (3,2), which is 1.4142 (6+3)/2 + 2.5 = 8.8639.
The minimum accumulative cost of all neighboring cells is assigned to the output
grid. In this example, it is 2.5 at cell (1,5). Then the accumulative cost of the new neigh-
boring cells is computed. For cell (1,4), the accumulative cost is (1+4)/2 + 2.5 = 5.0. And
for cell (2,4), the accumulative cost is (4+1)/2 + (1+4)/2 + (4+1)/2 = 7.5.
The minimum accumulative cost of all neighboring cells is assigned to the output grid.
In this example, it is 3.5 at cell (3,6). Then the accumulative cost of the new neighboring
cells is computed. For cell (4,5), the accumulative cost is NODATA because that is the
COST Grid cell value. And for cell (4,6), the accumulative cost is (1+2)/2 + 3.5 = 5.0
The minimum accumulative cost of all neighboring cells is assigned to the output grid.
In this example, it is 4.0 at cells (4,1) and (2,2). Then the accumulative cost of the new
neighboring cells is computed.
The minimum accumulative cost of all neighboring cells is assigned to the output grid.
In this example, it is 4.5 at cells (1,2) and (3,3). Then the accumulative cost of the new
neighboring cells is computed.
The iterative process continues until the minimum accumulative cost has been as-
signed to all cells. Then the grid is output as the result of the COSTDISTANCE function.
And a companion Backlink grid is output as a grid of the direction of minimum accu-
mulative costs. The following are the codes used for these directions.
130 PRACTICAL GIS ANALYSIS
Another companion grid that is output is a minimum cost allocation grid. This grid
indicates which resource is the least costly to get to from each cell.
COSTPATH
Estimates the least cost path from source cells to destination cells.
In order to estimate the optimal path, information about the minimum accumulative
cost and associated backlink direction is needed. Typically this cost information is sup-
plied by first running the COSTDISTANCE function.
As an example, imagine that we have the following cost information from our previ-
ous COSTDISTANCE example. You want to estimate the optimal path to get from one
cell location to a resource cell as cheaply as possible. The minimum accumulative
costs grid might be expressing the cost in terms of dollars, time, distance, or any other
cost measurement. The output optimal path grid has special codes as follows: 1 = a re-
source grid cell, 2= shared paths, 3= first optimal path, 4= second optimal path, 5= third
optimal path, and so on.
In the following example, there is one optimal path (the first which is coded with 3s)
and the total cost of reaching a resource is 13.1.
GRID ANALYSIS 131
In the next example, there are two optimal paths for the two from-cells. The first op-
timal path is from the 200 cell and is straight left and costs 7.5. The second optimal path
(coded with 4s) is the same as the path from our first example and costs 13.1. The Path
Grid cell coded with a 2 is a cell that contains both paths going through it.
In some applications, you may be interested in the optimal path from any cell instead
of from each individual cell. You can specify this by using the BYLAYER flag with the
COSTPATH function. As an example, imagine that you have an existing road and want
to estimate the optimal path from the existing road to any resource cell:
132 PRACTICAL GIS ANALYSIS
The COSTPATH function finds the minimum accumulative costs grid and looks at
the values for all road cells. The lowest minimum accumulative cost is 4.9 and the
COSTPATH function codes that cell location with a 3 as the optimal path from the road
to any resource.
TOPOGRAPHIC OPERATIONS: SLOPE, ASPECT
SLOPE
Estimates rate of maximum slope change for each cell.
The slope is computed from the 3 row by 3 column neighborhood of each cell. Basically,
a plane is computed for this neighborhood and the maximum slope of the plane is re-
turned to the cell being processed. At the edge of the grid, at least three cells (outside the
grids extent) will contain NODATA as their z values. These cells will be assigned the
center cells value. The result is a flattening of the 3x3 plane that is fit to these edge cells,
which thus usually leads to a reduction in the slope.
As an example, imagine that you have a grid of elevation values and you want to es-
timate the percent slope of each cell. In this example, the cell size is 100 meters wide and
long and the elevation values are in meters above sea level.
GRID ANALYSIS 133
ASPECT
Estimates the slope direction of each cell.
Aspect estimates the maximum slope direction of each cell. The output value is the com-
pass direction expressed from 0 to 360 (for example, 0=north, 45=northeast, 90 = east).
If the estimate slope is zero, there is no slope direction and a value of 1 is returned for
aspect of flat cells.
As an example, imagine that you have a grid of elevation values and you want to es-
timate the slope direction of each cell. In this example, the cell size is 100 meters wide
and long and the elevation values are in meters above sea level.
GRID ANALYSIS EXERCISES
1) You run REGIONGROUP on the following grid. Fill in the output grid:
134 PRACTICAL GIS ANALYSIS
2) What would the output grid contain if you use TEST with the following logical ex-
pression: Value in {1,3,8,9}
3) What would the output grid contain if you use SELECT with the following logical
expression: Value in {1,3,8,9}
GRID ANALYSIS 135
4) You have a grid of willow cells. You want to buffer all willow areas by 1 cell. Fill
the output grid with the appropriate values after the EXPAND function is executed.
5) You have the following grids output from running the COSTDISTANCE
function.
136 PRACTICAL GIS ANALYSIS
Fill in the output grid resulting from estimating an optimal path using the COST-
PATH function:
6) You have a grid of forest types and a grid of three watersheds. You want to pro-
duce the following information.
Fill in the following flowchart with the appropriate grid tools to solve your problem.
GRID ANALYSIS 137
7) You have a 1/0 grid of all lightning strike locations in interior Alaska during last
summer. You also have a grid of elevation values in meters for the same area. Each grid
cell is 1000 by 1000 meters. You want to determine the following information:
Fill in the following flowchart with appropriate grid operations to solve your GIS
problem:
138 PRACTICAL GIS ANALYSIS
8) You have one grid (veg) that contains values of 71,72,73,74,75 for open water and
values of 63,64,65,67 for wetland. The grids have a cell size of 25 meters. Draw a flow
chart showing how you would produce a grid of wetlands cells that are on the shorelines
of open water.
Fill in the following flowchart with appropriate grid operations to solve your GIS
problem:
GRID ANALYSIS 139
9) A forester is planning potential harvest units. She wants to find large (>100 ha)
stands of closed aspen, birch, or aspen/birch forest (VEGCLASS 16-19) The SIZECLASS
of the stands should be sawtimber or pole size (3 or 2) and the stands should be at least
100 meters from open water (VEGCLASS 80-89) and at least 50 meters from marsh
areas (VEGCLASS 77-79). All grids have a cell size of 25 by 25 meters.
Fill in the following flowchart with the appropriate Grid tools to solve your GIS
problem:
140 PRACTICAL GIS ANALYSIS
10) You have a grid of vegetation types (VEG), soil drainage class (DRAINAGE) and
a grid of elevation values (ELEV). You are checking for possible errors in the VEG grid.
Aspen typically grows on warm sites such as coarse, well-drained soils or warm,
southerly-facing slopes. You want to find any cells that have an aspen vegetation class
occurring on either poorly drained soils or cold slopes (slopes with an aspect from north
to east, northwest to west and a gradient greater than 10%).
Fill in the following flowchart with the appropriate Grid tools to solve your GIS
problem:
GRID ANALYSIS 141
Chapter 9
Image Analysis
INTRODUCTION
An image is a special type of grid taken from some remote sensing device like a camera
or digital scanner. Common image sources in GIS are digital orthophotos, satellite im-
ages, aerial video images, and scanned documents. Image processing techniques can be
very sophisticated and entire software packages for image processing of remotely sensed
images exist. Basic image processing operations are available in most GIS systems and
are therefore covered in this chapter. These operations include image display and en-
hancement, image rectification, unsupervised and supervised image classification, and
classification accuracy assessment.
IMAGE DISPLAY
A single band image is a grid with one value in each grid cell. These values are some-
times called digital numbers or DNs in remote sensing. To display a single band image,
you control your output display intensity with image DNs.
Lets start with a simple example. Imagine that you have a display device that can
only display in four patterns.
The range of digital numbers for your particular sensor is 0 to 255. Therefore one
possible display rule is to use this range to assign pixel display as follows:
143
The rule from the previous linear function means the following:
Using the above linear function we could display your digital image as follows:
144 PRACTICAL GIS ANALYSIS
Imagine that coniferous vegetation has DNs ranging from 80 to 90, and broadleaf
vegetation has DNs ranging from 90 to 100. The conifer pixels do not appear different
than broadleaf pixels on your displayed image.
You could enhance the visual contrast between broadleaf and conifer vegetation by
changing the linear contrast function as follows:
The rule from the previous linear function means the following:
And apply your linear function to display your digital image in the four output
shades:
IMAGE ANALYSIS 145
Notice that with the increased image contrast, the conifer and broadleaf pixels are
displayed with different patterns.
LINEAR CONTRAST STRETCH
The linear function that you have been using is called a linear contrast stretch. Typically
it is used with an image and a computer display that displays in video intensity values
from 0 to 255. The GIS can compute the output video intensity for any pixel by using
the following formula:
Video Intensity = [( Pixel Value Minimum Pixel Value) / (Max Min)] * 255
Here is a picture of the same linear contrast function:
Lets say that you want to display your image to your computer screen using the lin-
ear contrast stretch. The minimum pixel value of your image is 10 and the maximum
pixel value is 98. You could calculate the linear stretch output values and then fill in the
appropriate video intensity values. As an example, the first row video intensity values
are:
146 PRACTICAL GIS ANALYSIS
By applying a linear contrast stretch, you can dramatically increase the range or con-
trast of your image display. For example, the original range of pixel values from your
image was : 98 10 = 88 and yet by using the linear contrast stretch, you can stretch the
pixel display intensity values to 0 to 255.
The linear contrast stretch is analogous to film exposure. As an example, here is the
same digital image of the Baltimore area displayed using different linear contrast
stretches:
IMAGE ANALYSIS 147
In the above example, each pixel value is controlling the video intensity
value with no stretch. But since there are few pixels with values greater than
150, essentially the video intensity of 150 to 255 is being wasted on those
few pixels.
148 PRACTICAL GIS ANALYSIS
Most of the land pixels have values between 25 and 100. The linear contrast function
above enhances the contrast of these pixels. Any pixels with values less than 25 will be
black (the water pixels for example). And any pixels with values above 100 will be dis-
played with a video intensity of 255. The pixels between 25 and 100 are displayed with
stretched video intensities ranging from 0 to 255.
IMAGE RECTIFICATION
Image rectification is the process of transforming a digital image from pixel (column,
row) coordinates into your GIS coordinate system. It is a relatively simple process with
satellite imagery that has little relief displacement. Typically a transformation called the
affine transformation is used. This transformation is the same transformation that is
used in GIS work when registering tics before digitizing new theme features.
THE AFFINE TRANSFORMATION
The affine transformation is a model of 2 equations that describe or fit a flat plane as
follows:
Xmap = a + b1(image col) + b2(image row)
Ymap = a + c1(image row) + c2(image col)
or
Xmap = a + b1(pixel X) + b2(pixel Y)
Ymap = a + c1(pixel Y) + c2(pixel X)
Imagine the following simple example. You identify four locations on your image and
estimate their map coordinates in the field using a GPS receiver.
From these four ground control points or GCPs, you can develop linear models to
predict any map coordinate for any image pixel (col,row) coordinate as follows:
IMAGE ANALYSIS 149
The affine transformation equations for these control points are:
XMAP = 427,000 + 10(XPIXEL) + 0(YPIXEL)
YMAP = 7,188,000 -10(YPIXEL) + 0(XPIXEL)
Notice that the slope of the line for the X linear model is positive while the slope of
the line for the Y linear model is negative. This is because the image col and map X co-
ordinate system increase in the same direction, while the image row and map Y coordi-
nate system increase in opposite directions as follows:
150 PRACTICAL GIS ANALYSIS
As a check, you could plug in any of the original image (col,row) coordinates and get
your map coordinate estimates for that ground control location:
The affine transformation can unwarp an image by changing the following proper-
ties:
Change in X, Y scale
Translation in X,Y space
Rotation of image in X,Y space
SCALING
The affine transformation can stretch the image in X and/or Y directions. For exam-
ple, imagine that you located four image pixels and estimated their map coordinates.
Xmap = 0 + 10(Xpixel) + 0(Ypixel)
Ymap = 05 (Ypixel) + 0(Xpixel)
In this example, the X stretch factor is (5000 2500) / ( 250 125) = 2500 / 125 =
20. This means that the rectified pixels will have a width of 20 in the map coordinate
space. The Y stretch factor is (5000 2500) / ( 5001000) = 2500/ 500 = -5. This
means that the rectified pixels will have a length of 5 in the map coordinate system.
TRANSLATION
The affine transformation can translate or shift the image in X and/or Y directions.
For example, imagine that you located four image pixels and estimated their map coor-
dinates.
IMAGE ANALYSIS 151
Xmap = 200 + 1(Xpixel) + 0(Ypixel)
Ymap = 5001(Ypixel) + 0(Xpixel)
In this example, the affine transformation picks up the image and shifts it 200 units to
the right and 100 units up into map space.
ROTATION
The affine transformation can rotate the image. For example, imagine that you lo-
cated four image pixels and estimated their map coordinates.
Xmap = 0.8 (Xpixel ) + 0.4 (Ypixel)
Ymap = -0.4(Xpixel) + 0.8(Ypixel)
152 PRACTICAL GIS ANALYSIS
Rotation is especially important in rectifying polar-orbiting satellite imagery. Because
the earth is rotating on its axis, a straight north to south orbit would result in images
taken later in the day for southern regions relative to northern regions of the globe. To
compensate for this, many satellites travel in a northeast to southwest orbit in an at-
tempt to maintain a constant local time in images acquired during the orbit. These im-
ages need to be rotated in order for the north-south axis to be oriented up and down.
IMAGE ANALYSIS 153
RECTIFICATION MODEL ERRORS
There will always be some amount of error in selection of ground control pixels and in
estimating their associated map coordinates. For example, imagine that you selected
eight ground control locations on an image as follows. The pixel coordinates in bold
represented pixels that you selected that were not exactly the correct pixels for that
ground control location.
Using the eight GCP links, the GIS computes the best linear rectification model from
your control point information . . .
Xmap = 432511.259 + 25.010 (Xpixel ) + -0.074 (Ypixel)
Ymap = 7187448.139 24.906 (Ypixel) + 0.087 (Xpixel)
And the GIS computes the model X,Y and total error for each control point. For ex-
ample, the predicted X,Y map coordinate for GCP#1 is:
Xmap Predicted = 432511.259 + 25.010 (100 ) + -0.074 (98) = 435005.007
Ymap Predicted = 7187448.13924.906 (98) + 0.087 (100) = 7185016.051
And the X or Y errors are the GPS map coordinate minus the predicted map coordinate.
Once the GIS has computed the X error and Y error for each ground control location,
it can compute the total error for each GCP link as follows:
154 PRACTICAL GIS ANALYSIS
And the overall model error, called the Root Mean Squared Error is computed as:
RMS error = Square Root [ (Sum of Total Errors Squared)/ Number of Links ]
= SQRT [ (16.81
2
+ 21.21
2
+ 16.03
2
+ 66.43
2
+ 55.31
2
+ 227.25
2
+ 58.60
2
+
47.80
2
) / 8 ]
= 90.7 meters
IMAGE RESAMPLING OPTIONS
Once you have developed a rectification model with a satisfactory RMS error, the GIS
applies the model to the image to translate the image from image (col, row) space to
map (X,Y) space. The original image must be resampled because there is typically not a
perfect match between image grid cells and the output map grid cells. With most GISs
you can choose from three resampling options: nearest neighbor, bilinear interpolation,
or cubic convolution.
Nearest Neighbor resampling selects the image pixel value that is closest to each out-
put grid cell. It has the advantages of being the fastest resampling method and of pre-
serving original pixel values. The disadvantage of nearest neighbor resampling is that the
rectified image tends to have stair-step shaped linear features.
Bilinear Interpolation resampling selects the four of the image pixels that are closest
to each output grid cell and outputs a weighted average based on these four pixels. The
advantage of this resampling option is that it produces a smoother image and an image
with better positional accuracy relative to the nearest neighbor option. The disadvantage
of bilinear interpolation is that it is slower relative to nearest neighbor resampling.
Cubic Convolution resampling selects the 16 of the image pixels that are closest to
each output grid cell and outputs a weighted average based on these 16 pixels. The ad-
vantage of this resampling option is that it produces a smoothest image relative to bilin-
ear interpolation or nearest neighbor resampling. The disadvantage of bilinear interpola-
tion is that it is very slow compared to bilinear interpolation or nearest neighbor
resampling.
UNSUPERVISED CLASSIFICATION
Once an image has been rectified, you can classify it into a grid of land cover classes.
There are two basic classification strategies: unsupervised and supervised classification.
In unsupervised classification, you ask the GIS to group pixels that have similar spectral
values into spectral classes. Then you decide what each spectral class represents in terms
of land cover.
The first step in unsupervised classification is to ask the GIS to create spectral classes
by grouping pixels that have similar values. You specify the parameters that the classifier
should use such as the number of spectral classes, the number of computational itera-
tions, the mininum number of pixels allowed in a spectral class, and the pixel sampling
interval.
There are many clustering algorithms for unsupervised classification. One commonly
used algorithm is the ISODATA or Iterative Self-Organizing Data Algorithm. As an ex-
ample, imagine that you want to produce a land cover grid from a two-band image. You
would first specify the clustering parameters as follows. A sample interval of 2 means
every second pixel will be sampled starting with the upper left pixel and proceeding left
IMAGE ANALYSIS 155
to right, row by row. This speeds up the processing time for initial spectral clustering.
The shaded pixels will be skipped in this step.
User-specified clustering parameters:
4 maximum number of classes, 2 iterations, sample_interval= 2
Start with 4 means at arbitrary intervals, then assign each sample to the closest class
mean. The range of pixel values is 0 to 120. 120/4 classes = 30, so the starting means are
(30,30), (60,60) (90,90), (120,120). Each sample pixel is assigned to the nearest mean.
156 PRACTICAL GIS ANALYSIS
Interation#2: Compute new class means and plot new class means, then assign class
members closest to each new class mean.
IMAGE ANALYSIS 157
The next step is to use some rule to predict the spectral class of every pixel in the
image. In this example, each pixel is assigned based on the closest spectral class means:
The final step would be for you to decide what each spectral class represents in terms
of cover type and assign symbology for each class. For example, imagine that you deter-
mined that spectral class 1 represents water, 2 represents vegetation, and 3 represents
sand. You could then assign symbols to each spectral class.
Classified Spectral Class Grid:
158 PRACTICAL GIS ANALYSIS
SUPERVISED CLASSIFICATION
In supervised classification, you sample known cover class areas and the GIS uses the
sample statistics from these areas to predict the cover class of each pixel in the image.
The sample areas are typically called training fields or training areas and you need at
least one training field for each cover class in your image.
Supervised classification is not necessarily better than unsupervised classification, it is
simply a different strategy in classification. Supervised classification has the following
advantages relative to unsupervised classification:
1) The user can delineate training fields from cover classes that could be mixed spec-
tral classes in unsupervised classification, such as shadow versus water versus a
burned area.
2) With maximum likelihood classifier, the user can produce probability estimates for
each classified pixel and thus have an estimate of the confidence in the cover
class predictions.
Supervised classification has the following disadvantages relative to unsupervised
classification.
1) Rare cover types that exist in the image area may not be known by the user and
therefore are not included in the classification.
2) An error in training field delineation can significantly alter the classification re-
sults. Each training field should be from large, homogeneous cover class areas and
this is sometimes difficult to achieve.
SEED PIXEL TRAINING FIELDS
You have several options in delineating training fields. You could use existing land cover
polygons to delineate training fields or you could digitize training polygons directly from
your displayed image. One problem with these two approaches is that sometimes it is
difficult to get a large training field without including some unrepresentative pixels.
Another approach, called the seed pixel approach, was invented to minimize this
problem. As an example, imagine that you want to delineate a training field of a burned
area from the following area of a large image. The first step is to point to a seed pixel
that most represents the burned area. In this example, the seed pixel is shaded in gray.
The seed pixel algorithm then searches neighboring pixels and develops a parallelepiped
or rectangle based on the minimum and maximum values within the neighboring pixels.
In this example, the parallelepiped would be as follows:
IMAGE ANALYSIS 159
The GIS would then search all surrounding pixels and include any touching pixels
that fall inside the seed pixel parallelepiped as a training field pixel representative of a
burned pixel. These pixels are shaded in gray.
160 PRACTICAL GIS ANALYSIS
IMAGE CLASSIFIERS
There are many different types of classifiers that have been invented over the years.
Many are complicated and available in sophisticated image processing software. In GIS,
the more common classifiers include the Parallelepiped Classifier, and the Maximum
Likelihood Classifier. The parallelepiped classifier is a very simple, quick classifier that is
used in GIS. Because it is a fast classifier, it is commonly used in the seed tool algorithm.
However, it has several disadvantages in image classification:
1) The parallelepiped classifier uses minimum and maximums that can be rare ex-
treme values.
2) Overlapping cover class parallelepipeds commonly occur and must be dealt with
in classification decisions.
3) Pixel values that fall outside of all cover class parallelepipeds result in unclassified
pixels.
4) The parallelepiped classifier does not represent spectral correlation or class vari-
ability, which is common in remotely sensed images. For example, the following
parallelepiped was developed using two correlated spectral bands and two classes
that had different spectral variability. Because the parallelepiped classifier ignores
spectral correlation and variability, it would incorrectly classify the candidate pix-
els into Class A instead of Class B in the example below.
5) The parallelepiped classifier classifies each pixel with no estimate of confidence
in its prediction.
MAXIMUM LIKELIHOOD CLASSIFIER
The maximum likelihood classifier assumes that pixel values from a given cover class are
distributed as a bell-shaped curve centered around the mean for that cover class. For ex-
ample, imagine that you are trying to classify water bodies on an alluvial floodplain. As
your first step in the classification, you want to stratify pixels as either water or land
pixels. You develop training fields for both water and land pixels as follows:
Training Polygon Statistics:
The maximum likelihood classifier assumes that most of the land pixels will have a
value of 30 and most of the water pixels will have a value of 20. The standard deviation
is an estimate of variability and can be used with the formula for a bell-shaped curve to
compute likelihood estimates for any pixel value as follows:
Likelihood(CoverClass) = [1/( sd * sqrt( 2*pi) )] exp [ (pixelvalue mean )
2
/ (2* sd
2
) ]
Likelihood Values for a given pixel being from water or land classes:
The likelihood values can be plotted as a bell-shaped curve with the total area under
each class curve summing to 1.0.
IMAGE ANALYSIS 161
Once the likelihood values have been estimated, the GIS can look up the likelihood
value for water versus land for any pixel value. The pixel is assigned the class that has
the highest likelihood.
162 PRACTICAL GIS ANALYSIS
In this example, any pixel with a value less than 25 would have a higher likelihood
for water relative to land. And any pixel with a value greater than 25 would have a
higher likelihood for land relative to water. The likelihood values are equal at a pixel
value of 25 . . . this type of tie rarely occurs in GIS applications since the likelihoods are
computed as floating point values and the tie would rarely occur at a whole number
value.
CLASSIFICATION ACCURACY ASSESSMENT
It is relatively easy to create a land cover grid by classifying remotely sensed images with
a GIS. A thorough classification accuracy assessment is critical before accepting the clas-
sification as useful. In order to conduct a classification accuracy assessment, you need
validation data to compare with your classification predictions. Such validatation data
are often called ground truth or reference data.
REFERENCE DATA
How do you obtain reference data? One approach would be to randomly select image
pixels and then locate these pixels in the field to determine their actual cover type. There
are some problems with this approach:
1) It is often impossible to find single pixels in the field because of positional error in
rectification, and errors in field navigation.
2) By randomly selecting pixels, most sampled pixels will be from dominant cover
classes and yet the most important cover classes may be relatively rare on the land-
scape and not likely to be randomly selected.
3) Travel to randomly selected locations can be very expensive in remote areas and
this could lead to an inadequate sample size of reference data, especially for the
uncommon land cover classes.
Another approach is to use vector vegetation polygons as reference data in your accu-
racy assessment. There are some problems with this approach.
1) Vector polygons have a larger minimum mapping unit area relative to many clas-
sified images. For example, your vector polygons may have been mapped down to
a minimum mapping unit area of 1 hectare while a 30-meter pixel is .09 hectares
in area. So comparing polygons with grid cells can lead to a conservatively biased
accuracy assessment.
2) Vector polygons may be from an older source like aerial photography or maps
decades before the land cover source was taken. Changes in land cover due to the
temporal differences between the polygons and the classified image may lead to a
conservatively biased accuracy assessment.
3) There is positional error inherent in both vector polygons and in classified images.
The positional error may lead to positional differences between correctly classified
pixels and correctly classed polygons which may also lead to a conservative esti-
mate of classification accuracy.
ERROR MATRIX
The error matrix is the foundation of accuracy assessmentit compares predicted
classes with actual classes. The error matrix is typically structured as follows:
Reference Data
IMAGE ANALYSIS 163
Each diagonal cell represents the total number of correctly predicted pixels for the
cover class. For example, there were 29 correctly predicted clear water pixels. The sum
of the diagonal cells divided by the total gives the overall classification accuracy.
Overall
Classification = [ ( 29 + 30 + 25 + 21 + 24 + 20 + 26 ) / 210 ] * 100 = 83%
Accuracy
PRODUCERS AND CONSUMERS ACCURACY
For each class there are two types of accuracy: producers and consumers accuracy. Pro-
ducers accuracy is computed by looking at the predictions produced for a class and de-
termining the percentage of correct predictions. For example, for the shrub class there
were 39 pixels that were predicted to be shrub and the associated producers accuracy
for shrub is:
Shrub
Producers = (24 correct predictions/ 39 total predictions) * 100 = 62%
Accuracy
Consumers accuracy is computed by looking at the reference data for a class and de-
termining the percentage of correct predictions for these samples. For example, for the
shrub class there were 30 pixels that were actually shrub and the associated consumers
accuracy for shrub is:
Shrub
Producers = (24 correct predictions/ 30 total shrub pixel ) * 100 = 80%
Accuracy
IMAGE ANALYSIS EXERCISES
1) Match the images with their companion linear contrast stretch functions:
Image#1
164 PRACTICAL GIS ANALYSIS
Image#2
Image#3
IMAGE ANALYSIS 165
Image-_______
Image-_______
Image-_______
166 PRACTICAL GIS ANALYSIS
2) Circle the appropriate resampling method (either nearest neighbor or cubic convo-
lution) for the following two rectified images:
Nearest Neighbor or Cubic Convolution?
IMAGE ANALYSIS 167
3) You are trying to co-register a 30-meter Landsat Thematic Mapper image with a
geo-rectified digital raster graphics topographic map. Using five ground control pixels
from your 30-meter TM image, you find the following:
Using these data, develop a linear transformation model for image rectification. Fill in
the proper transformation coefficients.
UTM_X = __________ + _________(Image Column) + 0(Image Row)
UTM_Y = __________ _________(Image Row) + 0(Image Column)
4) You have the following table of ground control links for developing a linear recti-
fication model. The output image grid cell size is 25 meters. What is the RMS error of
the rectification model in image pixels?
5) A consulting group supplies you with rectified satellite imagery. The contract spec-
ifies the rectification RMS error to be less than 1 pixel. The rectified imagery has a pixel
size of 25 meters. Circle any of the following statements if they are correct:
The positional accuracy of each pixel in the rectified image is +/- 1 pixel.
Every pixel in the rectified image is within 50 meters of its true map location.
The positional accuracy of the rectified image is 25 meters.
6) You are delineating training fields using a parallelepiped seed-pixel approach. You
pick the shaded pixel as your seed pixel and the parallelepiped is based on the eight
neighboring pixels. Circle all pixels that would be included in your training field.
168 PRACTICAL GIS ANALYSIS
7) The following likelihood contours were developed using these training classes:
Using the maximum likelihood rule, classify the following image, by plotting the values:
IMAGE ANALYSIS 169
8) A consulting company conducts a 2001 image classification to locate all sawtim-
ber white spruce stands. The image pixel size is 25 meters.
The accuracy of the classification is assessed by several groups. Match the following
accuracy assessment scenarios with the most likely overall classification accuracies.
_____50% overall classification accuracy
_____90% overall classification accuracy
Group A: A guided approach was conducted in conjunction with the fieldwork for
training data collection, thereby reducing the costs of data collection. Reference data
were collected adjacent to training field areas. To minimize nonclassification errors (such
as slight positional shifts) only homogeneous polygons of at least 5 hectares were used as
reference data. The center pixel of these polygons was selected for each reference data
sample location. The photos were taken the same month and year as the satellite image
was acquired.
Group B: Using high altitude color infrared photographs (taken by NASA in 1979),
we stereoscopically interpreted cover types to polygons with a minimum mapping unit
of 2 hectares. The covertypes were transferred to the USGS 1:63,360 quadrangle series
using a digitizing tablet and stored in a vector GIS. For accuracy assessment these GIS
polygons were rasterized to correspond to the classified image. Classification accuracy
assessment was conducted on a pixel by pixel basis between the reference data and the
classified image.
9) You have an error matrix as follows. What is the overall classification accuracy?
Reference Data
10) You have an error matrix as follows. What is the Producers and Users accuracy
for the muskeg class?
Reference Data
170 PRACTICAL GIS ANALYSIS
Chapter 10
Vector Exercises
INTRODUCTION
In this chapter, you can test your understanding of GIS tools for application of problem-
solving in the vector GIS world. This world can include tools for tabular, point, line, and
polygons analysis, as well as network analysis and dynamic segmentation. The following
table summarizes these tools from the previous vector GIS chapters.
171
172 PRACTICAL GIS ANALYSIS
VECTOR EXERCISES 173
174 PRACTICAL GIS ANALYSIS
REAL ESTATE APPLICATION
You are interested in purchasing property where you can build a retreat cabin as a vaca-
tion home. You have the following polygon themes: parcels, elev_zone, vegetation. You
also have a line theme of roads. You want to find land that is privately owned so that
you can make an offer to purchase. You want to purchase an area of at least 1 hectare
and is totally above 500 feet elevation (to avoid flooding problems). You want decent
road access : at least partially within 100 meters of a road. And you want the area to
contain some aspen vegetation (which typically grows on warm sites). Fill in the follow-
ing flowchart with the appropriate vector GIS tools to solve your problem:
VECTOR EXERCISES 175
176 PRACTICAL GIS ANALYSIS
MOOSE HABITAT ANALYSIS
You are hired as the GIS guru in a moose habitat study. You have the following polygon
themes: rivers, vegetation. You also have a point theme of radio-collared moose loca-
tions from the winters of 19952001. You are to test the hypothesis that mature bull
moose spend most of their time in large riparian Salix alexensis (willow) stands when-
ever there is deep snow. The winters of 1995, 1997, and 2000 were deep snow winters
while the winters of 1996,1998,1999, and 2001 were light snow winters. The moose bi-
ologist wants the following GIS analyses:
1) For mature bull moose, compute the mean distance to riparian Salix alexensis (ri-
parian is defined as within 100 meters of a river) for light versus deep snow years.
2) Compute percentage of mature bull moose locations that were within 100 meters
of riparian Salix alexensis for light versus deep snow years.
3) For the Salix alexensis polygons that had mature bull moose locations in them,
compute the mean Salix alexensis polygon area in hectares for light versus deep
snow years.
Fill in the following flowchart with the appropriate vector GIS tools to solve your
problem.
For mature bull moose, compute the mean distance to riparian Salix alexensis (ripar-
ian is defined as within 100 meters of a river) for light versus deep snow years.
VECTOR EXERCISES 177
178 PRACTICAL GIS ANALYSIS
Compute percentage of time that mature bull moose locations were within 100 meters
of riparian Salix alexensis for light versus deep snow years.
VECTOR EXERCISES 179
180 PRACTICAL GIS ANALYSIS
For the Salix alexensis polygons that had mature bull moose locations in them,
compute the mean Salix alexensis polygon area in hectares for light versus deep snow
years.
VECTOR EXERCISES 181
FIRE HYDRANT INSPECTION APPLICATION
You have a table of county-owned fire hydrants and their mileage recorded by driving
along all county roads. The table contains information such as date of last inspection,
age of hydrant, maximum flow rate, and so on. You also have a county-wide roads line
theme and a parcels polygon theme. You also have a table of parcel owners; the table
contains parcel-IDs, owner names, addresses, and phone numbers. All hydrants in the
county are inspected on a five-year basis. Your job is to create a text file of names and
addresses of parcel owners with property within 100 meters of a fire hydrant that is
scheduled to be inspected in 2001.
Fill in the following flowchart with the appropriate vector GIS tools to solve your
problem.
FOREST FIRE APPLICATION
You have a line theme of flight lines over the center of lakes, a polygon theme of lakes,
and a point theme of the centers of five currently burning wildfires. In order for tankers
to reload water, they need a lake that is at least 3 km long. Develop a map showing all
lakes available for tanker reloading and which fire they are the closest to.
EMERGENCY PHONE APPLICATION
You are hired as a GIS guru for the facilities management department of a large uni-
versity. You have a polygon theme of buildings, a line theme of sidewalks, and a point
theme of outside emergency phones. Your job is to show the area of the campus that is
most lacking in emergency phone service. Fill in the following flowchart with the appro-
priate GIS tools to solve your problem.
182 PRACTICAL GIS ANALYSIS
Chapter 11
Grid Exercises
INTRODUCTION
In this chapter, you can test your understanding of GIS tools for application of problem-
solving in the Grid GIS world. The following table summarizes the Grid GIS tools that
were covered in Chapter 8.
183
REAL ESTATE APPLICATION
You have grids of elevation, roads, and ownership. You want to find land owned by
the Fairbanks Northstar Borough that is above 500 feet elevation (out of the ice fog), on
a easterly, southerly, or westerly slope, and within 100 meters of a road. Your elevation
values is in meters.
Fill in the following flowchart with the appropriate Grid tools:
184 PRACTICAL GIS ANALYSIS
GRID EXERCISES 185
WATERSHED APPLICATION
You have a grid of vegetation types, and a grid of watersheds. You want a table showing
the area of each watershed and the area of black spruce in each watershed. Each grid cell
is 25 by 25 meters.
Fill in the following flowchart with the appropriate grid operations to solve your GIS
problem:
WATERFOWL HABITAT ANALYSIS
You have grids of lakes. One grid is from early summer and the other grid is from late
summer. You want to create a grid of the best lakes for waterfowl habitat . . . lakes that
had at least 1 hectare of exposed mudflats in late summer due to drawdown of the lake
water level. The cell size of each grid is 10 meters. Fill in the following flowchart to solve
your GIS problem.
186 PRACTICAL GIS ANALYSIS
GROUNDWATER APPLICATION
You have a grid of well locations, gas station locations, soils, and parcel ownership that
contains the address of the owner of each parcel.
Outline what tools you would use to develop a text file of all well owners that have
wells on sandy textured soils within a kilometer of a gas station.
GRID EXERCISES 187
MOOSE HABITAT ANALYSIS
You have grids of moose locations, random locations and vegetation types. You want to
estimate the mean distance of moose and random locations to the nearest willow cell. All
grid cells are 10 meters in size.
Fill in the following flowchart with the appropriate grid operations to solve your GIS
problem:
188 PRACTICAL GIS ANALYSIS
Chapter 12
Saving Time In GIS Analysis
INTRODUCTION
In this chapter are tips on how to improve efficiency in using GIS. Although the exam-
ples are from using Arc/Info, many of these ideas can be applied to any GIS.
WORK WHILE YOU PLAY
Some GIS processes such as image rectification, spatial join, and buffering operations
take considerable processing time. You can run these as batch jobs to be excuted while
you sleep or play. In the unix environment, it is a three-step process:
1. Edit a file telling the batch processor the type of unix shell script being used. Set
your ARC environment variables by sourcing your .cshrc and then list your arc
commands. For example:
Edit rectify.batch
2. Make your batch file executable using the unix chmod command. For example:
chmod u+x rectify.batch
3. Use the unix at command to submit your batch job to start at 10 p.m.
at -f rectify.batch 10:00pm Feb 10
If you work in the windows NT or windows 2000 environment, there is a similar AT
command available for batch processing.
189
AVOID TELEPHONE TAG
We live in a digital world. E-mail instead of phoning probems to GIS customer support.
This allows you to:
1) Avoid telephone tag.
2) Avoid time-zone differences.
3) Attach sample error messages and data.
4) Store a file detailing your question and customer supports solution.
5) Document poor customer support, if you feel justified in complaining.
PAINLESS DOCUMENTATION
It is easy to talk about good documentation and how important it is. But it is human na-
ture to put off documentation until the current crisis is overand in many cases GIS
work can be crisis management where clear, thorough documentation may be put on the
back burner for a long, long time.
Documentation should never be on the back burner because it is painless! The trick is
to document as you work, through the use of documentation files. A documentation file
is a edited file where all GIS commands are entered into and then copied to the GIS com-
mand window as in the following example:
Documentation files are a good practice for the following reasons:
1) You can work on your commands while the GIS is currently executing.
Why wait for a GIS to execute a command? In the above example, arc may take several
minutes to generate the point coverage. During that time you could enter into your doc-
umentation file comments and the next commands (build gps-utm points).
2) Trivial facts that may become important later are captured.
190 PRACTICAL GIS ANALYSIS
3) Repeating operations is easy.
Repetitive commands can easily be copied and modified as a block of commands to
be pasted to the GIS command window. For example, just before you take a coffee
break, you could copy and paste the following to be excuted while you are gone . . .
SAVING TIME IN GIS ANALYSIS 191
4) How did you do that last year?
I dont remember exactly, but I can easily and efficiently E-mail you the documenta-
tion file . . .
For example, what workspace was the coverage in? Was the datum NAD27 or
NAD83? What was fuzzy tolerance? etc.
ASSUME YOUR GIS LIES
You can save yourself time by being skeptical about the results from GIS operations.
Sometimes a lack of understanding can be the source of trouble. Fortunately, GISs allow
you to visually check out the results of most analysis operations. Here are a few exam-
ples where a lack of understanding is the source of trouble, but visual checks saved the
day:
1) Moose habitat analysis. A biologist has a point theme of moose locations and a
polygon theme of a series of wildfire burns of various ages. He uses the NEAR tool
to estimate the distance to the nearest wildfire polygon for each moose observa-
tion. He then computes the mean distance to each age class burn and gets the fol-
lowing table:
It seems strange that the mean distance for the burn 10-20 years old is so much larger
than the other burns. The biologist visually checks out moose locations relative to that
burn . . . and selects the moose location with the greatest distance from the 10-20 year
old burn.
By visually checking out the results of the analysis, the biologist discovers that for
moose inside the burn, the distance values were all greater than zero and not zero as ex-
pected. The problem was easily solved once the biologist recognized that there was a
flaw in his interpretation of the analysis.
192 PRACTICAL GIS ANALYSIS
2) Groundwater well analysis. A series of shallow wells have been established to
monitor possible contamination from a chemical spill. All wells are spaced at least
10 km away from each other. For this application, the wells to be used for a spe-
cial chemical analysis have to be at least 100 meters away from any oak stand to
avoid tannins associated with oak leaves. The analyst buffers the wells by 100 me-
ters and intersects the buffered theme with a theme of nonoak polygons. The ana-
lyst then reselects polygons that represent wells that are at least 100 meters away
from any oak forest (polygons with an area equal to pi *100
2
). However, no in-
tersected buffers meet this criteria.
By visually checking out the results of the analysis, the analyst finds that there are
indeed wells that meet this criteria.
The problem was that the buffer circle is approximated by a series of arcs and thus
the area is less than pi *100
2
(31415.9265). So the analyst recognizes the problem
and easily solves it by selecting buffered wells with areas greater than or equal to
31255.
3) Vegetation Index Analysis. A common vegetation index used in remote sensing is
the Normalized Difference Vegetation Index (NDVI) which is:
(NIR Reflectance Red Reflectance) / (NIR Reflectance + Red Reflectance)
where the index ranges from 1 to +1 with values less than 0.10 representing
sparsely vegetated or unvegetated grid cells.
A remote sensing analyst wants to select all vegetated pixels prior to a classifica-
tion by selecting those pixels with an NDVI greater than 0.10. Using two grids, he
does the analysis as follows:
Grid: NDVI = (NIR_Grid Red_Grid) / (NIR_Grid + Red_Grid)
Grid: Veg_pixels = Select( NDVI, Value gt 10)
Where the NIR_Grid, Red_Grid are integer grids of percent reflectance, ranging
from 0 to 100.
Instead of boldly going straight to the Veg_pixels = Select( NDVI, Value gt 10) the
analyst should first check to make sure the NDVI values seem reasonable. He
could do this by sampling some pixels and looking at the input Red_Grid,
NIR_Grid, and output NDVI values . . .
SAVING TIME IN GIS ANALYSIS 193
100 M BUFFER
AREA = 31255.789
NON-OAK POLYGON
All the grid cells have an NDVI value of zero! The problem is that the two input grids
are integer and therefore the output grid is integer by default. Once you recognize that
this is the problem, the solution is easy to figure out. You would have to use the FLOAT
function to specify that you want a floating point calculation.
Grid: NDVI = FLOAT(NIR_Grid Red_Grid) / FLOAT(NIR_Grid + Red_Grid)
As a check we could sample some pixels to make sure the calculation was correct:
So far we have seen examples where the GIS really was not lying . . . poor under-
standing of the GIS tools was the problem. Sometimes, the GIS really does lie . . . espe-
cially with the release of new tools. As an example, the GeoProcessing Wizard in Ar-
cView 3.2 allows for operations such as clipping. It is a good idea to always test new
tools to make sure they give results that you expect. As an example, we generate two
polygon themes with 50% overlap:
The correct area of the clip of these 2 test polygons should be 100 X 50 = 5000 m
2
and the correct perimeter should be 100+100+50+50 = 300 m. We test-drive the Geo-
processing Wizard to see if it gives us the correct answer..
194 PRACTICAL GIS ANALYSIS
100 meters
100 meters
100
meters
The resulting Test_clip theme looks reasonable . . . but did the area and perimeter get
correctly computed?
SAVING TIME IN GIS ANALYSIS 195
The GIS gives the wrong Area and Perimeter! Instead of calculating the correct area
after a clipping operation, it copies the original input polygon area and perimeter!
Once you recognize the problem, you can solve it by using Avenue [Shape].Return-
Area and [Shape].ReturnLength requests with the table calculator as follows:
And the correct Area and Perimeter are calculated for the output polygons.
196 PRACTICAL GIS ANALYSIS
LESS CAN BE BETTER
GIS work is fun because it is challengingyou have to figure things out! Often frustra-
tion can be minimized and efficiency maximized if you work with a reduced spatial data
set.
For example, a Landsat Thematic Mapper scene contains over 40 million grid cells.
You could wait for 40 million grid cells to be processed before you find out you made an
error. It would be more efficient to subset a small area from the scene to debug all your
image processing methods with. In ARC/INFO Grid, you can do this with the SETWIN-
DOW command. Quickly figure out how to correctly do your analysis on a 10 row by
10 column grid. After you are successful, you can then apply the same analysis opera-
tions to your huge grid.
Another example is using digital orthos for on-screen digitizing of wildfire burn scars.
If you find your digital ortho display too slow (and you can live with 2 -meter pixels),
you could reduce a 200 mb 1-meter cell size image to a 50 mb 2-meter cell size image by
using the Grid RESAMPLE function.
In the vector environment, instead of wasting time struggling with tens of thousands
of points, lines or polygons, first use the arc RESELECT command to create a small sub-
set to work with. Once you have successfully figured out the correct analysis operations
on a few features, then apply the same operations to your theme that contains tens of
thousands of features.
GIS TABLES ISSUES
Consider keeping some of your attributes external to the GIS. You can always relate
these attributes to your feature attribute table. Why would you want to do this? First,
have you ever deleted the wrong coverage by mistake? If you had all your attributes
stored in the feature attribute table, you lost them along with the geography. When you
use external tables that can be linked, your attribute data are preserved when the cover-
age is accidentally deleted. You may have to redigitize the coverage features, but your at-
tribute table is already constructed.
Secondly, related tables can make quality assurance a much easier process. Consider a
coverage with 40,000 polygons, each of which is attributed with one of 22 soil types.
You could have the actual name of the soil type entered for each polygon. Or you could
have integer soil codes for each polygon and then use a related Look-up table where the
actual names for the soil types exist. If you discover that some polygons are mislabeled,
it is relatively simple to find and correct them, because the integer codes are short and
errors will tend to stand out when the values are listed. If you go the other route, and
choose to store the soil names in the polygon attribute table, hystic pergellic cryoquept
may be entered 27,000 times, hystic pergellic cryoqupt may be entered 110 times,
hystc pergellic cryoquept 2 times, hystic pergelic cryoquept 10 times, and hystic
pergellic c8ryoquept 4 times. When you query for polygons where soil_type = hystic
pergellic cryoquept you will miss 126 polygons that should have been selected. When
you do recognize that there are errors in the database, correcting them will be time-con-
suming because the errors are scattered among 40,000 records and the exact spelling er-
rors are unknown.
CONSIDER ABANDONING YOUR GIS
GIS is great software for spatial analysis. It can be pretty lame however for tabular
analysis, graphics, and statistical analysis beyond very simple applications. Therefore,
you should always keep your options open for using other software when you need the
tools. Consider exporting your GIS table into a real spreadsheet program if you need so-
phisticated spreadsheet operations. Export your data into a statistical package if you
need powerful statistical analysis. And if you require graphics beyond basic charts and
graphs, think about exporting your GIS data into a graphics program. Use the best suite
of tools available to you, and your work will be more fun and efficient.
DO NOT REPEAT YOURSELF
GIS work should not be a mundane process of repeating commands; there is usually a
way to minimize repetition.
SAVING TIME IN GIS ANALYSIS 197
1) Startup files. Use a startup file to automatically execute common tasks. For ex-
ample, the arc processor executes a startup file associated with each module when you
start that module. For example, the .arc startup file is executed when you start Arc,
while the .arcedit startup file is executed when you start Arcedit. By creating a .startup
file, you can avoid repeating yourself every time you start a module. Here is an exam-
ple .arcedit startup file for the Arcedit module. Note that the user does not have to
specify any of this information as these commands are automatically executed from
the .arcedit file.
/***.arcedit startup file to avoid specify session settings:
/**hardware specifications:
&terminal 9999 /***Xwindows
&display 9999 4 /***fullscreen X-windows canvas
digitizer 9100 /dev/ttyb /**calcomp 9100 digitizer to serial port b
/***set snapping tolerances :
&echo &on /***echo back the settings
nodesnap closest 50 /***snap to closest node within 50 meters
arcsnap on 50 /***snap to an arc within 50 meters
intersectarcs add /***automatically put a node when arcs intersect
weedtolerance 2 /***vertex minimum spacing is 2 meters
/***set drawing rules:
mapextent %.nrm338%/studyarea
drawenvironment arcs nodes tics ids links /***drawing rules
nodecolor node red /***color nodes red
&echo &off /***turn the echo off
/****user selects coverage to edit:
&sv cover = [getcover * -all Select a coverage to edit ]
editcoverage %cover%
&return /***done with .arcedit startup so return to arcedit
2) Use AML Repeat and Expansion Character
In ARC/INFO, the AML processor interprets the (! as the beginning of a repetitive
process. For example,
projectcopy cover /net/shemp/export/data/veg-utm newveg-utm
projectcopy cover /net/shemp/export/data/veg-utm newsoils-utm
projectcopy cover /net/shemp/export/data/veg-utm newburns-utm
Can be replaced by the following:
projectcopy cover /net/shemp/export/data/veg-utm (! newveg-utm newsoils-utm newburns-utm !)
3) Overwriting Temporary Grids
In Grid, by default you cannot overwrite an existing grid. For example:
198 PRACTICAL GIS ANALYSIS
Grid: temp = select (ndvi, Value gt 100)
GRD ERROR Unable to create the output grid: temp
Output exists
Grid: kill temp
Killed TEMP with the ARC option
Grid: temp = select (ndvi, Value gt 100)
Running...
You can tell Grid that you want to overwrite existing grids by using the VERIFY OFF
command. That way you do not have to constantly delete old grids before you write to
them:
Grid: verify off
Grid: temp = select (ndvi, Value gt 100)
Running...
INDEXING FEATURES AND ATTRIBUTES
Imagine that you just picked a gallon of blueberries and you pull your favorite cookbook
off the shelf. To find a recipe for blueberry pie, you could start at the first page of the
cookbook and sequentially search page by page through the book until you find what
you are looking for. A more efficient approach would be to look in the book index under
Blueberry. In an analogous matter, it is more efficient to search GIS features and at-
tributes if they are indexed.
Attributes can be indexed in ARC/INFO by using the INDEXITEMcommand and in
ArcView by first making the table field active, and then selecting Field Create Index.
By indexing attributes, you can speed up operations that select or relate features by at-
tribute values.
Spatial features can be indexed in ARC/INFO by using the INDEX command and in
ArcView by first making the Shape field of the attribute table active, and then selecting
Field Create Index. By indexing spatial features, you can speed up operations that re-
trieve features by spatial location such as drawing and graphical selection operations.
SAVING TIME IN GIS ANALYSIS 199
Appendix
GIS DATA MODEL EXERCISE SOLUTIONS
1) You have the following line theme:
Fill in the following information from using the DESCRIBE operation.
There are three nodes. An arc starts at a from-node and ends at a to-node.
Therefore there are 5 arcs . . . the arrows in the figure are arbitrary directions the arcs
were originally digitized as. Arc#5 has two vertices at 0,500 and 0,750 while Arc#1 has
a vertex at 1000,1000. Therefore the XMIN is 0, YMIN is 500, XMAX is 1000, and
YMAX is 1000.
201
2) You have the following points. Point 100 is a well that is 200 feet deep, on parcel
356, while point 101 is a well that is 300 feet deep on parcel 387. Fill in the correct val-
ues for the point attribute table associated with this theme.
By definition, a point has no area or perimeter so the attribute table contains zeros for
those values. The Well-ID is the user-specified identification tag while the Well# is gen-
erated by the GIS as a unique ID for each point.
3) You have the following lines. Arc#1 has a length of 100 meters.
Fill in the correct values for the arc attribute table associated with this theme.
202 PRACTICAL GIS ANALYSIS
4) You have the following polygons. These parcels have the following tax informa-
tion:
Fill in the correct values for the polygon attribute table associated with this theme.
The GIS assigns a unique ID to each polygon, including the universe polygon which is
assigned an area as negative of the summed area. The area of parcel 203 is 500 X 250 =
125000 and the perimeter is 500+250+500+250 = 1500. The area of parcel 208 is 500
X 750 = 375000 and the perimeter is 500 + 750 + 500 + 750 = 2500. The area and
perimeter of parcel 209 is the same as parcel 203. The universe polygon has an area of
the sum of the three parcel polygons (125000 + 375000 + 125000) and is assigned a
negative value to indicate that it is an artificial polygon.
APPENDIX 203
5) You have the following grid of slope classes. Build a Value Attribute Table for
your grid.
For each unique value, count the total number of grid cells and fill in the Value Attribute
Table. There are 11 cells with 0, 41 cells with 1, and 48 cells with 2 as values. As a
check we have 10 rows and 10 columns and all cells contain valid values, therefore the
total count should be 100.
204 PRACTICAL GIS ANALYSIS
6) You have the following grid of soil pH. Build a Statistics Table for your grid.
The descriptive statistics of minimum, maximum, mean, and standard deviation are
computed using all grid values:
TABULAR ANALYSIS EXERCISE SOLUTIONS
TABULAR ANALYSIS EXERCISES
1) You have an attribute table about soil polygons. You run the statistics program to
create a new table summarizing the area of soil polygons by texture class. The output
table is as follows:
Why is the Frequency for texture class 3 not displayed in the table? The value for the
attribute is larger than the output width allocated for that column (output width of 5).
How can you solve the problem so that Frequency for texture class 3 is displayed in
the table? Use ALTER to make the output width of frequency larger than 5 . . . try an
output width of 10 to accommodate values up to 9,999,999,999.
2) You have the following arc attribute table:
Which records would be selected in the following expression:
Tables: SELECT stream.aat
Tables: RESELECT Ownership in {1,3} AND Trout_count GT 200
AND means the conditions of ownership of 1 or 3, and trout_count >200 both have to
be true. Only the 5
th
record satistifies those criteria:
APPENDIX 205
Which records would be selected in the following expression:
Tables: SELECT stream.aat
Tables: RESELECT Ownership in {1,3} OR Trout_count GT 200
OR means that either one side or the other side or both sides of the expression need to
be true. Therefore all 5 records are selected since they either have an ownership of 1 or
3, or they have a trout count > 200.
3) You have a street arc attribute table containing two attributes: Speed_Limit which
is the maximum allowable speed in miles per hour and Length which is the length of
each arc in meters. You add another attribute column called Time . How would you use
calculate the time in minutes it would take to travel across each arc at the maximum
speed limit? There are 5280 feet in a mile and 3.281 feet in a meter.
You can convert Speed_Limit in miles/hour to feet/hour by multiplying by 5280 feet/ 1
mile. You can convert feet/hour to feet/minute by multiplying by 1 hour/60 minutes.
And you can convert Length in meters to feet by multiplying by 3.281 feet/ 1 meter.
CALCULATE FT_PER_MIN = Speed_Limit * 5280 * ( 1.0 / 60.0 )
CALCULATE Length_FT = Length * 3.281
CALCULATE TIME = Length_FT / FT_PER_MIN
As a check we could create a test table with one record having a speed limit of 25
miles per hour and the arc length of 1 mile. The second record could have a speed
limit of 60 miles per hour and an arc length of 10 miles. We do the calculations
and then list the table:
Our test table checks out . . . it takes 10 minutes to travel a 10 mile (52800 ft) arc at a
speed of 60 minutes per hour. And it takes 2.4 minutes at 25 miles per hour to cross a
mile-long arc (5280 feet). (Check 25 miles/hr * 2.4 minutes * 1 hour/60 minutes = 1
mile).
4) Correct the following logical expression:
RESELECT SPECIES CN KING SALMON OR CN SOCKEYE SALMON
The problem with the above expression is that the GIS does not know which at-
tribute to search to see if a record contains SOCKEYE SALMON
The correct expression is:
RESELECT SPECIES CN KING SALMON OR SPECIES CN SOCKEYE
SALMON
5) Correct the following calculation:
CALCULATE ACRES = HECTARES X 2.471
The problem is that the operator for multiply is * and not X
The correct expression is:
CALCULATE ACRES = HECTARES * 2.471
206 PRACTICAL GIS ANALYSIS
6) Correct the following calculation:
CALCULATE ACRES = AREA / 43,560
The problem is that the character, is unknown to the expression translator
The correct expression is:
CALCULATE ACRES = AREA / 43560
7) Correct the following logical expressions:
The problem is that the all records need to be selected before each record selec-
tion.
Otherwise, you would get a message like 0 Records selected
RESELECT VEGCODE = 1
CALCULATE SHADECOLOR = 27
RESELECT VEGCODE = 2
0 Records selected
CALCULATE SHADECOLOR = 35
RESELECT VEGCODE = 3
0 Records selected
CALCULATE SHADECOLOR = 67
The correct expression is:
ASELECT
RESELECT VEGCODE = 1
CALCULATE SHADECOLOR = 27
ASELECT
RESELECT VEGCODE = 2
CALCULATE SHADECOLOR = 35
ASELECT
RESELECT VEGCODE = 3
CALCULATE SHADECOLOR = 67
8) You have selected a FOREST.PAT polygon attribute table. You create a new at-
tribute called SITE_CLASS based on an existing SITE_INDEX attribute. SITE_CLASS
of 1 would be any polygon with SITE_INDEX less than 50, SITE_CLASS of 2 would be
any polygon with SITE_INDEX between 50 and 75, and SITE_CLASS of 3 would be
any polygon with a SITE_INDEX greater than 75.
APPENDIX 207
9) You have a point attribute table of waterfowl nests containing the following items:
You want to produce a table with the following information:
What would you use for frequency and summary items to generate this information?
Frequency item(s): These are the categories that the GIS summarizes by:
UNIT, SPECIES, AGECLASS
Summary item(s): These are the summed or totaled attributes: CLUTCH_SIZE
This would give you the total number of eggs for each category. Since each record
represents a nest, FREQUENCY is always an output attribute from the program and
will represent the total number of nests in each category.
208 PRACTICAL GIS ANALYSIS
10) There is a proposal to purchase some land for an experimental forest research
site. Your job is to produce a table listing the hectares and percent of area for each veg-
etation class in this area. You have a vegetation polygon attribute table and another
table of vegetation names as follows:
APPENDIX 209
Vegetation Polygon Attribute Table
Fill in the appropriate tools in the following flowchart that would produce a sorted
table of total hectares and percent of area for each vegetation class.
210 PRACTICAL GIS ANALYSIS
POINT ANALYSIS EXERCISE SOLUTIONS
1) You have a point coverage of manhole covers and a line coverage of streets. You
want to determine the street name for each manhole cover.
Fill in the following flowchart with the appropriate GIS tools to solve your problem:
APPENDIX 211
2) You have a point theme of wells to sample groundwater. You want to calculate the
mean distance for each well to all other wells. Fill in the following flowchart to solve
your problem:
212 PRACTICAL GIS ANALYSIS
3) You have a point theme of lightning strike locations and a polygon theme of ele-
vation classes. Your goal is to produce a table like the following:
Fill in the following flowchart with the appropriate GIS tools to solve your problem:
APPENDIX 213
4) You have 5 GPS survey monuments set-up in the Tanana Flats area.
Draw 5 Thiessen polygons associated with these GPS monuments:
5) You have a point coverage of bufflehead nest box locations along the Chena river.
Outline the GIS analysis tools that you would use to select all nest box points that are at
least 1 km away from any other nest box.
214 PRACTICAL GIS ANALYSIS
APPENDIX 215
6) You have a point coverage in meters of snowshoe hare locations. You want to
know how many of these locations are within 50 meters of a willow polygon . . . .
216 PRACTICAL GIS ANALYSIS
From the buffer table above, we can infer that all points had full buffers since the areas
are all the same and slightly less than pi*50
2
. Notice also that the attribute BUF_50M-
ID contains the original hare ID numbers.
How many of the ten hare locations are within 50 meters of a willow polygon?
The attribute BUF_50M-ID contains the original hare IDs. Therefore the following
hares are within 50 meters of a willow polygon: Hare#2, Hare#3, Hare#4, Hare#5, and
Hare#6 and Hare#10 . . . so a total of 6 Hares.
Notice that Hare#4 is within 50 meters of willow 101 twice . . .
7) You have a point coverage of radio-collared moose. The point attributes include
Moose-ID, sex, and age. You also have a line coverage of rivers. You want to produce a
table as follows:
Fill in the following flowchart to produce the above table:
APPENDIX 217
8) You have a point coverage of Polar Bear locations stored in longitude/latitude. You
want to create a new theme of bear locations north of 70 degrees and between 170 and
100 degrees of latitude. Fill in the following flowchart to produce this new theme.
9) You have a theme of endangered plant species locations and a theme of randomly
located points. You want to know what percent of plant points are within clay loam soil
polygons compared to the percent of random points in clay loam polygons. Fill the fol-
lowing flowchart with the appropriate GIS tools to solve this problem
218 PRACTICAL GIS ANALYSIS
In the final step, you would use DESCRIBE to determine the total number of original
points and the total number of points in clay loam polygons. Then you could calculate
the percentage of points inside the clay loam polygons using a calculator.
10) You have a theme of endangered plant species locations. You want the mean,
minimum, and maximum distance of the plant locations to the nearest clay_loam poly-
gon. Fill the following flowchart with the appropriate GIS tools to solve this problem.
APPENDIX 219
LINE ANALYSIS EXERCISE SOLUTIONS
1) You have the following line coverage of pipes. You run DISSOLVE using
Pipe_Class as the dissolve item. What will your output theme look like?
220 PRACTICAL GIS ANALYSIS
The only arcs that you can dissolve that contain Pipe_Class of 1 are arcs 5 and 6.
They get dissolved to arc# 5 which is 300 + 600 = 900 in length. The arcs that contain
Pipe_Class of 2 are arcs 8 and 13. They intersect the main pipe arcs 3 and 4 and there-
fore cannot be dissolved. The arcs that contain Pipe_Class of 3 are arcs 7, and 9-12. Of
these, arcs 9 and 10 can be dissolved together and arcs 11 and 12 can be dissolved to-
gether. And there is only one arc with Pipe_Class of 4, so it cannot be dissolved with
any other arcs.
APPENDIX 221
2) You have a line coverage of roads and a point coverage of cabin locations. You
want to find all cabins that are within 1 mile of a road. Your GIS coordinate system is in
meters.
Fill in the following flowchart to solve the problem:
Your GIS coordinate system is in meters. There are 1609.344 meters per mile, therefore
your buffer distance should be 1609.344
222 PRACTICAL GIS ANALYSIS
3) You have a line coverage of streams with an arc attribute called King_count repre-
senting the count of king salmon observed along each arc. For each stream, determine
the total density of king salmon per mile. Your GIS coordinate system is in meters.
APPENDIX 223
4) You have a line theme where each line represents the border between two vegeta-
tion types as follows:
224 PRACTICAL GIS ANALYSIS
From searching the polygon attribute table, you know there is one stand of Sweet
Gum and one stand of Pin Oak in the theme. What is the length of the border between
the Sweet Gum and Pin Oak stand? Fill in the appropriate TABLES tools to solve the
problem:
From the polygon attribute table, you can see that the Pin Oak polygon internal ID is
Vegetation#2 and the Sweet Gum polygon internal ID is Vegetation#3. Therefore, find
any arcs where the left polygon is Pin Oak (Lpoly#2) and the right polygon is Sweet
Gum (Rpoly#3) or the right polygon is Pin Oak (Rpoly#2) and the left polygon is Sweet
Gum (Lpoly#3)
5) You have a line theme of roads and a point theme of auto accidents. Each road arc
has a highway code attribute. You want to store all the accidents that occurred on the
Parks Highway (highway code= 2) as a new point theme.
Fill in the following flowchart to solve your GIS problem:
APPENDIX 225
6) You have a polygon theme of parcels and a line coverage representing a utility
right of way. Each parcel polygon has the owners phone number and address stored in
the polygon attribute table. You want to generate a text file of owner phone numbers
and addresses for all parcels within the right of way.
Fill in the following flowchart to solve your GIS problem:
226 PRACTICAL GIS ANALYSIS
7) You have a line theme of gas pipelines and a polygon theme of ownership. You
want to produce a table showing for each ownership, the total length of pipeline by pipe
diameter class and by pipe age class.
Fill in the following flowchart to solve your GIS problem:
APPENDIX 227
8) You have a line theme of contour lines with an attribute of elevation. Elevations
are recorded at contour intervals of 10 meters, ranging from 10 to 980 meters. You want
to assign a dashed line to minor contour elevations that are tens of meters (for example:
10,20,30,40, etc. ) and a solid line to major contour elevations that are hundreds of me-
ters (for example 100,200,300).
Fill in the following flowchart to solve your GIS problem:
228 PRACTICAL GIS ANALYSIS
9) You have a line coverage of streets and a point coverage of fire hydrants. You have
a line coverage of streets and a point coverage of fire hydrant locations. You want to
find all hydrants where the distance between fire hydrants is greater than 1 km and the
hydrants are within 10 meters of a street.
Fill in the following flowchart to solve the problem:
APPENDIX 229
10) A grizzly bear biologist wants you to use a line theme of streams and a point
theme of bears to produce the following three tables. The GIS coordinate system is in
meters. Your job is to figure out the total area within 1 km of each stream and the total
number of bears within these 1 km areas
Fill in the appropriate tools to solve the problem on the next page:
230 PRACTICAL GIS ANALYSIS
APPENDIX 231
NETWORK ANALYSIS EXERCISE SOLUTIONS
1) You have a report of a fire at 1721 First Avenue West. Draw the location of that
address on the following streets theme. The first step is to search the arc attribute table
and match the address components to the street arc attributes Name=First, Type=Av-
enue, Suffix= West . . .
232 PRACTICAL GIS ANALYSIS
Arc# Length F_node T_node Left_from Left_to Right_from Right_to Name Type Suffix
1 400 1 2 200 300 201 299 Main Street
2 400 2 3 302 400 301 399 Main Street
3 400 3 4 402 500 401 499 Main Street
4 400 5 1 100 2 99 1 Third Avenue
5 400 6 2 100 2 99 1 Second Avenue West
6 400 8 3 100 2 99 1 First Avenue West
7 400 11 4 100 2 99 1 Center Street
8 400 6 5 1814 1998 1813 1997 First Avenue West
9 200 7 6 1744 1812 1743 1811 First Avenue West
10 200 8 7 1692 1742 1691 1741 First Avenue West
11 100 9 8 1612 1690 1611 1689 First Avenue West
12 200 10 11 1552 1610 1551 1609 First Avenue West
13 100 11 10 1500 1550 1549 1551 First Avenue West
14 400 12 5 200 102 199 101 Third Avenue West
15 400 13 6 200 102 199 101 Second Avenue West
Arc# Length F_node T_node Left_from Left_to Right_from Right_to Name Type Suffix
6 400 8 3 100 2 99 1 First Avenue West
8 400 6 5 1814 1998 1813 1997 First Avenue West
9 200 7 6 1744 1812 1743 1811 First Avenue West
10 200 8 7 1692 1742 1691 1741 First Avenue West
11 100 9 8 1612 1690 1611 1689 First Avenue West
12 200 10 11 1552 1610 1551 1609 First Avenue West
13 100 11 10 1500 1550 1549 1551 First Avenue West
Then find the record that has the correct address range.
Then find Arc#10 and determine which side is the right side of the arc:
And then interpolate between the arc addresses to estimate the location of 1721.
1741 1691 = address range of 50 over a distance of 200.
1721 1691 / 50 = 60 percent of the arc or .60 * 200 = distance of 120 from start-
ing node.
2) Given the following network theme, determine the path that is the quickest to get
from point A to point B. First, we label the time it takes to travel across each arc from
the information in the arc attribute table
We build two tables, one of nodes that have already been processed, and one of adja-
cent nodes to process. We start at Point A, node#1:
We then pick the adjacent node with the least cumulative cost (node#4) and add that
to the processed nodes list. Then add the nodes adjacent to the last processed node to the
adjacent nodes list. . . nodes 5 and 7.
APPENDIX 233
We then pick the adjacent node with the least cumulative cost (node#2) and add that
to the processed nodes list. Then add the nodes adjacent to the last processed node to the
adjacent nodes list.
We then pick the adjacent node with the least cumulative cost (node#5) and add
that to the processed nodes list. Then add the nodes adjacent to the last processed node
to the adjacent nodes list.
We then pick the adjacent node with the least cumulative cost (node#8) and add
that to the processed nodes list. Then add the nodes adjacent to the last processed node
to the adjacent nodes list.
234 PRACTICAL GIS ANALYSIS
We then pick the adjacent node with the least cumulative cost (node#7) and add that
to the processed nodes list. Then add the nodes adjacent to the last processed node to the
adjacent nodes list.
We then pick the adjacent node with the least cumulative cost (node#3) and add that
to the processed nodes list. Then add the nodes adjacent to the last processed node to the
adjacent nodes list.
We then pick the adjacent node with the least cumulative cost (node#6) and add that
to the processed nodes list. Then add the nodes adjacent to the last processed node to the
adjacent nodes list.
APPENDIX 235
We then pick the adjacent node with the least cumulative cost (node#11) and add that
to the processed nodes list. Then add the nodes adjacent to the last processed node to the
adjacent nodes list.
We then pick the adjacent node with the least cumulative cost (node#9) and add that
to the processed nodes list. Then add the nodes adjacent to the last processed node to the
adjacent nodes list.
We then pick the adjacent node with the least cumulative cost (node#10) and add
that to the processed nodes list. Then add the nodes adjacent to the last processed node
to the adjacent nodes list. All the nodes adjacent to node 10 (nodes 7,11) have already
been processed. So we add node 12 as the final entry to the processed nodes table.
236 PRACTICAL GIS ANALYSIS
The quickest path from node 1 to node 9 is 98521 and the optimal route would
take 170 seconds.
APPENDIX 237
3) You are managing wilderness cabins on a network of backcountry ski trails. The
average skiing time per 100 meters of trail for various slopes is as follows:
Delineate service network that represents all trails that are within a half hour from the
Colorado Creek Cabin.
The first step is to compute the time in seconds it would take to travel across each arc.
For example, arc#1 would take : 40 seconds/100m * 1100 m = 440 seconds. Arc#3
would take 20 seconds/100 m * 1200 m = 240 seconds to travel from node#6 to node#5
and 120 seconds/100 m * 1200 m = 1440 seconds to travel up slope from node#5 to
node#6. Draw these times on the network figure.
238 PRACTICAL GIS ANALYSIS
The time limit is 30 minutes * 60 seconds/minute = 1800 seconds. Starting at node#7,
allocate network arcs until you reach that limit.
APPENDIX 239
DYNAMIC SEGMENTATION EXERCISE SOLUTIONS
1) You have a route system of irrigation canals as follows. Circle the first section of
route#1 and the second section of route#3. The first section of route#1 is from 2/3s of
the arc to the end of arc#1 (66.66 percent to 100 percent). The second section of
route#3 is from 0 to 50 percent of arc#9.
240 PRACTICAL GIS ANALYSIS
2) You have the following point theme of auto accidents and network of streets.
APPENDIX 241
You enter the ADDROUTEMEASURE command and specify Accidents as the name
of your point theme, Streets as the name of your route theme and route system, and
acc_events.tbl as the name of your output event table. You specify a search radius of 100
meters. Fill in your output event table:
Accident number 2 is not within 100 m of any street so it is not included in the out-
put event table. Accident#1 is closest to halfway down arc#1 route 1, measure 7.5,
Accident#7 is halfway down arc#5 route1, measure 52.5. Accident#4 is at the end of
arc#6 route 1, measure 65. Accident#3 is halfway down arc#8 route 3, measure
7.5. Accident#6 is halfway down arc#11 route 5, measure 10. And Accident #5 is
halfway down arc#12 route 6, measure 10.
acc_events.tbl
242 PRACTICAL GIS ANALYSIS
3) You have a route system of streams as follows.
APPENDIX 243
Circle the stream section that has the highest density of Sockeye Salmon.
Circle the stream section that has the highest density of King Salmon.
The first step is to add 2 columns to your event table for density of Sockeye Salmon and
density of King Salmon. Density is calculated as count / length where length is the length
of each section (To From):
The highest density for both Sockeye and King Salmon is the section from 190 to 195 on
route#2. So we circle that section.
244 PRACTICAL GIS ANALYSIS
4) You have a route system of streams as follows. What would be the output if you
run RESELECT on your streams with the following expression:
The output would be section# = 3 and would look like:
APPENDIX 245
5) You have the following route system. Draw what the output theme would look
like if you applied the ROUTEARC command to this route system.
6) You have the following route system. Draw what the output theme would look
like if you applied the SECTIONARC command to this route system.
246 PRACTICAL GIS ANALYSIS
ARC #5 ARC #6
ARC #5 ARC #6
7) You have the following bus route system:
Bus route system tables:
What would be the results if you run ROUTESTATS on your route system with the
following options?
RouteStats: ARCLENGTH
RouteStats: MEASURELENGTH
ARCLENGTH will return the total length of each route in GIS coordinate units. For
example, route 3 is composed of arc# 1,2,8,9,10 and 25 percent of arc#11. Therefore
the ARCLENGTH returned for route 3 is: 500 + 500 + 500 + 1000 + 500 + 0.25(1000)=
3,250.
MEASURELENGTH will return the total length of each route in measure units. For ex-
ample, route 1 has a total of 200 measure units.
APPENDIX 247
8) You have the following point and linear events along a street route system. Draw
the output themes that would result from using EVENTPOINT and EVENTARC.
248 PRACTICAL GIS ANALYSIS
EVENTPOINT would take the pole events and determine their correct location in the
GIS coordinate system based on the event location attribute. The output point theme
would look like the following:
APPENDIX 249
EVENTARC would take the speed limit events and determine their correct location in
the GIS coordinate system based on the event location attribute. The output line theme
would look like the following:
9) You have two event tables of King Salmon counts along a stream route. Fill in the
output table from executing OVERLAYEVENTS with these two event tables.
With the UNION option, the first step is to combine the From- and To- segments of the
section: 60 to 65, 65 to 75, 110 to 130, 165 to 175. Then tally up the counts for these
stream stretches . . .
250 PRACTICAL GIS ANALYSIS
With the INTERSECT option, the first step is to find the stretches that completely over-
lap in both event tables: 60 to 65, 110 to 130, and 165 to 175. Then tally up the counts
for these stream stretches . . .
Notice that the output may be misleading due to some stretches not perfectly overlap-
ping. For example, for 60 to 65, the table contains 1 for Count_2000. This may be true,
but the original stretch was 60 to 75 that year so the King may have been counted any-
where in that stretch. To avoid this type of possible error, it would be best to have the
From- and To- measures consistent from year to year.
10) You have the following route system and irrigation event table. Fill in the output
event table resulting from using DISSOLVEEVENTS with Pipe_Type as the dissolve
item.
APPENDIX 251
POLYGON ANALYSIS EXERCISE SOLUTIONS
1) Polygons are assumed to be pure in their content and discrete and absolute in their
boundary locations. In other words, they are assumed to be homogeneous with sharp,
distinct boundaries. This is rarely the case. Rank the following polygon themes in terms
of their likely purity and boundary sharpness.
2) You have polygon themes of a wildfire burn from 1963 and a burn from the year
2000. Fill in the following flowchart to determine the total number of hectares from the
1963 burn that burned in 2000.
252 PRACTICAL GIS ANALYSIS
3) You have a line theme of rivers and streams and a polygon theme of three water-
shed areas (1= Rock Creek Watershed, 2=Clear Creek Watershed, 3=Willow Creek Wa-
tershed). For each watershed, you want to know the density of rivers and streams in me-
ters per hectare of the watershed. Your GIS coordinate system is in meters.
Fill in the following flowchart to solve your GIS problem.
APPENDIX 253
4) You have a point theme of mallard nest locations with three age classes of nesting
mallards. You also have a polygon theme of land cover types (1= tussock 2=low shrub
3=high shrub 4=barren). Your GIS coordinate system is in meters.
You want to estimate the density of nests in each land cover type by mallard age class
as follows:
Fill in the following flowchart to solve your GIS problem.
254 PRACTICAL GIS ANALYSIS
5) You have a polygon theme of vegetation types (vegpolys) and a polygon theme of
watershed areas (watersheds). You want a table showing the area of each watershed and
the area of black spruce in each watershed.
APPENDIX 255
6) You have three polygon themes: SOIL, VEG, and TINPOLYS. SOIL contains soil
information, VEG contains vegetation information, and TINPOLYS are triangles where
the slope and aspect of each triangle has been estimated.
You are checking for possible errors in the vegetation theme. You want to find if there
are any aspen polygons that occur over poorly drained soils or on northerly facing slopes
that are over 10 percent in slope gradient. These are most likely errors in vegetation
mapping, because aspen typically grows on relatively dry, warm sites.
Fill in the following flowchart to solve your GIS problem.
256 PRACTICAL GIS ANALYSIS
7) You have three candidate routes for a new power line. The cost of building the line
varies by amount of land owned privately, by soil type to build on, and by amount
of wetlands to cross, and also total distance. You want to know this type of information
for the three routes. Fill in the following flowchart with the most appropriate GIS tools:
APPENDIX 257
8) You have a pretimber harvest polygon theme of alder clumps and a posttimber
harvest point theme of planted spruce seedlings. You want to find the mean height
growth of the spruce seedlings within 10 meters of preharvest alder, versus seedlings that
were at least 10 meters away from the preharvest alder clumps. Fill in the following
flowchart with the most appropriate GIS tools:
9) You have polygon themes of lakes. One theme is from early summer and the other
theme is from late summer. You want to create a new theme of the best lakes for water-
fowl habitat . . . lakes that had at least 1 hectare of exposed mudflats in late summer
due to drawdown of the lake water level. Fill in the following flowchart to solve your
GIS problem.
258 PRACTICAL GIS ANALYSIS
10) You have a point theme of well locations, a point theme of gas station locations,
a soils polygon theme, and a parcels polygon theme containing the address of the owner
of each parcel.
Outline what tools you would use to develop a text file of all well owners that have
wells on sand, loamy sand, or sandy loam soils within a kilometer of a gas station.
APPENDIX 259
GRID ANALYSIS EXERCISE SOUTIONS
1) You run REGIONGROUP on the following grid. Fill in the output grid:
Start with the upper left cell, and assign all cells that have a value of zero and are con-
necting as group 1, proceed to the next cell on the row and assign all cells that have a
value of 90 and are connecting as group 2, and so on. Once you have completed your
group assignments, you can build your output value attribute table.
The output grid has 36 cells, so as a check, add up your count values to make sure they
sum to 36.
2) What would the output grid contain if you use TEST with the following logical
expression: Value in {1,3,8,9}
Test returns a 1 if the cell meets the logical expression criteria, 0 if it does not. There-
fore the cells containing 1,3,8, or 9 receive 1s and the other cells receive 0s.
As a check, the total count should sum to 36 . . . and it does.
260 PRACTICAL GIS ANALYSIS
3) What would the output grid contain if you use SELECT with the following logical
expression: Value in {1,3,8,9}
SELECT returns the cell value if the cell meets the logical expression criteria; NO-
DATA if it does not. Therefore the cells not containing 1,3,8, or 9 receive NODATA.
As a check, the total of number of valid (not NODATA) cells is 15. Therefore the
sum of count should be 15 and it is (5 + 5 + 4 + 1 = 15).
APPENDIX 261
4) You have a grid of willow cells. You want to buffer all willow areas by 1 cell. Fill
the output grid with the appropriate values after the EXPAND function is executed.
BUFWILLOW = EXPAND ( WILLOW, 1, LIST, 4, 5, 6, )
5) You have the following grids output from running the COSTDISTANCE
function.
First draw the direction arrows associated with the codes in the Backlink Grid.
Then use the Backlink Grid and Accumulative Cost Grid to determine the optimal path
starting at cell (3,4). The optimal path has a cost of 786
Fill in the output grid resulting from estimating an optimal path using the COST-
PATH function:
262 PRACTICAL GIS ANALYSIS
6) You have a grid of forest types and a grid of three watersheds. You want to pro-
duce the following information.
Fill in the following flowchart with the appropriate grid tools to solve your problem.
APPENDIX 263
7) You have a 1/0 grid of all lightning strike locations in interior Alaska during last
summer. You also have a grid of elevation values in meters for the same area. Each grid
cell is 1000 by 1000 meters. You want to determine the following information:
Fill in the following flowchart with appropriate grid operations to solve your GIS
problem:
264 PRACTICAL GIS ANALYSIS
8) You have one grid (veg) that contains values of 71,72,73,74,75 for open water
and values of 63,64,65,67 for wetland. The grids have a cell size of 25 meters. Draw a
flow chart showing how you would produce a grid of wetlands cells that are on the
shorelines of open water.
Fill in the following flowchart with appropriate grid operations to solve your GIS
problem:
APPENDIX 265
9) A forester is planning potential harvest units. She wants to find large (>100 ha)
stands of closed aspen, birch, or aspen/birch forest (VEGCLASS 16-19) The SIZECLASS
of the stands should be sawtimber or pole size (3 or 2) and the stands should be at least
100 meters from open water (VEGCLASS 80-89) and at least 50 meters from marsh
areas (VEGCLASS 77-79). All grids have a cell size of 25 by 25 meters.
Fill in the following flowchart with the appropriate Grid tools to solve your GIS
problem:
266 PRACTICAL GIS ANALYSIS
10) You have grid of vegetation types (VEG), soil drainage class (DRAINAGE) and a
grid of elevation values (ELEV). You are checking for possible errors in the VEG grid.
Aspen typically grows on warm sites such as coarse, well-drained soils or warm,
southerly-facing slopes. You want to find any cells that have an aspen vegetation class
occurring on either poorly drained soils or cold slopes (slopes with an aspect from north
to east, northwest to west and a gradient greater than 10 percent).
Fill in the following flowchart with the appropriate Grid tools to solve your GIS
problem:
APPENDIX 267
IMAGE ANALYSIS EXERCISE SOLUTIONS
1) Match the images with their companion linear contrast stretch functions:
Image#1
Image#2
268 PRACTICAL GIS ANALYSIS
Image#3
Image1 since it is the lightest, and all pixels above the mean are
displayed at a video intensity of 255
APPENDIX 269
Image2 since it is the darkest, and all pixels below
the mean are displayed at a video intensity of zero
Image3 since it has the best contrast and the stretch is for most of the image pixels
270 PRACTICAL GIS ANALYSIS
2) Circle the appropriate resampling method (either nearest neighbor or cubic convo-
lution) for the following two rectified images:
Cubic Convolution Resampled Image . . .
smoothest due to weighted averaging of pixels
Nearest Neighbor Resampled Image since linear features have staircase shape.
APPENDIX 271
3) You are trying to co-register a 30-meter Landsat Thematic Mapper image with a
geo-rectified digital raster graphics topographic map. Using five ground control pixels
from your 30-meter TM image, you find the following:
Using these data, develop a linear transformation model for image rectification. Fill in
the proper transformation coefficients.
The slope of a line is the rise/run. For the X model, the slope would be rise/run or
change in UTM X relative to a change in image columns:
slope = ( 518,120518,000 ) / ( 5 1 ) = 120 / 40 = 30
Once you determine the slope of the line, you can use algebra to determine the intercept:
518,000 = ____________ + 30( 1)
518,000 30 = _________
517,970 = intercept
UTM_X = 517,970 + 30(Image Column) + 0(Image Row)
As a check, you can plug in any column value to see if you get the correct X Map
Coordinate:
517,970 + 30(2) + 0(Image Row) = 518,030
517,970 + 30(3) + 0(Image Row) = 518,060
517,970 + 30(4) + 0(Image Row) = 518,090
517,970 + 30(5) + 0(Image Row) = 518,120
For the Y model, the slope would be rise/run or change in UTM Y relative to a change
in image rows:
slope = (5,190,8805,191,000) / ( 5 1 ) = -120 / 40 = -30
Once you determine the slope of the line, you can use algebra to determine the intercept:
5,191,000 = ____________30( 1)
5,191,000 + 30 = _________
5,191,030 = intercept
UTM_Y = 5,191,03030(Image Row) + 0(Image Column)
272 PRACTICAL GIS ANALYSIS
As a check, you can plug in any column value to see if you get the correct X Map Coor-
dinate:
5,191,030 30(2) + 0(Image Column) = 5,190,970
5,191,030 30(3) + 0(Image Column) = 5,190,940
5,191,030 30(4) + 0(Image Column) = 5,190,910
5,191,030 30(5) + 0(Image Column) = 5,190,880
4) You have the following table of ground control links for developing a linear recti-
fication model. The output image grid cell size is 25 meters. What is the RMS error of
the rectification model in image pixels?
RMS error = Square Root [ (Sum of Total Errors Squared)/ Number of Links ]
= SQRT [ (26.81
2
+ 11.21
2
+ 26.03
2
+ 66.43
2
+ 45.31
2
+ 37.25
2
+ 58.60
2
+
47.80
2
) / 8 ]
= SQRT [ 15094.305 / 8 ] = 43.44 = 1.7 pixels
5) A consulting group supplies you with rectified satellite imagery. The contract spec-
ifies the rectification RMS error to be less than 1 pixel. The rectified imagery has a pixel
size of 25 meters. Circle any of the following statements if they are correct:
The positional accuracy of each pixel in the rectified image is +/- 1 pixel.
Every pixel in the rectified image is within 50 meters of its true map location.
The positional accuracy of the rectified image is 25 meters.
None of these statements are correct. The positional accuracy of an individual pixel is
unknown and can be either less than or greater than 25 meters depending upon the pixel
you choose. The model RMS error is an estimate of the expected error but the actual
distribution of errors is unknown and the RMS error is an estimate based on relatively
few ground control locations. The ground control locations may have been selected
from flat, roaded areas of the image. If this was the case, then the positional error for
pixels from mountainous areas of the image may be much greater.
APPENDIX 273
You are delineating training fields using a parallelepiped seed-pixel approach. You
pick the shaded pixel as your seed pixel and the parallelepiped is based on the eight
neighboring pixels. Circle all pixels that would be included in your training field.
274 PRACTICAL GIS ANALYSIS
7) The following likelihood contours were developed using these training classes:
Using the maximum likelihood rule, classify the following image, by plotting the values:
APPENDIX 275
8) A consulting company conducts a 2001 image classification to locate all sawtim-
ber white spruce stands. The image pixel size is 25 meters.
The accuracy of the classification is assessed by several groups. Match the following
accuracy assessment scenarios with the most likely overall classification accuracies.
_____50 percent overall classification accuracy . . . Group B since they used old refer-
ence data and the land cover may have changed between 1979 and the 2001 imagery
date. Also since they used polygons with a minimum mapping unit, many of the 0.0625
hectare pixels could have been correctly classified but incorrectly assessed as misclassi-
fied because of the large polygons used in the classification accuracy assessment. Both
the temporal difference and the minimum mapping unit area difference between the ref-
erence data and the classified image would lead to a conservative estimate of the classifi-
cation accuracy.
_____90 percent overall classification accuracy . . . Group A since they sampled ex-
clusively areas that would be easy to classify . . . homogeneous areas near training fields
and ignored the heterogeneous areas that would be difficult to classify.
Group A: A guided approach was conducted in conjunction with the fieldwork for train-
ing data collection, thereby reducing the costs of data collection. Reference data were
collected adjacent to training field areas. To minimize nonclassification errors (such as
slight positional shifts) only homogeneous polygons of at least 5 hectares were used as
reference data. The center pixel of these polygons was selected for each reference data
sample location. The photos were taken the same month and year as the satellite image
was acquired.
Group B: Using high altitude color infrared photographs (taken by NASA in 1979), we
stereoscopically interpreted cover types to polygons with a minimum mapping unit of 2
hectares. The cover types were transferred to the USGS 1:63,360 quadrangle series using
a digitizing tablet and stored in a vector GIS. For accuracy assessment these GIS poly-
gons were rasterized to correspond to the classified image. Classification accuracy as-
sessment was conducted on a pixel by pixel basis between the reference data and the
classified image.
276 PRACTICAL GIS ANALYSIS
9) You have an error matrix as follows. What is the overall classification accuracy?
10) You have an error matrix as follows. What is the Producers and Users accuracy
for the muskeg class?
Producers accuracy is computed by looking at the 26 predictions produced for a muskeg
and determining the percentage of correct predictions. Consumers accuracy is computed
by looking at the reference data of 30 true muskeg pixels and determining the percent-
age of correct predictions for these samples.
Producers Accuracy = ( 20/26 ) * 100 = 77 percent
Consumers Accuracy = ( 20 / 30 ) * 100 = 67 percent
APPENDIX 277
VECTOR EXERCISE SOLUTIONS
Real Estate Application. You are interested in purchasing property where you can build
a retreat cabin as a vacation home. You have the following polygon themes: parcels,
elev_zone, vegetation. You also have a line theme of roads. You want to find land that is
privately owned so that you can make an offer to purchase. You want to purchase an
area of at least 1 hectare, that totally is above 500 feet elevation (to avoid flooding prob-
lems). You want decent road access : at least partially within 100 meters of a road. And
you want the area to contain some aspen vegetation (which typically grows on warm
sites). Fill in the following flowchart with the appropriate vector GIS tools to solve your
problem:
278 PRACTICAL GIS ANALYSIS
Moose Habitat Analysis. You are hired as the GIS guru in a moose habitat study. You
have the following polygon themes: rivers, vegetation. You also have a point theme of
radio-collared moose locations from the winters of 1995-2001. You are to test the hy-
pothesis that mature bull moose spend most of their time in large riparian Salix alexen-
sis (willow) stands whenever there is deep snow. The winters of 1995, 1997, and 2000
were deep snow winters while the winters of 1996,1998,1999, and 2001 were light
snow winters. The moose biologist wants the following GIS analyses:
1) For mature bull moose, compute the mean distance to riparian Salix alexensis (ri-
parian is defined as within 100 meters of a river) for light versus deep snow years.
2) Compute percentage of time that mature bull moose locations that were within
100 meters of riparian Salix alexensis for light versus deep snow years.
3) For the Salix alexensis polygons that had mature bull moose locations in them,
compute the mean Salix alexensis polygon area in hectares for light versus deep
snow years.
Fill in the following flowchart with the appropriate vector GIS tools to solve your prob-
lem.
APPENDIX 279
For mature bull moose, compute the mean distance to riparian Salix alexensis (riparian
is defined as within 100 meters of a river) for light versus deep snow years.
280 PRACTICAL GIS ANALYSIS
Compute percentage of mature bull moose locations that were within 100 meters of ri-
parian Salix alexensis for light versus deep snow years.
APPENDIX 281
For the Salix alexensis polygons that had mature bull moose locations in them, com-
pute the mean Salix alexensis polygon area in hectares for light versus deep snow years.
282 PRACTICAL GIS ANALYSIS
Fire Hydrant Inspection Application. You have a table of county-owned fire hydrants
and their mileage recorded by driving along all county roads. The table contains infor-
mation such as date of last inspection, age of hydrant, maximum flow rate, and so on.
You also have a county-wide roads line theme and a parcels polygon theme. You also
have a table of parcel owners; the table contains parcel-IDs, owner names, addresses,
and phone numbers. All hydrants in the county are inspected on a five-year basis. Your
job is to create a text file of names and addresses of parcel owners with property within
100 meters of a fire hydrant that is scheduled to be inspected in 2001.
Fill in the following flowchart with the appropriate vector GIS tools to solve your
problem.
APPENDIX 283
Forest Fire Application. You have a line theme of flight lines over the center of lakes,
a polygon theme of lakes, and a point theme of the centers of five currently burning
wildfires. In order for tankers to reload water, they need a lake that is at least 3 km long.
Develop a map showing all lakes available for tanker reloading and which fire they are
the closest to.
Emergency Phone Application. You are hired as a GIS Guru for the facilities manage-
ment department of a large university. You have a polygon theme of buildings, a line
theme of sidewalks, and a point theme of outside emergency phones. Your job is to show
the area of the campus that is most lacking in emergency phone service. Fill in the fol-
lowing flowchart with the appropriate GIS tools to solve your problem.
284 PRACTICAL GIS ANALYSIS
CHAPTER 11 GRID EXERCISE SOLUTIONS
Real Estate Application. You have grids of elevation, roads, and ownership. You want to
find land owned by the Fairbanks Northstar Borough that is above 500 feet elevation
(out of the ice fog), on a easterly, southerly, or westerly slope, and within 100 meters of
a road. Your elevation values are in meters.
Fill in the following flowchart with the appropriate Grid tools:
APPENDIX 285
Watershed Application. You have a grid of vegetation types, and a grid of watersheds.
You want a table showing the area of each watershed and the area of black spruce in
each watershed. Each grid cell is 25 by 25 meters.
Fill in the following flowchart with the appropriate grid operations to solve your GIS
problem:
286 PRACTICAL GIS ANALYSIS
Waterfowl Habitat Analysis. You have grids of lakes. One grid is from early summer
and the other grid is from late summer. You want to create a grid of the best lakes for
waterfowl habitat . . . lakes that had at least 1 hectare of exposed mudflats in late sum-
mer due to drawdown of the lake water level. The cell size of each grid is 10 meters. Fill
in the following flowchart to solve your GIS problem.
APPENDIX 287
Groundwater Application. You have a grid of well locations, gas stations locations,
soils, and parcel ownership that contains the address of the owner of each parcel.
Outline what tools you would use to develop a text file of all well owners that have
wells on sandy textured soils within a kilometer of a gas station.
288 PRACTICAL GIS ANALYSIS
Moose Habitat Analysis. You have grids of moose locations, random locations, and
vegetation types. You want to estimate the mean distance of moose and random loca-
tions to the nearest willow cell. All grid cells are 10 meters in size.
Fill in the following flowchart with the appropriate grid operations to solve your GIS
problem:
APPENDIX 289
abandoning GIS 197
accuracy assessment 162, 163
address geocoding 7, 59, 601; exercise
734, 2323
ADDROUTEMEASURE 81, 82, 95, 175, 242
ADDXY 27, 173, 218
affine transformation 14951; rotation
1523; scaling 151; translation 1512
ALTER 21, 172, 205
APPEND 9, 171, 284
arc attribute tables 5, 19, 23, 44, 45, 50, 52,
74, 846, 88, 95, 96, 202, 205, 206, 220,
224, 232, 243, 245
ARC/INFO 2, 196, 198
ARCPOINT 27, 173, 284
ArcView 24, 194, 195
area analysis tools 27, 2830, 173
ASCII characters 16
ASELECT 18, 19, 172
ASPECT 1334, 183, 267, 287
attribute analysis tools 27, 335, 173
attribute tables: tools for building 1011; see
also arc...; point...; polygon...; route...;
section...; value attribute tables
auto accidents 53, 94, 225, 241
batch jobs 189
batch processor 189
bears see grizzly bears; polar bears
best path/route 6, 59, 66, 75, 2337
bilinear interpolation resampling 155
binary integer items 1517
birds 91; see also nest box location;
waterfowl habitat
BUFFER: in line analysis 43, 457, 174, 222,
225, 229, 278; in point analysis 27, 2930,
173, 214, 215, 283; in polygon analysis
103, 174, 259, 281
BUILD 10, 171
BUILDSTA 11
BUILDVAT 11
bus routes 99, 247
cabin locations 51, 222
character items 15, 17
classification accuracy 162, 169
classification of images 1559; see also image
classification
CLIP 103, 106, 174, 258
closest-facility finding 7, 59, 678
CON function 121, 1234, 183
connectors 18
consumers accuracy 1634, 277
continuous events 79
contour lines 55, 228
COPY 9, 171, 212; in tabular analysis 21,
173
COSTDISTANCE function 12731, 183, 262
COSTPATH function 1313, 183
COUNTERVERTICES 43, 45, 103, 174
cubic convolution resampling 155
cut-and-paste operations 103, 1067, 109
data models 111; exercises 1214, 2014
DESCRIBE 9, 10, 171; examples of use 181,
201, 218, 231
descriptive statistics tools 1921, 172
Dijkstra algorithm 61
DISSOLVE 434, 103, 104, 174, 2201
DISSOLVEEVENTS 84, 92, 175, 251
distance analysis tools 27, 303, 173
distance tools: in grid analysis 1267, 183
documentation files 1901
dynamic segmentation 7, 7792; events
analysis tools 8992, 175; exercises
93102, 24051; routes/sections analysis
848, 175
ELIMINATE 103, 105, 174
e-mail 190
emergency phones 182, 284
endangered plants 412, 21819
ERASE 103, 108, 174, 258
error checking 115, 141, 192, 256, 267
error matrix 163, 170, 277
errors, image rectification 1535
estate agency see real estate
EUCDISTANCE function 1267, 183, 289
event location 59, 67
event tables 81, 82, 83, 89, 90; combination
of adjacent records in 92; combination of
912; in exercises 95, 100, 101, 102, 243,
244, 248
Index
EVENTARC 84, 901, 175, 250
EVENTPOINT 84, 8990, 175, 249, 283
events 79; analysis tools for 84, 8992;
creating 803; types 7980
EVENTSOURCE 80, 175, 283
EXPAND function 126, 183; examples of use
262, 265, 266, 285, 288
exporting of tables 21
facilities 59, 67
fields 15
fire hydrants 56, 181, 229, 283
fire location 734, 2323
fire stations 678
FLOAT function 120
floating-point items 15, 17
forest 234, 245, 140, 209, 266
forest fires 182, 284
FREQUENCY 201, 172; examples of use
213, 227, 254, 255
From-measure attributes 79
from-nodes 6, 60
gas pipelines 55, 227
gas stations 118, 187, 259
generic tools for managing GIS features
910, 171
grid analysis 11934; distance tools 1267,
183; exercises 13441, 2607; grouping
tools 1246, 183; optimal path tools
12733, 183; selection tools 1214, 183;
topographic tools 1334, 183
grid arithmetic operators 1201, 183
grid data model: compared with vector model
11920; exercises 1838, 2859
grid themes 8; and DESCRIBE operation 10;
statistics tables 8, 11; value attribute tables
8, 11
grizzly bears 567, 2301
ground control points (GCPs) 149, 273
ground truth 162; see also reference data
groundwater 37, 187, 193, 212, 288
grouping tools, in grid analysis 1246, 183
hiking trails 19, 20, 45, 489, 616, 823,
108
IDENTITY 103, 110, 174
image analysis 14364; exercises 16470,
26877
image classification: accuracy assessment
162, 169, 276; supervised 1589;
unsupervised 1558
image classifiers 1602; maximum likelihood
classifier 1612; parallelepiped classifier
1601
image contrast 146
image display 1434
image enhancement 1459
image rectification 14953, 1678, 2713;
errors 1535, 168
image resampling 155
image sources 9, 143
image themes 9; and DESCRIBE operation
10
indexing: of attributes 22, 173, 199; of
features 199
INDEXITEM 22, 173, 199
INTERSECT 14; in line analysis 43, 489,
174, 222, 2257, 229, 231, 278; in point
analysis 27, 35, 173, 21315, 218, 219; in
polygon analysis 103, 110, 174, 2527,
259, 27884
irrigation applications 812, 93, 102, 240,
251
ISODATA algorithm 155
item type 15
item width 15
items 15
JOINITEM 22, 173; examples of use 211,
213, 254, 255
KILL 9, 21, 171, 172
lightning strikes 378, 138, 213, 264
likelihood contours 169, 275
line analysis 439; exercises 507, 22031;
tools 439, 174
line themes 56, 43; and DESCRIBE
operation 10
linear contrast stretch 1469
linear contrast stretch functions 146, 1646,
26870
linear events 7980; conversion to new line
theme 901
LIST 9, 11, 171; in tabular analysis 21, 172
mallard nest locations 113, 254
manhole covers 36, 211
maximum likelihood classifier 1612, 169,
275
merging, tables 212
moose habitat 312, 40, 17780, 188, 192,
217, 27982, 289
NEAR 27, 302, 173; examples of use 211,
217, 219, 280
nearest neighbor resampling 155
nest box locations 38, 214
nest locations 24, 113
network analysis 5973; and address
geocoding 59, 601; and closest-facility
finding 59, 678; exercises 736, 2329;
and optimal routing 59, 616; and resource
allocation 59, 6972
292 PRACTICAL GIS ANALYSIS
network themes 67
NODATA 120, 121, 122, 123, 133
node 6
NODEPOINT 27, 173
Normalized Difference Vegetation Index
(NDVI) 1934
NSELECT 18, 19, 172
operands 17
operators 17
optimal path tools: in grid analysis 12733,
183
optimal routing 6, 59, 616; exercise 75,
2337
output width 17
OVERLAYEVENTS 84, 912, 175, 2501
overwriting 1989
parallelepiped classifier 1601
petrol stations see gas stations
pipes 501, 55, 220, 227
point analysis 2735; area analysis tools 27,
2830, 173; attribute analysis tools 27,
335, 173; distance analysis tools 27,
303, 173; exercises 3642, 21119; point
management tools 27, 173
point attribute tables 4, 24, 29, 30, 35, 39,
202, 208
point events 79; conversion to new point
theme 8990
point themes 45, 27; and DESCRIBE
operation 10
POINTDISTANCE 27, 323, 173, 212
POINTNODE 27, 173
polar bear locations 40, 218
polygon analysis 10310; exercises 11118,
2529
polygon attribute tables 8, 25, 29, 30, 39,
46, 52, 83, 104, 197, 203, 224
polygon themes 78; boundary sharpness
rank 111, 252; and DESCRIBE operation
10; purity rank 111, 252
POLYGONEVENTS 82, 175
power lines 116, 257
prefix 60
producers accuracy 163, 277
real estate applications 68, 1756, 184, 278,
285
RECLASS function 1245, 183, 263, 264
record selection tools 1719, 172
rectification model errors 1535
reference data 1623
REGIONGROUP function 1256, 183;
examples of use 260, 266, 287
RELATE 22, 173, 211, 213, 283
relating attributes 197
RENAME 9, 171
repetition 1979
RESELECT: in line analysis 43, 478, 174,
225, 229, 231, 284; in point analysis 27,
334, 173, 214, 215, 218, 279, 283; in
polygon analysis 103, 174, 255, 256, 258,
259, 278, 279, 284; for routes and sections
845, 96, 175, 245; for table records 18,
172, 205, 206, 207
resource allocation 7, 59, 6972; exercise
76, 2389
right of way 54
rivers 44, 48, 112
roads/streets 23, 467, 48, 206, 211, 225,
241
root mean squared (RMS) error 155, 168,
273
rotation of image 1523
route attribute tables 78, 97, 98, 246
route systems 78; exercises 95100, 102,
240, 243, 24551
ROUTEARC 84, 856, 175; in exercises 97,
246
routes 78; analysis tools for 848, 175
ROUTESTATS 84, 878, 175; in exercises
99, 247
salmon 52, 95, 101, 223, 2434, 2501
scaling of image 151
section attribute tables 78, 93, 94, 97, 98,
240, 241, 246
SECTIONARC 84, 867, 175; in exercises
98, 246
sections 789; analysis tools for 848, 175
seed pixel training fields 15960, 168, 274
seedlings 109, 117, 258
SELECT function 122, 183; examples of use
205, 261, 263, 266, 287, 289
selection tools: in grid analysis 1214, 183
service areas 69, 70
service network 70, 76, 238
ski trails 76, 108, 238
sliver polygons: removal of 103, 105
SLOPE 133, 183
snowshoe hare locations 39, 21617
soil attributes 18, 22, 345, 187, 205
SPLIT 103, 107, 174
startup files 198
STATISTICS command 1920, 172; examples
of use 213, 217, 223, 253, 255, 258, 289
statistics table 8, 204; tool for building 11
streams 23, 52, 848, 956, 112, 205, 223,
2435
street route 100, 248
streets see roads/streets
stretching of image 151
suffix 60
summarizing tables 201
INDEX 293
supervised classification 1589; compared
with unsupervised classification 1589
table records: selecting 1719
tables: adding information to 21, 172;
deleting 21, 172; descriptive statistics
1921, 172; exporting 21, 173; indexing
22, 173; managing 21, 1723; merging
212, 173; modifying 21, 172;
summarizing 201, 172; viewing 21, 172,
see also attribute tables
tabular analysis 1522; exercises 225,
20510; tools 1722, 1723
TEST function 121, 183; examples of use
260, 263, 265, 266, 267, 285, 286, 288
THIESSEN 27, 28, 173, 284
Thiessen polygons 28, 38, 214
time-saving tips 18999
To-measure attributes 79
to-nodes 6, 60
topographic tools: in grid analysis 1334,
183
total model error 154
training areas/fields 158, 15960
translation of image 1512
turn tables 723
UNION 103, 110, 174
universe polygon 8, 203
unsupervised classification 1558; compared
with supervised classification 1589
UPDATE 103, 109, 174
utility right of way 54, 226
value attribute table 8, 204; tool for building
11
vector data model: compared with grid model
11920; exercises 17582, 27884; tools
1715
vector polygons 1623
vegetation 245, 523, 91, 109, 114, 137,
140, 209, 224, 255, 263; error checking
115, 256
vegetation index analysis 1934
vertices, of arc 56, 45
visual check of GIS results 1924
Voronoi polygons see Thiessen polygons
waterfowl habitat 24, 113, 118, 186, 208,
254, 258, 287
watersheds 112, 114, 137, 185, 253, 255,
263, 286
wells 2930, 118, 187, 212
wetlands 139, 265
wildfires 111, 182, 252
willow 312, 39, 136, 17780, 262
X error 154
Y error 154
294 PRACTICAL GIS ANALYSIS