Introduction to Spatial Data Resources
and Analysis for research in Urban Design and Planning
2-day workshop notes
27-28 September 2004
Introductions – please put on
a name tag to help me know who you are
Background and Objectives:
1.
What is your
experience with computers?
2.
What is your
experience with GIS?
3.
What are you
expectations for this workshop?
4.
What do you hope
to be able to do by tomorrow afternoon?
Workshop Outline
Monday (8-4 with 1 hr break for lunch)
8-9: Lecture
introduction to GIS
9-11: Introduction
to ArcGIS 9.0 GIS software and “Exploring ArcCatalog and ArcMap” example
11-12: Representing
the geometry of spatial phenomena (lecture)
12-1: Lunch
break, questions
1-3: ArcTutorial
example: basics of geoprocessing
3-4: Questions
and ideas for day 2
Tuesday: (9-3:30 with 1 hr break for lunch)
9-10: Web resources, Downloading and understanding data
10-11: Vector Analysis –
types of analyses
11-12: Hands-on example 1.
12-1: Lunch Break
1-3: Hands-on example 2 or 3
or “user suggestion”
I.
Introduction to Geographic Information Systems and spatial analysis (
A.
What is spatial analysis? Definitions
- Process for looking at spatial patterns in
your data and relationships between features.
- The actual methods used can be simple (making
a map) or complex, involving multiple processing steps and combining may
data layers to mimic the real world.
What is GIS?
Three parts:
1. Database – attribute data
2. Map or spatial information
3. A way to link 1 and 2
A GIS is a toolbox – a set of
tools for analyzing spatial data
GIS is “a powerful set of
tools for storing and retrieving at will, transforming and displaying spatial
data from the real world for a particular set of purposes (Burrough 1986, p. 6)
– “toolbox definition”
GIS are “automated systems
for the capture, storage, retrieval, analysis, and display of spatial data”
(Clarke 1995, p. 13) – “process definition”
GIS is “a special case of information
systems where the database consists of observations on spatially distributed
features, activities or events, which are definable in space as points, lines,
or areas. A geographic information system manipulates data about these points,
lines, and areas to retrieve data for ad hoc queries and analyses” (Dueker
1979, p 106) – “information system definition”
Examples of spatial analysis
relevant to Urban Ecology and Planning
1. Determining the frequency of patches of remnant forest
at various distances from downtown
2. Determining the change in human density at different
distances from major town centers
3. Calculating the amount of water that flows through a
drainage basin
4. Identifying parcels that are within a specified
distance from parks
5. Calculating the total length of streams in a watershed
6. Others ?
Several examples of analyses that are not
necessarily spatial
1. Calculating the number of parcels with more than 200
people in them
2. Calculating the density of people in census blocks (provided
you know the area of each census block)
3. Others ?
What is "Image Processing"
or "Remote Sensing" and how is it different from GIS?
B.
Introduction to ArcGIS 9.0 software to do GIS
ArcGIS 9: (Getting Started with ArcGIS workbook: Chapter 1)
1. ArcMap - mapping, basic spatial analysis commands,
similar to ArcView
2. ArcCatalog - similar to Windows Explorer, but the
difference is IMPORTANT
3. ArcToolBox - access to vector commands, topology
tools, database tools
4. ArcInfo / command line - access to ALL the commands
for analysis
Example differences in viewing directories
and files with Windows Explorer and with ArcCatalog
|
|
|
|
DEMO:
Exploring ArcCatalog and ArcMap
Note: For the following demo, you will need
to first do the following steps:
1. Copy the tutorial data to your local drive for use
today – copy the entire udp-gis-intro
directory to your local (c: or d: ) drive under the workspace directory
2. Demo: Running through parts of Chapter 2 in Getting
Started with ArcGIS workbook: pages 17-34 in workbook (pg 23 in the pdf
file) – open the pdf file and substitute the root directory that you just
copied when told to access data in the workbook.
3. After doing the demo, Compare for yourself the
difference in views that Windows Explorer and ArcCatalog give of the Tongass
Directory (under D:\udp-gis-intro\ArcTutor\Toolbox\Data\Tongass\)
II. Representing the geometry of spatial
phenomena (10-11 lecture)
A. Abstracting real-world
entities: representing the geometry of spatial phenomena
Coordinate systems and Projections (Notes and Images taken from "Understanding Map Projections" by ESRI)
Geographic coordinate system - uses a 3-D spherical surface to define locations on the earth. Each point on the earth has a longitude and latitude which are angles measured from the earth's center to the point on the earth's surface.
Sphere and Spheroids - the shape and size of a geographic coordinate system's surface is defined by the sphere or spheroid used to represent it. In general, the data we will be using will use speroids (ellipseoids) to represent the shape of the earth.
Datums - the datum defines the position of the spheroid relative to the center of the earth, providing a frame of reference for measuring locations on the surface of the earth.
Data present you will download will likely be in WGS 1984, NAD 1927, or NAD 1983 Datums. The NAD (North American Datum) datums are usually preferred for spatial data of North America. While the generally preferred datum is NAD 1983 (with the GRS 1980 spheroid) because of its accuracy, many datasets will be in NAD 1927 and require reprojection into NAD 1983.
- Some data may be in State Plane Coordinate System
(SPCS) which divides each state in a series of numbered sections.
- Raw GPS data is reported in the WGS 1984 coordinate system, but because of the similarity of WGS 1984 and NAD 1983, the data are directly comparable.
Projected Coordinate Systems - are defined on a flat, 2-D surface, and therefore has constant lengths, angles, and areas across the two dimensions. A projected coordinate system is always based on a geographic coordinate system that is based on a sphere or spheroid. Locations are identified by x,y coordinate pairs on a grid.
To translate a 3-D surface to a 2-D plane, you must transform the 3-D surface (= map projection). Representing the earth's surface in 2-D causes distortions of shape, area, distance, or direction of the data. Different projections cause different types of distortions.
Conformal Projections - maintain local shape
Equal area projections - maintain area of local features
Equidistant projections - preserve distances between certain points
True-direction projections (azimuthal) - direction/asmuths are correct
Creating a projection starts by creating a point of contact with the 3-D object called a point or line of tangency. Projections can be Conic, Cylindrical, or Planar
The most commonly used map projection for our purposes is the Universal Transverse Mercator (UTM). It is a conformal projection (shapes accurate) that does not maintain true directions, based on a cylindar tangent to a meridian where the globe has been divided into 60 zones, each spanning 6 degrees of longitude. Specific properties - accurate representation of small shapes, minimal distortion of large shapes at edges of zones, minimal distortion of area within zone, local angles are true, minor distortion of lateral distance.
Things to remember:
a. All data layers used MUST BE in the same map projection and coordinate system so that features in each layer will be comparable spatially to do any spatial analysis that involves more than one layer.
b. Programs we will be using may allow DISPLAY of data in different projections/coordinate systems, but on-the-fly reprojection for analysis is NOT recommended.
c. Data from several sources need to be checked to be sure they are all in a common projection and coordinate system and reprojected to a common system before analysis proceeds.
Examples: opening files, determining metadata, projection, etc.
B. Ways of representing spatial data: Raster and Vector
(Also refer to
Chapter 3 in Getting
Started with ArcGIS) – (
The
raster model represents features as a matrix of cells in
continuous space
Each
cell (or pixel) is a square that represents a specific portion of an area
·
All cells in a raster must be the same size
·
Cells are arranged in rows and columns, producing an x,y Cartesian
plane, with each cell having a unique x,y value
·
All locations are covered by the matrix
·
Each layer represents one attribute - attributes are tied to each cell
rather than to an area
· Cells with the same value belong to the same zone which do not have to be spatially contiguous (as with most vector features)
· Both integer and floating-point values are supported - continuous data can be represented as either types; categorical data as integer only
Rasters can represent continuous values such as Elevation
Rasters can also represent discrete values such as buildings, parking lots, and roads, or soil types.
Categorical rasters can also store additional attribute data
Most
analysis occurs by combining the layers to create new layers with new cell
values
The
cell size used for analysis with affect the results of the analysis
Potential Uses:
Inherent Limitations:
Types of raster data
1.
satellite imagery
2.
scanned aerial
photographs
3.
digital
orthophotos
4.
grids - ArcGIS
raster files
Vector (Vector data are stored in Coverages, Geodatabases, and Shapefiles in ArcGIS)
Types of Vector data (data models):
The
vector model represents features as x,y locations – points
These
locations are connected together to make arcs (lines) and polygons (areas)
Points represent single locations
and are
Lines (arcs) represent a series
of coordinate pairs
Areas are represented as closed
polygons
Coverage – the traditional
(ArcInfo, ArcGIS) format for complex geoprocessing, building high-quality
geographic datasets, and sophisticated spatial analysis. Primary features are
label points, arcs, and polygons. Composite features (Routes - complex lines and Regions - overlapping polygons)
are built from primary feature types.
Shapefile
(ArcView) – a vector data storage format for storing the location, shape, and
attributes of geographic features, stored in a set of related files and
contains one feature class, topology is not inherently stored.
Geodatabases
– a geographic data model that represents real-world geographic features as
objects in an object-relational database, features are stored as rows in a
table, geometry is stored in a shape field. Objects may have custom behavior.
Potential Uses of Vector Data
1. Representing linear objects - roads, streams, fault
lines
2. Representing area - lakes, cities, census blocks
3. Representing overlapping areas - the overlapping
habitat within the home range of multiple crows ("Regions")
4. Networks - transportation networks
("Routes")
Advantages
1. can more accurately represent some spatial objects
(e.g., stream lines, roads)
Inherent Limitations
1. slower processing than raster database
2. difficult to represent "soft" or gradual
edges
- The branch of geometry
that deals with the properties of a figure that remain unchanged even when the
figure is bent, stretched, or otherwise distorted.
-
Defines the spatial relationship between geographic features
-
Topological relationships:
Adjacency - implies whether two polygon features are adjacent to one another
Containment - indicates whether a single feature is contained within the boundaries of a polygon
Connectivity - indicates whether two line segments are connected
GISs are used for spatial analysis because
they can integrate all three elements of spatial information in a logically
consistent manner
Tabular: Tabular manipulation of attribute data
Working
with data tables is an important part of GIS analysis.
RDBMS: relational database management
system – a type of database in which the data is organized across several
tables. Tables are associated with each other through common fields. Data items
can be recombined from different files.
1.
These tables can contain the attribute data of the features, or derived
summary statistics.
2.
Tabular data can
be generated within a GIS and then exported and opened in a spreadsheet (Excel)
or statistics software package (SPSS) for further non-spatial analyses
Types of features
1. Discrete
a. For discrete locations and
lines, the actual location can be pinpointed
b. At any given spot the
feature is present or it is not present
c. Streams, tree locations,
bear dens
2. Continuous
a. Continuous phenomena can be
found or measured anywhere, there are no gaps
b. A value can be determined
at any given location
c. Precipitation, temperature,
elevation, slope, water depth
3. Features summarized by area
a. Summarized data represents
counts or density of individual features within area boundaries
b. Density of trees in a
stand, population size within a region, number of businesses within a Zip Code
a. The density measure applies to the entire summarized
area, even if the individual (trees, people) are clustered in certain
sub-sections of the area
c. GIS allows you find the
spatial coincidence of areas with features to perform summaries
b. Unless these are standardized to a density per unit
area, features summarized by area cannot be directly compared to other features
summarized by a different area (e.g., it does not make sense to compare the
density of people in Seattle with the average income of people in all of King
County to make a statement about the relationship between density and income
for all of King county)
Representation of
geographic features
1. Discrete features (e.g., points and lines) and data
summarized by area are usually represented by vectors, but can be converted to
raster when combined with other raster layers.
2. Continuous categories
(categorical) are represented by either vectors (e.g., Isopleth maps) or
rasters.
3. Continuous numeric values
(e.g., elevation) are represented by rasters.
Hands-on Example: Chapters
5 and 6 (Assembling the database and Preparing data for analysis) from “Getting
Started with ArcGIS” (1-3).
Review of Objectives and
ideas for Tuesday’s session
Tuesday, September 28th
I.
Planning a project and Data Available on the Web (
Planning a Spatial Analysis
When
performing an analysis you go through 5 major steps:
ii.
1. Framing the question
1.
How will the results be used
2.
Who will use the results
3.
Understanding your data
4.
What type of data, features, and attributes are available
iii.
2. What will need to be generated
iv.
3. Choosing a method(s)
1.
Many different methods exist
2.
Some methods are more approximate than others
v.
4. Processing the data
1.
Many steps required for most analyses
2.
A flow chart of the steps required and the expected results from each
step is helpful in organizing an analysis
vi.
5. Examining the results
1.
Results can be displayed as a map, a table, or a chart.
2.
Looking at the results is important to determine if the information is
valid or useful
3.
Often the analysis will need to be rerun with slightly different
methods to achieve the desired results
Web sites Sites containing spatial data
http://ads.ahds.ac.uk/project/goodguides/gis/ -
The GIS Guide to Good Practice - advice on conducting GIS analysis
http://wagda.lib.washington.edu/
WAGDA -
http://wa-node.gis.washington.edu/
http://www.metrokc.gov/gis/index.htm
http://dnr.metrokc.gov/topics/map/carto.htm">King
Resource Maps</a></p>
<p><a href="http://www.metrokc.gov/assessor/">King County Assessors Office</a></p>
<p><a href="http://www.ecy.wa.gov/services/gis/index.html">Washington Department
of Ecology GIS data</a></p>
<p><a href="http://www.seattleurbannature.org/">Seattle Urban Nature Project</a></p>
<p><a href="http://www.fish.washington.edu/naturemapping/">Seattle NatureMapping</a></p>
<p><a href="http://www.dnr.wa.gov/">Washington Department of Natural Resources</a></p>
How to connect to