Introduction to GIS concepts and issues

Overview of GIS operations

Spatial data are described by their spatial characteristics and their attribute characteristics –Spatial characteristics

Feature type (point, line, polygon, surface)
Area, extent, length, position –Attribute characteristics

Attributes or variables describing an entity

# households in a county (polygon)
% organic matter for a soil sample (point)
DBH of a single tree (point)

Critically important to document attributes of spatial data in addition to the spatial characteristics

Common to document scale at which a vector dataset was collected
As important to identify definition used for a “household” in reporting # households in a county and so on

Spatial features are tied to attribute records through a identifier field (key field, id field, unique identifier) that uniquely identifies each object. This is simply a common field between the spatial data and the attribute data records linking them together.

Spatial Data Input - Database Construction

Digitizing - manual construction of digital features from a hard copy map
Scanning - heads up digitizing, automatic feature extraction
Data import/export
Data formats (shapefiles, coverages, geodatabases, .img files, .sid files, DEM...)

Attribute Data Management

Spatial Data vs. Attribute Data
GIS software can interface with external databases or use internal databases

Data Display

Screen and hardcopy - WYSIWYG, sometimes...
Black and White printers in lab
Color printer at central STC station
Large format color plotter at Information Commons and Geography/Map Library

Data Exploration

Spatial Data Exploration (SDE) and Exploratory Data Exploration (ESDE)

Examining a spatial dataset looking for patterns and relationships that can lead to hypothesis generation and testing

Data Analysis

Data Integration

Is soil contamination greater near rivers/streams that are close to industrial areas than rivers/streams that are more isolated?

Are trees affected by Sudden Oak Death spatially clustered (SOD map viewer)?

How is the patchiness of a forest related to species distributions?

Do areas characterized as urban sprawl require greater per capita emergency response resources than other areas?

Examples of GIS Applications

Crime Mapping - Department of Justice- National Institution of Justice

Spatial Modeling

Data Exploration

Prediction/Forecasting

Scenario Testing

Data in ArcGIS are associated through the use of a spatial reference system. A spatial reference system refers to the map projection, datum and coordinate system used to represent spatial locations of features in a dataset. The projection is the representation of the earth's surface used in the representation of features in the GIS, the coordinate system is the method of applying coordinate locations within the area of interest and the datum is the origin point for the coordinate system. More description of these terms and formal definitions will follow in the next section of the course.

Structure of a Geographic Information System

Common GIS software packages

Introduction to ArcGIS

Data in a GIS are stored in a specific data structure (also sometimes called data model) and data format

A Data Structure is a representational system used to represent spatial data. The most common data structures are the vector and raster systems

A Data Format is an implementation of a Data Structure. Examples of a data format include 'coverages', 'shapefiles' and 'grids'.

A fundamental difference between the vector and raster data structures is the ability to represent data across space. Some features in a landscape are discrete objects such as a bus stop location, a road feature or a polygon representing a park boundary. Other spatial features are continuous rather than discrete. That is they are more gradients in the landscape than isolated objects. Examples include surface slope (topography), soil characteristics (i.e. % organic matter) and population density. Note the difference between representing population as discrete objects (i.e. household locations, county demographic data) vs. a gradient (population density). Population density data is generated from discrete values through spatial interpolation that assigns values to a spatial area between discrete locations to produce spatially continuous data.

In general, raster data structures are well suited to the representation of gradients and vector data structures are not (exceptions include isolines or contour line data that are a means of representing gradients through a vector data structure). The vector data structure is useful for representing linear features (or objects that lend themselves to linear representations (rivers?)) and discrete objects like point locations for cities.

Data formats and ArcGIS

Shapefile - legacy ESRI (ArcView 3.x) vector format
Coverage - legacy ESRI (ArcInfo 7.x) vector format
Geodatabase - most recent (ArcGIS 8.x onward) vector format
Grid - raster data format (support level hasn't changed much over the years)

Topology

One of the most important concepts in a GIS - Broadly defined as the relationship between entities

Adjacency (polygons)
Containment (e.g. points in polygons)
Connectivity (lines)

Critical because of the need to explore spatial relationships between features in the landscape

Which parcels are adjacent to which
Which roads are connected to which
Which wells fall within a certain municipal boundary
Which streams are in a watershed

Some GIS data structures allow for topological relationships and some do not

Major differences between data formats supported in ArcGIS:

1) Shapefiles are not topological
2) Coverages are topological (but you can't edit coverages with ArcGIS 8.3!!!)
3) Geodatabases are now topological (but with ArcGIS 8.1 and 8.2 they weren't)
4) ArcTools functions primarily work with coverages and shapefiles
5) ArcMap functions work with all formats but some functions are limited to shapefiles and geodatabases (i.e. less support for coverages)
6) Coverages use INFO as the database engine but can link to some SQL compliment DBS
7) Geodatabases use MS Access as the database engine but can also link to some SQL compliant DBS (e.g. Oracle, SYBASE...)
8) Geodatabases are part of an object oriented database structure - point features and polygon features can all exist in the same dataset and be related to each other through spatial relation rules (i.e. accident incidents that are point locations are linked in the DB to locations on roads that are line features; well locations in a certain watershed are linked to that watershed through a spatial relation (containment))

Organization of ArcGIS, ArcView and ArcEditor

Main Modules in ArcGIS

ArcCatalog - data management (copying, moving), data documentation
ArcMap - data visualization, editing, some analysis, some conversion
ArcTools - data conversion, analysis

Important terms: topology, topography, spatial reference system, map projection, attribute data, spatial data, data structure, data format, spatially continuous data, gradient, discrete data, coverage, shapefile, geodatabase, Grid

Small note: Don't confuse topology and topography -

Topology - relations between objects, in GIS refers to adjacency, containment and connectivity

Topography - surface slope (i.e. ridges, valleys...)

GIS Glossary in ArcGIS Desktop Help

Active help in ArcGIS (ArcMap, ArcCatalog)

Logistics

1) Starting ArcMap, ArcCatalog

2) Explore file structure of coverages, geodatabases and shapefiles

Notes re: *.mxd files and spatial data files

3) When working on labs

4) A note about data management - NEVER USE WINDOWS EXPLORER TO COPY ESRI BASED SPATIAL DATA FILES (Coverages, Shapefiles, Grids) FROM ONE LOCATION TO ANOTHER, ALWAYS USE ArcCatalog. It's possible to do so, but the potential for missing an important file is substantial enough that it's best just to use ArcCatalog until you get comfortable with the file structure.