Raster Based GIS Analysis. What is a RASTER GIS? 4 A graphic representation of features and attributes 4 Often looks more like an image than a map 4.

Raster Based GISAnalysis

Raster Based GISAnalysis

What is a RASTER GIS?

A graphic representation of features and attributes

Often looks more like an image than a map Made of grid cells which contain

information Cells can be any size but are often limited

by several criteria

While feature themes use Coordinates and lines (vectors) to represent geographic features

Raster Grid themes use cells to represent geographic

features

Raster GIS represents elements of the real world for analysis:

Vector GIS

We have already used a vector based GIS– It is superb at drawing maps– It is excellent for managing attribute data– It is very efficient for performing calculation

along routes and for areas with matching geographies

Points, Lines and Areas represent entities– Points - City, Tree– Lines - River, Road– Areas - Forest, Lake

Analysis includes:– Buffering– Intersecting– Network Analysis (shortest path)

In A Vector GIS...

Raster GIS

Very different from the Vector GIS– Uses more images and grids– Stores data differently– Uses square grids to hold data– Uses more “codes” to interpret data– Uses much more processing power– Can generate many more mathematical outputs

Raster Themes Include:

Images Scanned maps, Air photos, Satellite images

GridsDEM, Surfaces

In A Raster Based GIS...

Grids represent entities– Grids made of cells– Value applied to cell

Analysis Includes:– Buffering/Proximity– Reclassification– Hill-shade/Slope– Interpolation– Surface Calculation

In A Raster GIS...

•Map Algebra•Hill Shades

•Slope•Aspect

•Raster Modeling•Raster and Vector Integration

•Raster to polygon conversion•Contour Generation•Surface Interpolation from point data

Tools for Raster Data Analysis...

How does it work?

A imaginary grid is placed over an area Each cell in the grid is given a numeric

code for the attribute that is dominant in that cell

Each cell can be... – numeric– qualitative– a feature identifier

A Vector to Raster Comparison...

Vector Model

Raster Model

Vector GIS Manages Discrete Features

•Distinct boundaries•Stored as integer values•Land use, zoning,

vegetation, lakes, roads, rivers

A Raster GIS Manages Continuous Data

•Stores data as floating point values•Examples: Elevation, noise pollution, rainfall, slope and temperature

The Raster Grid

Raster grids are always square but can be displayed as other shapes

Each group of cells can be considered a layer Various layers can be overlain on each other Calculations can be done between layers if

the grids line up.

Row00 11 22 33

332211

00

Column

Cell (2,3)

•Grid themes are an organized matrix of cells•Cells are organized into rows and columns•Rows and columns have an index position number •Top left cell is at the 0,0 position

•Cells Store a numeric value•Numeric values are codes that represent geographic features

1-Wetlands2-Water3-Forest4-Recreation

•Cells with the same value make up a zone

Raster Grid Structure

Geographic Analysis

Raster systems are superb at analyzing some geography problems such as...– Data that changes continuous over a surface

• elevation• vegetation• rainfall• temperature, etc.

– There can be used to create new information such as erosion potential.

Questions and answers about something geographic– Leads to decisions that affect people

– Involves budgeting money and resources

Spatial query examples – Show me the pine trees

– Select the freeway

Spatial analysis examples – Which trees are ready to harvest?

– Identify some corridors for the new freeway

Raster GIS Analysis

What is GIS Analysis– From simple to

Complex– Simple - Sometimes by

just making a map you are doing analysis

– Complex - Sometimes we use many layers of data to simulate real world events

Data AnalysisWhat can we map using GIS (Spatial Analysis Tools)?

Map where things are- Location AnalysisMap the most and least- Quantitative AnalysisMap concentrations- DensityMap what is inside- MonitoringMap what is nearby- RelationshipsMap Change- Prediction

What are the

effects of

Global

Warming?

Land cover and temperature relationships are made clear when the data are seen at once.

Will the new

building spoil

the Park

Scenery?

Park planners use GIS to determine if a new Visitor Centre can be seen from the peak.

How can I

reduce my

fertilizer Input?

GIS is used in making crop management decisions to maximize yields and minimize fertilizer input.

Managing Raster Data

Raster systems are data intensive– they must record data at every cell even if

nothing is present– use large amounts of hard drive space

Raster systems have the advantage of being uniformly defined.– This means that very powerful analytical tools

can be used with a raster system.

Summary

Raster systems use grids to store information

Images are raster grids Images can be given attributes by pixel Grid can be analyzed for various purposes Raster data can be overlaid with other grids

for greater analysis power.

VECTOR! Vector Data Represented by point, line and polygon. Relatively small file size (small data volume) Excellent representation of networks. A large no. of attributes can be attached, hence

more information intensive and a number of thematic maps can be prepared from a single layer.

Features are more detailed & accurate. Creating, cleaning and updating data is more time

and labour consuming.

Vector con’t Topology-based analysis & operations are easier to

perform (like network analysis etc.). Can not represent continuous values like land use,

elevation etc very well. Assigning projection and transformations are less

time taking and consumes less memory of the computer system.

Topology makes data structure complex.

RASTER! Points, line & polygons everything in the form of

Pixels. Large file size. Networks are not so well represented. Only one pixel

value represents each grid cell. Generalization of features (like boundaries) hence

accuracy may decrease. Simulations and modeling is easier (spatial analysis,

terrain modeling etc.).

Raster con’t Maintaining is easier. Excellent for representing data containing

continuous values (like land use, elevation etc.) Coordinate-system transformations take more time

and consume a lot of memory. Grid cells or pixel makes simpler data structure.