Вы находитесь на странице: 1из 25

9/25/2013

Introduction to GIS
BIS342 Fall 2013
Santiago Lopez, PhD
cslopez@uw.edu slopez@uwb.edu

Outline
Self-enroll in the course Blackboard http://www.uwb.edu/learningtech/help/how-to/bb-student. Create an account (for new users) http://bb.uwb.edu Log into BB with your usrname and pwd and click on the Courses tab. Search for this course: BIS342AIntro to GIS, Fall 2012 Click on enroll Class structure (syllabus) GIS: Concepts and Notions

9/25/2013

GIS Fundamentals and Applications

Geographic Representation Scales of Representation Conceptualizations

GEO-SPATIAL TECHNIQUES

Geodesy

GPS
Surveying

Cartography

GIS

Database management

Photogrammetry

Geo Statistics and Spatial Statistics

Remote Sensing

9/25/2013

What is GIS?
Geographic Reality Location, scale (both spatial and temporal), and other attributes Information Data and meaning of data Collection of facts, knowledge, personal or collective values Systems Computers and methods, organization. Physical and conceptual entities

GIS Notions
GIS is a set of tools GIS is an information system GIS is an approach to science GIS is a business GIS is a social activity GIS is a course of study
http://www.kingston.ac.uk/undergraduate-course/geographical-information-systems-2011/

9/25/2013

Six (?) Parts of GIS


Data acquisition and management Digitization, storage, retrieval Software Codes and scripts People Users, developers Hardware Computers, scanners, plotters Procedures Analysis, manipulation, product generation The internet the digital network

Geographic Information Systems and Science


Main Purpose Propose geographic questions Provide answers to geographic questions through maps, tables, reports, creation of knowledge

9/25/2013

Geospatial Questions
Where is good place to build a new health clinic in Seattle? How has land cover in the state of Washington changed over time? What is the amount of deforested land in the Amazon region? Where are the highest concentrations of Latin American immigrants in the western United States? What areas around Seattle are susceptible to flooding and landslides?

Geospatial Data: Spatial is Special


Geographic Information associated in some way with location. Human interaction with place Earth-based

9/25/2013

Geospatial Data: Spatial is Special


Spatial Location, size, direction, proximity, shape, pattern, distribution Spatial attributes of the earth or any place

Geospatial Data: Spatial is Special


Georegistered Associated with a coordinate system

9/25/2013

Geographic Entities
Representation (in digital form binary system) Geographic entities or features Social Data Person House Block City Country Continent Environmental/Physical Data Well Lake Wetland system Sea Ocean

Scales of Representation
What is the length (R) of the shoreline of Puget Sound from Fort Lawton to Alki Point, Seattle?

1 mile interval R=11.46 miles

9/25/2013

Scales of Representation
What is the length (R) of the shoreline of Puget Sound from Fort Lawton to Alki Point, Seattle?

0.5 mile interval 11.65 miles

Scales of Representation
What is the length (R) of the shoreline of Puget Sound from Fort Lawton to Alki Point, Seattle?

0.25 mile interval 12.35 miles

9/25/2013

Scales of Representation
Answer: The exact length of the shoreline is . An approximation or representation of length (or any other geometric property) is scale-dependent.

Common spatial data models


Spatial data models begin with a conceptualization A view of real world entities Two main conceptualizations: Vector data model Raster data model

9/25/2013

Vector data model


Defines discrete objects
Uses discrete elements such as points, lines, and polygons to represent real geographic entities

Representing Geographic Entities Vector Model


Points (x, y) Lines (connected points) Polygons (connected lines)

10

9/25/2013

Non-topological Vector Models


Early vector data models that were originally developed to organize and manipulate line data. CAD formats (DGN, DWG, DXF) Spaghetti model

Non-Topological Vector Model


Pros Simple data structure Easy to create and store Rapid display and retrieval Good for cartographic representation were detail is needed Cons Slower analysis Redundant (e.g. duplicate boundaries) Inefficient to represent spatial relationships Produce slivers

11

9/25/2013

Slivers
Extra polygons created when two boundaries should be the same but arent

Topological Vector Model


What is topology? Topology is the study of the mathematical relationships between points, lines (arcs) and polygons. Topology studies the geometric properties that do not change when the forms or shapes are bent, stretched or undergo similar transformations. For example: Polygon adjacency or containment or Line connectivity

12

9/25/2013

Topological Vector Model


There is no single, uniform set of topological relationships.
Different researchers and software have incorporated different topological information in their data structures

Planar topology is often included, as are representations of adjacency (which polygons are next to which), containment (which polygons contain which points, lines, or other polygons) and connectivity (which lines connect to which).

13

9/25/2013

Topological Vector Model


Generally, topological information is stored in internal tables.

14

9/25/2013

Vector Features and Attribute Tables


Tables are used to organize feature attributes. There is a linkage between rows in the table and the spatial data the feature layer. The most common relationship is a one-to-one linkage between each entry in the attribute table and feature in the data layer.
For each feature in the data layer there is one (and only one) entry in the table.

One-to-one relationships

15

9/25/2013

Raster data model


Defines continuous spaces or continuous variation of spatial phenomena
Uses cells or pixels arrayed in a row and column pattern to provide wall-to-wall coverage of the study region

Raster Data Model


Examples of continuous data: Elevation

Precipitation

Pollutant concentration

16

9/25/2013

Raster Data Model


Raster data sets have cell dimension (i.e. spatial resolution) The cell dimension defines the length and width of the cell in surface units. 30 m resolution = cell of 30 x 30 m Cells are usually oriented parallel to the X and Y directions

Raster Data Model


The calculation of a cell location is a simple process of counting and multiplication A cell location may be calculated from the cell size, known corner coordinates, and cell row and column number

17

9/25/2013

Raster Data Model

What is the location of this cell?

= 30 500 + (5*30) = 650 (x, y) = (200, 500) 200 + (5*30) = 350

Raster Data Model


The number of cells in a raster data set depends on the cell size. A linear decrease in cell size, causes an exponential increase in cell number (file size). For example, halving the cell size causes a four-fold increase in cell number

18

9/25/2013

Raster Data Model


Each raster cell represents a given area on the ground an is assigned a value the may be considered to apply to the entire cell.

Raster Data Model


However, under most conditions there is within-cell variation, and the raster cell value represents the average, central, or most common value found in the cell.

19

9/25/2013

Raster Data Model


Raster Features and Attribute Tables Raster layers also have associated attribute tables Maintaining a one-to-one relationship between raster cells and rows in the attribute table comes at some costs: the large size of the attribute table. To avoid large attribute tables, a many-to-one relationship is usually allowed between raster cells and the attribute table.

Raster Data Model

20

9/25/2013

Raster Data Model


Data compression Raster datasets are often compressed to save space. Compression algorithms may be lossless, in that all information is maintained during compression, or lossy in that some information is lost. A lossless compression algorithm will produce an exact copy of the original when it is applied and then the appropriate decompression algorithm applied. Lossy algorithms are most often used with image data and are uncommon applied to thematic spatial data.

Raster Data Model


Run-length encoding compression

21

9/25/2013

Raster Data Model


Quad-trees compression

Raster vs. Vector

22

9/25/2013

Object Data Model


Recent alternative. Incorporates much of the philosophy of object-oriented programming into a spatial data model. Object models attempt to encapsulate the information and operations (often called methods) into discrete objects. These objects could be geographic features such as a city, a watershed, a metro route. Geometry is another attribute. Object models can incorporate behavior.

Object Data Model

23

9/25/2013

Other Data Models


Triangulated Irregular Networks (TIN)
Used to represent terrain variation and heights.

Multiple Models
Digital data may be represented using several data models. The analyst must choose which representation to use. Digital elevation data are perhaps the best example of the use of multiple data models to represent the same theme: Digital Elevation Models (DEMs), TINs, and contour lines can be used to represent elevation.

24

9/25/2013

Multiple Representations and Models

Spatial Data Formats

25

Вам также может понравиться