Lecture 2

Introduction to GIS, Data models, Data structures

REPRESENTATION AND DATA STRUCTURESCoordinates and Attributes

Common Data Models

REPRESENTATION AND

DATA STRUCTURES

•Most common data

models define thematic layers

•Typically, layers, one

layer for each distinct view of a theme

Part 1: Coordinates (Cartesian)

Spherical coordinate can be expressed two ways

e.g.,

44o or

44.51o

30’ 35”(DMS)

(DD)

We Can Convert

Three Types of Vector Features

One-to-one,because of Attributes

Enforced Uniformity

Planar Topology – no overlaps

On a blank sheet of paper, draw a

topologically correct (not necessarily

geometrically correct) rendition of the 6

coastal (name counties, if you can).

Identify all the nodes.

Create a table corresponding to the

geographic data you drew (make the

linkages clear).

Include an attribute for county area, in

square kilometers, and square miles

Rasters – Fixed Cell Size, Grid Orientation

Connecting data, contrast with vector

Raster, one-to-one correspondence, how

many rows in the attribute table?

Number of cells = 100!Raster, one-to-one correspondence, rarely used

Many-to-one much more common, to tame the attribute table

How can we mimic a vector one-to-one relationship between individual polygon codes and table rows, when using a raster, without keeping track of individual cells?

Raster – The Mixed Pixel Problem

Landcover map – Two classes, land or water

Cell A is straightforward

What category to assignFor B, C, or D?

Resampling Ambiguity

Changing Resolution (resampling) with Categorical

Data Assignment Method Matters!

Orientation and/or Cell Size May Differ

Resampling - Distance-weighted averagingcontinuous data:

bilinear interpolation

z

No Decision is Final – We Can Convert

Data and File Structures

Data are stored as binary numbers

Bits are 0 or 1

Bytes are 8 bits

Data (e.g., raster cells)are often references as 1 byte, two byte, etc.

Cells Have a Type, Size

Type – e.g., Real, unsigned integer, signed integer, text

Size – 8 bit, 32 bit, 64 bit, long, short, character width

These control the size of the datasets, and type of data that may be stored

Mixing types, sizes, often requires some care

Raster data set properties

We can store a number upTo 2 or4,293,967,296 in a cell

32

Data andFile Structures

Data often have specificorganization to

•reduce size

•speed access

•ease updates

Example: ESRI ShapefilesLandcover dataset, wash_lcis a cluster of files,

wash_lc.shp - containing the coordinates wash_lc.dbf - containing the attributes wash_lc.shx - containing linkages, other infowash_lc.prj - optional, containing projection information wash_lc.sbn - an optional indexing file

Thursday, August 22, 13

Data and File Structures CompressionReducing size – e.g., raster run-length

coding

Rasters – Discrete or Continuous

Features Can we compress either

without loss?discretecontinuous