Two primary data structures in GIS - Vector and Raster data

Vector - each object is composed by a series of x,y coordinate pair(s)

• Points - single x,y pair
• Lines - nodes at beginning and end, vertices for changes in direction along a line
• Polygons - closed set of lines's

Vector structure: The start and end points of a line are called nodes, any change of direction on a line is a vertex. Lines are joined by sharing a common node. Polygons are joined (adjacent) if they share a common line.

Vector objects are all created from the general graphic primitives of single x,y coordinates. points are single coordinates, lines are strings of coordinates, polygons are a closed string of coordinates (often composed of multiple lines). See here.

Vector representation of hydrography from an air photo

Vector objects all have an ID that ties the spatial data object to the attribute data describing that object

Raster - set of rows and columns that form individual cells. Each cell has a unique data value and explicitly has an area coverage proportional to the cell size. Most rasters use square cells and a uniform cell size.

A graphical comparison of vector and raster data structures

Another comparison of vector and raster representations

Critical issues with vector data representation:

Topology

Topology broadly refers to the relation between features. In a GIS this refers to these specific qualities:

Connectivity of line features
Directionality of line features
Containment of features within polygons

Some implementation of vector data structures are topological (e.g. coverages) and some are not (shapefiles, geodatabases in ArcGIS 8.2)

Lines can not overlap without a node in a topological data structure. Lines can overlap without nodes in a non-topological data structure (e.g. spaghetti). See here

Scale

The number of vertices determines the precision with which a line feature can be represented.

More vertices allows for more line curvature to be represented. Straight line segments do not require vertices along their extent since additional vertices does not change the representation of the line.

Vector data are input into a GIS by digitizing maps or through input of GPS based coordinates.

The scale of the source of the data is related to the spatial accuracy of the line feature in the GIS. Data from coarse scale maps are highly generalized. Data from fine scale maps (e.g. 1:10,000) are highly detailed.

The same feature coming from different sources can have widely varying representations.

Graphical depiction of vector features with different levels of precision (Heywood 1998) (as might be the case if they were derived from source datasets of different scales)

Vector Data

Graphical depiction of simple vector data

Graphical depiction of more complex vector data

Regions

Basically a workaround to allow for overlapping polygons using Coverage data format. Regions are composed of combinations of vector polygons created by the intersection of a set of component polygonal areas.

Dynamic Segmentation

Means to more efficiently handle 'events' or locations along a line segment. This can be done by creating a line intersection and node at each location along a line segment but this is inefficient because it increases data storage and data cleaning. Instead a distance marker is used to identify a location along a line segment. Since line features in topological data implementations have directionality it is apparent where the object lies in coordinate space by locating its position along a line feature.

Applications includes identifying the location of accidents along a road network or the location of manhole covers along a sewer system database.

TIN Triangulated Irregular Network

Quasi three dimensional representation using a vector data implementation. Sometimes called 2.5 dimensional data. TINs consist of point data between which are edge features that collectively create triangular areas of constant gradient representing gradients in the landscape. Most often used for terrain mapping applications but raster data are as or perhaps more commonly used for these applications.

Discussion questions:

Would you use a point or polygon to represent student building for a GIS of a) campus, b) city of Bloomington, c) Monroe County?

Would you use a line feature or a polygon feature to represent 7th St. and Highway 37 in a GIS of a) City of Bloomington, b) Monroe County, c) state of Indiana?

For what applications would you need a topological data implementation?

What applications might not require a topological data implementation?

If you have a GIS that includes road data derived from 1:50,000 scale maps can you use these data to produce a map at a) 1:250,000 scale, b) 1:100,000 scale, c) 1:50,000 scale, d) 1:30,000 scale, e) 1:10,000 scale?