Sunday, May 13, 2007


GIS data can be broadly described as- Spatial data and Non-spatial data.


Spatial data is geographical representation of features. In other words, spatial data is what we actually see in the form of maps (containing real-world features) on a computer screen. Spatial data can further be divided into two types- vector and raster data.

Vector Data

Vector data represents any geographical feature through point, line or polygon or combination of these.

1. Point

A point in GIS is represented by one pair of coordinates (x & y). It is considered as dimension-less object. Most of the times a point represent location of a feature (like cities, wells, villages etc.).

2. Line

A line or arc contains at least two pairs of coordinates (say- x1, y1 & x2, y2). In other words a line should connect minimum two points. Start and end points of a line are referred as nodes while points on curves are referred as vertices. Points at intersections are also called as nodes. Roads, railway tracks, streams etc. are generally represented by line.

3. Polygon

In simple terms, polygon is a closed line with area. It takes minimum three pairs of coordinates to represent an area or polygon. Extent of cities, forests, land use etc. is represented by polygon.

Raster Data

Raster data is made up of pixels. It is an array of grid cells with columns and rows. Each and every geographical feature is represented only through pixels in raster data. There is nothing like point, line or polygon. If it is a point, in raster data it will be a single pixel, a line will be represented as linear arrangement of pixels and an area or polygon will be represented by contiguous neighbouring pixels with similar values.

In raster data one pixel contain only one value (unlike vector data where a point, a line or a polygon may have number of values or attributes) that’s why only one geographical feature can be represented by a single set of pixels or grid cells. Hence a number of raster layers are required if multiple features are to be considered (For example- land use, soil type, forest density, topography etc.).

As discussed earlier digital satellite images are also in raster format.


Attributes attached to spatial data are referred to as non-spatial data. Whatever spatial data we see in the form of a colourful map on a computer screen is a presentation of information which remains stored in the form attribute tables. Attributes of spatial data must contain unique identifier for each object. There may be other field also containing properties/information related a spatial feature. Attribute table of spatial data also contains ‘x’ and ‘y’ location (i.e. latitude/longitude or easting/northing) of features; however in some GIS software these columns may remain ‘invisible’.

For example- if we are doing demographic analysis of villages then attributes of each point (representing a village) must have a unique village ID and other demographic information like total population, number of males & females, number of children etc.

In another example- if we are doing some GIS analysis related to road then each road must have its unique Road ID. Other attributes may include like road length, road width, current traffic volume, number of stations etc.

Please see INDEX for complete list of topics

No comments: