Skip to main content icon/video/no-internet

Raster is a method of data organization for spatial data. The data structure for raster data consists of rows and columns of discrete data points that inherently preserve the spatial relationships between the locations on the earth represented by the data points. While it is not always the case, the data “points” are usually represented as areas or “cells,” which are often rectangular. Thus, a digital camera image is a common example of a raster data set in which a rectangular pixel (picture element) is used to represent each data point. The terms pixel and cell are used some-what interchangeably, though pixel is more correctly used with respect to raster images.

The raster data structure is a widely used method of representing geospatial information. The format in general is very flexible and can be used for a number of different types of geospatial data, including thematic layers, such as elevation, land use or watersheds, or images captured by satellites or airplanes. The key concept is that every point within a raster file has an implied map coordinate. This is in contrast to vector data, in which each spatial element must be associated explicitly with its location. Many of the display and analysis functions within GIS handle either raster or vector data and in some circumstances a combination of both. This entry describes the function and characteristics of raster data sets within the context of GIS.

Cell (Pixel) Dimensions

Raster data have a number of characteristics that determine how GIS operations may be performed. A raster file has a specified number of rows and columns of data that represent the values of a particular phenomenon over a specific area on the earth's surface. If each cell has the same spatial extent in x and y, the area represented by each cell is consistent throughout the raster. Areas within the image may be calculated easily by multiplying the area covered by each cell by the number of cells. This is the most widely used pixel shape in remote sensing and GIS.

However, raster data may have other shapes. Raster data sets are often created using collection points separated by distances measured in terms of degrees, minutes, or seconds of longitude and latitude. In this case, the area of all pixels is not the same, changing as one moves north and south through the data set. Thus, if the raster covers a significant area on the earth's surface, the ground area of a cell is less the farther it is away from the equator because of convergence of lines of longitude at the poles.

Another shape for cells in a raster data set is hexagonal, which is used for numerical modeling to ensure uniform spacing between points in diagonal as well as rectangular directions. Other shapes (triangles, irregular rectangles, etc.) can be used, each with their own characteristics in regard to spatial data functions, such as angular, linear, and area measurement.

The geographic coordinates of a raster data set are normally stored in an ancillary file or in a header record within the raster file. Since a raster is (usually) a regular grid, only the top-left geographic coordinates are needed to compute the geographic position of any cell or pixel within the raster file. A raster file may contain more than one spatial variable (land cover, roads, slope, etc.); however, the cell size of each layer represented within a single raster file must be the same. This same architecture supports the ability of a raster file to contain geocorrected multiband images. Each spectral band of the multispectral image may be treated as a separate variable.

...

  • Loading...
locked icon

Sign in to access this content

Get a 30 day FREE TRIAL

  • Watch videos from a variety of sources bringing classroom topics to life
  • Read modern, diverse business cases
  • Explore hundreds of books and reference titles

Sage Recommends

We found other relevant content for you on other Sage platforms.

Loading