Geographic Information Systems: An Introduction, 3rd Edition

4.5 Vector Versus Raster Models

One of the basic decisions in GIS design involves the choice between vector and raster models, each of which has advantages and disadvantages. In the vector model, the observation units are end points and/or variable line or polygon magnitudes, whereas the raster model presupposes fixed observation areas in a grid. Otherwise, the models are, in principle, identical.

As discussed previously, vector and raster data have varying ability to represent reality. It is not always easy to recognize vector data's discrete objects out in the terrain. This applies especially to phenomena with diffuse borders and continuous changes, such as vegetation and population density. However, many real phenomena are related to locations. Measurements are often made at points, infrastructures are often related to lines, and administrative units are frequently described in terms of defined areas of various shapes and sizes.

Raster GIS emphasizes properties: Here, the basic units of observation are regular cells in a raster. As opposed to vector data, raster data are relatively well suited to represent continuous phenomena, provided that an appropriate cell size is selected. But not all phenomena are related directly to such grid patterns. At present, satellite data, digital ortho images, and digital elevation data account for the bulk of data available in raster form. In many countries, national elevation data models have been established based on a fixed grid (e.g., 100 ...