Structural Elements of a Geodatabase

The structure of a geodatabase is, in reality, not very complicated. The possible structural elements number less than a dozen. The most basic element is a feature class, a shapefile. Vector data (points, lines, and polygons) have attribute tables as part of its basic structure. At the same fundamental level are raster feature classes, consisting of continuous data such as aerial photographs, scanned images, or continuous elevation data. Feature classes are grouped together to form feature datasets. Datasets can hold an unlimited number of feature classes.

Domains are set-up in the properties of the geodatabase itself. They involve setting limits on possible inputs in the attribute tables of the feature classes (shapefiles). Basically, domains create drop-down menus in these tables that help decrease errors during data entry. Relationship classes connect fields in one feature class’ attribute table to fields in another feature class’ attribute table. A possible use of this would be an address table that is used for feature classes that contain different types of features at one address, such as polygons representing houses, points representing telephone poles, lines representing underground pipes, and so forth. A major benefit of using relationship classes is that it saves data entry time by eliminating the need to enter repetitive data in multiple feature classes.

A geometric network defines directional attributes, such as the flow of water through a network of pipes or electricity through a power grid. Other elements in a geodatabase not diagramed in figure 4-7 include topology, which locks features in shapefiles to features in other shapefiles and transmits actions performed on one feature to the other (such as a delete command). An example of this type of element would be light posts locked into their power boxes. Another is a survey dataset that holds survey data for creating meets-and-bounds style survey maps.