How to draw networks on a map

This document was published with and applies to ArcGIS 9.3.
A 9.2 version also exists.

In this topic

Drawing networks on a map

Drawing networks involves a variety of cartographic techniques. Depending on your application, you will apply some of these and other techniques:
Like other geographic features, network features can change dimension as you zoom in to a map. For example, a regional river network will be drawn as a set of lines at a scale of 1:1,000,000, but a map at scales such as 1:24,000 should begin to show the larger rivers with their areal shapes between shorelines.
Also, like other geographic features, network features are drawn with point, line, and fill symbols by a type classification. The following illustration shows sample symbols and types for hydrography layers on maps:

Techniques for roads

A large-scale street map has several special cartographic requirements. By custom, streets are drawn with casings, which symbolically represent the curb line (or edge) of a road. Also, the natural hierarchy of roads, spanning dirt roads, city roads, major roads, and divided highways, needs to be illustrated using colors, patterns, and line thicknesses. Road names and highway shields are labeled. Finally, where roads cross each other or objects such as rail lines and rivers, the correct stacking order of the objects needs to be drawn.
Drawing clean road casings is easy; define two layers on the same street feature class. On the bottom layer, set the line symbol to be the casing color, usually black, with a line thickness of three points or so. On the layer above, draw the fill with a lighter color, such as white, and a narrower thickness, about two points.
To draw a sophisticated road map like the one in the following illustration, you need a street feature class with two attributes: a type classification and an elevation level. To achieve fine control of stacking order, you need to precisely sort the drawing order of features. One way to do this is to create many layers on the street feature class and draw them in a controlled order.
The following road map shows the use of many finely defined layers in groups using elevation levels to show the correct order of roadways:
The idea is to draw feature casings (thicker, darker lines) by type for ground level. Then draw street fills (thinner, lighter lines) by type for ground level and repeat for each level. To organize these many layers, use group layers to collect layers by elevation. For more information on high-quality rendering of transportation and other layers, see Chapter 8, "Cartography and the Base Map" in "Designing Geodatabases: Case Studies in GIS Data Modeling."


Depicting the connectivity of a network is often the most important part of a map. In this view, geographic detail is unimportant.
Situations calling for schematics range from passengers who want a simple, clean map for navigating a subway system to power engineers who need to make control decisions regarding the operation of an electric system.
A schematic, like the one in the following illustration, is a simplified representation of a network that emphasizes connectivity over geographic shape. A schematic may preserve some geographic representation, such as keeping elements in relative geographic orientation, but detailed shape information is not shown. Some schematic graphic elements have no relation to geographic shapes; they show relationships along a network.
You work with schematic diagrams on the map in the same way as you work with charts or reports. A schematic is another graphic representation of a query against a database; it is composed of a set of schematic elements that have a graphic x,y location, maintain a link to their original feature, and store symbology and type information.
This link from a schematic element to the original feature is essential to automating schematics. In past practice, schematics were generated as one-time graphic products manually regenerated over time. Currently, you can automate and refresh schematics in a geodatabase when the source network features are edited.
Generating schematics requires an investment in training and configuration because the available types and properties of schematics is rich and complex. That investment can pay off for organizations, such as utilities, by reducing manual editing and improving clarity and accuracy of schematics.

See Also:

About network applications
About networks and graphs
Two core network models