Overview of routes
A collection of routes with a common system of measurement is called a route system. Route systems usually define linear features with similar attributes. For example, a set of all bus routes in a county would be a route system.
Many route systems can exist within a single coverage. For example, school bus, truck, and ambulance route systems could exist in a coverage of a city.
Route systems are built using arcs, routes, and sections and can accurately model linear features without having to modify the underlying arc–node topology.
The route illustrated below is defined using four arcs. Notice how the route's endpoints fall along the arcs. Routes need not begin and end at nodes.
Sections, as shown below, are the arcs or portions of arcs used to define each route. They form the infrastructure of the route system.
The diagram below shows an example of attributes containing distance measurements, such as milepost numbers or addresses, which can be used to locate events, such as accidents or pavement quality.
The routes and sections of a route system are stored in two INFO data files—the route attribute table (RAT), which defines a route feature class, and the section table (SEC), which defines a section feature class. Together, these feature classes define a route system. Route and section feature classes are often referred to as route and section subclasses because many may exist in a coverage, and each is dependent on the existence of an arc feature class.
The RAT and the SEC are feature attribute tables stored as INFO data files. A RAT and a SEC together comprise one route system. The route system must always have a name. For example, a coverage called ROADS may have a route system called BUS that contains school bus routes. The corresponding INFO data files are named ROADS.RATBUS and ROADS.SECBUS. Similarly, a route system called HAZARD has INFO files called ROADS.RATHAZARD and ROADS.SECHAZARD.
How the RAT and SEC fit in the coverage
The RAT and SEC of a route system are stored within the coverage they reference. The integrity of routes and sections is automatically maintained as arcs are edited or updated. Events are not part of a coverage but simply reference a route system from an external source.
This figure shows how RAT, SEC, and ARC files in a ROADS coverage fit together to implement a route system named BUS. The ARCLINK# item in the SEC points to the arc internal number in the ARC file. The items F-POS and T-POS in the SEC file reference the coordinates of the arc. The RAT and the SEC are related by the ROUTELINK# in the SEC and the <subclass># in the RAT.
Why use routes to represent linear features?
To associate events on linear features, a route system must be defined using routes, sections, and measures, built on arc features.
There are two major reasons why route systems are used to represent linear features:
- To avoid splitting arcs each time attribute values change (for example, to dynamically locate events on lines)
- To allow large collections of data recorded as route measures to be mapped, analyzed, and manipulated using dynamic segmentation
When arcs are used to represent linear features, a new arc must be used each time any attribute along the linear feature changes, even though an intersection has not been detected. Splitting arcs is not only redundant and wasteful but would not allow you to dynamically locate events along linear features.
Transportation, public safety, and fire departments, as well as utility and delivery companies, maintain large databases of events with route measures. For example, "Highway 400 kilometer 145" illustrates a common method for recording highway information, such as sign locations, exit ramps, pavement quality, lane closures, speed zones, accident locations, and so on. This data is usually maintained in external databases that can be accessed, mapped, and analyzed using ArcGIS.
Not only can these individual events be mapped, they can also be associated with other events. The underlying linear geography is the same. This ability to relate disparate data along a line is similar to topological overlay; it might be thought of as line-on-line or point-on-line overlay. The route measures provide the locational information; ArcGIS provides the tools to work with this data.
Route attribute tables
The RAT contains items that identify a route. Any number of user-defined attributes can be added to the RAT once it is created. At a minimum, the RAT contains the following items:
Route Attribute Table
Internal number of the route
User-defined identifier for the route
A RAT can consist of many routes that make up a particular route system. For example, the bus route system can be all the bus routes in the city. The following is an example of a RAT for a route system called BUS.
This route attribute table demonstrates the way a bus route system with one route is coded.
Routes can exhibit a variety of structural characteristics. Routes are most commonly continuous and nonbranching, but it is possible to have routes that branch or consist of disjointed sections. These types of routes are fully supported in the dynamic segmentation data model.
The measures assigned to routes can also exhibit a variety of characteristics. Measures typically monotonically increase along the route—that is, they increase without ever decreasing. Other characteristics exhibited by measuring systems may include discontinuous measures (measure gaps) or overlapping measures (more than one measure value for one location on the route).
The RAT is always created simultaneously with an accompanying SEC.
A section record in the SEC contains several items. The purpose of these items is to link the section to a particular route, define the measures for that portion of the route, and define the start and end positions of the section on the arc. The SEC contains the following items:
Identifies the route to which the section belongs (RAT internal number)
Identifies the arc that the section references (arc internal number from the ARC and AAT files)
Start measure of the section
End measure of the section
Start position of the section, defined as the percentage along the arc from the from-node
End position of the section, defined as the percentage along the arc from the from-node
Internal number of the section
User-defined identifier for the section
The value of ROUTELINK# for all the sections in a route is equal to the <subclass># of that route in the RAT. This is what links a route to its sections. A route cannot exist without its sections.
Any number of user-defined attributes can be added to the SEC once it is created. Normally though, attributes are not stored in the SEC. Attributes describing the whole route are stored in the RAT. Attributes describing portions of the route are stored as events.
How sections are defined
This section table shows how sections for a route system called BUS are coded. The table contains three sections belonging to one route. Section 1 is defined between 40 and 100 percent along arc 7. Section 2 is defined as all of arc 8. Section 3 is defined between 0 and 30 percent along arc 9.
The position of a section on an arc is stored in a SEC as from- and to- values. These values represent percentage distance along the arc, starting from the from-node of the arc. For example, in the figure above, the first section is coded as 40 to 100 for the F-POS and T-POS position items since it is located on arc 7, from 40 percent along the arc to 100 percent along the arc (that is, the end of the arc). The above figure assumes that the arcs are in the same direction as the route. If any arc is in the opposite direction of the route crossing it, the values of from and to position items in the SEC are reversed.
The measures in the SEC specify how the section is ordered within a route. These values are associated with the ends of a section. Measures can be computed automatically; by default, the length of the arc is used. For example, the first section in the above figure is the start of the bus route. The length of the arc is 90 units. Note that 60 percent of this length is 54 units. Therefore, the from measure is entered as zero and the to measure is entered as 54.
Normally, an associate RAT is created simultaneously with a SEC. It is possible, however, to have a SEC without an accompanying RAT. A SEC that does not have an associate RAT is said to have routeless sections. The practice of creating and using routeless sections is not recommended, because event data cannot be displayed and operated on without a RAT.