5.1 How Does GIS Represent Real-World Items?

GIS has a multitude of applications, ranging from mapping natural gas pipelines to determining which parts of the natural landscape are under the greatest pressure to be developed. No matter the application, the data (such as the locations and lengths of pipelines or the area and extent of land cover) needs to be represented somehow within the GIS. GIS provides a means of representing (or modeling) this data in a computer environment so that it can be easily analyzed or manipulated. A model is a kind of representation that simplifies or generalizes real-world information. For instance, if you were tasked to make a map of all of the stop signs in your city, you’d need to be able to represent the locations of all the signs on a map at the scale of the entire city. By laying out a map of the city and placing a dot at each sign location, you would have a good representation of them. The same holds true if you had to come up with a map of all the roads in your town. You could attempt to draw the roads at their correct width, but the roads’ locations can be represented simply and accurately by lines drawn on the map.

Whenever you’re considering how to model real-world phenomena with a GIS, you first have to decide on the nature of the items you want to represent. In the GIS, there are two ways of viewing the world. The first is called the discrete object view. In the discrete object view, the world is made up of a series of objects, each of which has a fixed location, or a fixed starting and stopping point, or some sort of fixed boundary. A street starts at one point and ends at another, or there is a definite property boundary between your property and your neighbor’s. Around a university campus, there will be numerous objects—representing buildings, parking lots, stop signs, benches, roads, trees, and sidewalks. By viewing the world in this way, real-world items can be modeled in the GIS as a series of objects (like a collection of lines to represent a road system, or a series of points to represent fire hydrant locations).

When adapting the discrete object view of the world to a data model, items in the real world are represented in the GIS by one of three objects (Figure 5.2):

• Points: These are zero-dimensional objects, a simple set of coordinate locations.
• Lines: These are one-dimensional objects, created from connecting starting and ending points (and any points in between that give the line its shape).
• Polygons: These are two-dimensional objects that form an area from a set of lines (or having an area defined by a line forming a boundary).

These three items are referred to as vector objects, as they make up the basis of the GIS vector data model. Basically, the vector data model is a means of representing and handling real-world spatial information as a series of vector objects—items are realized in the GIS as points, lines, or polygons. For instance, locations of fire hydrants, crime incidents, or trailheads can be represented as points; roads, streams, or power conduits are represented as lines; and land parcels, building footprints, and county boundaries are represented as polygons. When dealing with a much larger geographic area (for instance, a map of the whole United States), cities may be represented as points, interstates as lines, and state boundaries as polygons (Figure 5.3).

A lot of GIS data has already been created (such as road networks, boundaries, utilities, etc.), but in order to update or develop your own data, there are a few steps that need to be taken. Digitizing is a common way to create the points, lines, and polygons of vector data. With digitizing, you are (in essence) “tracing” or “sketching” over the top of areas on a map or other image to model features in that map or image in the GIS. For instance, if you have an aerial photograph of your house being shown on the screen of your GIS software, you could use the mouse to sketch the outline of your house to create a polygon and save it in the GIS as a representation of your house. If you sketch outlines of four of your neighbors’ houses as well, you’ll have several polygons stored in a single layer in the GIS. When you examine the data, you’ll find you have five polygon entries in that data layer (your house and four others). Similarly, you could sketch each house’s driveway as a line object by tracing the mouse over the appropriate place in the image, which would result in a second data layer that’s comprised of five lines. With digitizing, each of your sketches can be translated into a vector object.

Digitizing is commonly done through a process referred to as “heads-up digitizing” or “on-screen digitizing,” which is similar to the sketching just described. In “heads-up” digitizing, a map or other image (usually an aerial photograph or satellite image) is displayed on the screen as a backdrop. Decide if you’ll be sketching points, lines, or polygons and create an empty data layer of the appropriate type (to hold the sketches you’ll make). From there, you use the mouse to sketch over the objects, using as much detail as you see fit (Figure 5.4). Keep in mind that despite your best efforts, it may likely be impossible or unfeasible to sketch every detail in the image. For instance, when you’re tracing a river, it may not be possible to capture each bend or curve by digitizing straight lines.

When objects are created in GIS, they can be set up with coordinates, but the GIS still needs to have knowledge of how these objects connect to each other and what relation each object has to the others. For instance, when two lines are digitized that represent crossing streets, the GIS needs to have some sort of information that an intersection should be placed where the two lines cross. Similarly, if you have digitized two residential parcels, the GIS would require some information that the two parcels are adjacent to one another. This notion of the GIS being able to understand how objects connect to one another independent of their coordinates is referred to as topology. With topology, geometric characteristics aren’t changed, no matter how the dataset may be altered (by projecting or transforming the data). Topology establishes adjacency (that is, how one polygon relates to another polygon, in that they share a common boundary), connectivity (that is, how lines can intersect with one another), and containment (that is, how locations are situated inside of a polygon boundary).

Digital Line Graphs (DLGs) are examples of vector data applications in GIS. DLGs were created from United States Geologic Survey (USGS) topographic maps (see Chapter 13 for more information about topographic maps), featuring vector datasets representing transportation features (such as streets, highways, and railroads), hydrography features (such as rivers or streams), or boundaries (such as state, county, city, or national forest borders). Since these topographic maps are no longer being produced, neither are the corresponding DLGs, but the existing products are freely distributed by the USGS (see Figure 5.5 for an example of several DLGs in GIS). The DLGs were created by digitizing, using a USGS map as the source and the map features (like roads or rivers) turned into a digital format using GIS. See Hands-on Application 5.3: USGS Digital Line Graphs for more about how to obtain DLGs.

There’s also a second way of viewing the world, in which not everything has a fixed boundary or is an object. For instance, things such as temperature, atmospheric pressure, and elevation vary across Earth and are not best represented in GIS as a set of objects. Instead, they can be thought of as a surface that’s made up of a near-infinite set of points. This way of viewing the world is called the continuous field view. This view implies that real-world phenomena continuously vary, and rather than a set of objects, a surface filled with values is used to represent things. For instance, elevation gradually varies across a landscape from high elevations to low elevations filled with hills, valleys, and flatlands, and at every location we can measure the height of the land. Similarly, if you’re standing at a ranger station in a park and preparing a rescue plan for stranded hikers, you’ll want to know the distance from the station to every location in the park. Thus, you could have two layers in your GIS—one showing the elevation at every place and another showing the distance to the ranger station at each location.