ArcGIS Tutorial: Discrete data and continuous data

The values assigned to the surface cells can be expressed as discrete data or continuous data. Features and surfaces in ArcGIS can be represented as discrete data or continuous data.

Discrete data (also known as categorical or discontinuous data) primarily represents objects in the feature data storage system and raster data storage systems. Discrete objects have well-known boundaries that can be defined. It is easy to precisely define the starting and ending values of an object. A lake is a discrete object within the surrounding landscape. The intersection of lakes and continents can be clearly defined. Other examples of discrete objects include buildings, roads, and plots. Discrete objects are usually nouns.

　　

This phenomenon can be expressed through continuous data or continuous surfaces: each position on the surface can be used to measure the concentration level, or to measure the relationship between each location and the fixed point in space or with the emitter source. Continuous data also refers to field data, non-discrete data, or surface data.

One of the types of continuous surface data derives from a number of features that define the surface (measured by a fixed registration point for each position in the surface). These features include elevation (fixed point is sea level) and aspect (fixed point is direction: North, East, south and west).

　　

　　Discrete features and continuous features

Most ArcGIS applications use discrete geographic information, such as land ownership, soil classification, zoning, and land use. These types of data are represented by nominal, ordinal, interval, and ratio values. The surface is continuous data, such as elevation, rainfall, pollution concentration and groundwater level. Such data can be expressed as continuous surfaces, usually without upheaval and mutation.

　　Discrete features

Discrete features are discontinuous and have clear feature boundaries. For example, roads have widths and lengths that are represented as lines on a map. Cadastral maps can show the boundaries between parcels. The characteristics of each feature on the map, such as owner name, parcel number, and active area, are significantly different.

The following cadastral map gives an example of discrete features.

　　

Discrete map features can also be visualized as thematic data. These data or map features are simply represented as points, lines, or polygons in the map. Now you've learned how to use ArcGIS data structures to represent the topological relationships of two-dimensional features. These map features can be assigned attributes to describe, draw, symbolize, and annotate these map features. In addition, further analysis can be performed to define or identify new relationships between these features.

　　Continuous features

In contrast, continuous features are spatially non-discrete. Typically, transitions between possible values on a continuous surface have no abrupt or definite intervals. The properties of the surface are stored as Z-values, where the z-value is a separate variable in the vertical direction associated with the given x, y location. For example, the value of the surface elevation is continuous across the surface. Any representation of a surface is just an example (subset) of values in the entire surface.

　　Gradient Continuous data

Another type of continuous surface data is the phenomenon of some gradual change from one source through a surface. Examples of gradient continuous surface data include the movement of liquids and air. These surfaces are characterized by the type or manner in which the phenomenon moves.

One type of movement is through diffusion or any other type of specific movement in which the phenomenon manifests itself in moving from a high concentration area to a low concentration area until the concentration tends to balance. The change of salt concentration in the surface or water, the change of oil concentration during oil leakage and the change of heat in forest fire have the surface characteristics of this kind of movement. In such a continuous surface, there must be a source. The concentration near the source is always greater, and the concentration of the moving medium becomes smaller as the distance from source increases.

　　

In the above source concentration surface, the concentration at any position is determined by the ability of the event source to move in the medium.

Another type of movement is controlled by the intrinsic characteristics or movement patterns of the moving substance. For example, the movement of noise generated by a bomb explosion is controlled by the inherent characteristics of the noise and its passing media. Motion patterns may also limit and directly affect the surface concentrations of features, such as the propagation of plant seeds. The way the seeds are moving (whether by bees, humans, wind or water) will affect the surface concentration of plant seeds.

　　

Other examples of movement include the migration of animal populations, potential customers in stores (cars are sports, time is a limiting factor) and the spread of disease.

　　Discrete or continuous?

It is not obvious whether the boundary is continuous or discrete when many features are represented and modeled. Continuum is created in the process of representing geometric features, and extreme cases are pure discrete features and pure continuous features. Most of the elements are located between these two extreme situations.

Examples of features that are distributed along a continuum are: soil types, forest margins, wetland boundaries, and geographic markets affected by TV advertising campaigns. The determinant of where a feature falls from a continuous to a discrete range is the ease with which the feature boundary is defined. Regardless of where the feature falls in the contiguous region, the raster data can represent the feature with a higher or lower precision.

When making decisions based on generated values, it is important to understand the type of data being modeled (whether it is discrete or continuous). The exact location of the building should not be determined solely on the basis of soil maps. The square area of the forest cannot be used as a primary consideration when determining a suitable habitat for deer. Sales activities should not rely solely on the geographical market impact of television advertising. It is important to understand the validity and accuracy of the input data boundaries.

ArcGIS Tutorial: Discrete data and continuous data