Link to USGS Home Page

[Top] [Prev] [Next] [Bottom]



Annex B

(informative)

The SDTS Model of Spatial Data

The basis of SDTS is a model of spatial data sufficiently general so that any user data can be accepted, but at the same time sufficiently structured to provide an adequate basis for the organization of spatial phenomena. This model encompasses the spatial objects in section 2 and also the attributes associated with these objects (which are of equal importance), and the ways in which objects and attributes are organized and bound into other more comprehensive forms. The SDTS model of spatial data incorporates the model concepts of part 2 (Spatial Features), including the entity type and attribute concepts used to organize the definitions in part 2.

This spatial data model is presented as an informative annex, because the model may be considered an aid to understanding the normative concepts of part 1, rather than an absolute requirement for spatial data transfer.

The SDTS conceptual model has three parts: a model of spatial phenomena; a model of the spatial objects used to represent phenomena; and a model of spatial features, which explains how spatial objects and spatial phenomena are related.

Terms from the Object Oriented Programming System literature are used to define the parts of the model. These terms are given specific definitions in this standard as a way of distinguishing the various criteria for grouping elements into sets that are meaningful as spatial data abstractions. The terms are:

a) Phenomenon. A fact, occurrence or circumstance. Route 10, George Washington National Forest, and Chesterfield County are all phenomena.
b) Classification. The assignment of similar phenomena to a common class. An individual phenomenon is an instance of its class. Route 10 is an instance of the class road.
c) Generalization. A process in which classes are assigned to other classes. The general class includes all the instances of the constituent classes. Sewers are included in the more general class of utilities.
d) Aggregation. The operation of constructing more complex phenomena out of component phenomena. A lock is an aggregation of walls, gates, and a reservoir.
e) Association. The assignment of phenomena to sets, using criteria different from those used for classification. Concrete roads may be associated with concrete sewers, concrete locks, and other phenomena constructed of concrete.

B.1 Model of Spatial Phenomena

SDTS transfers information about phenomena that are defined in space and time, and so are described using a fixed location-- spatial phenomena. All phenomena are defined as belonging to a class of phenomena. (Smith's Farm belongs to Farm.) A characteristic of such a class is called an attribute. (Acreage is an attribute for Farm.) An attribute value is a specific quantity or quality of the attribute assigned to a phenomenon in that class. (Smith's Farm has an Acreage of 160 acres.)

Whether a given phenomenon belongs to a class is determined by the definition of the class. The definition consists of a statement about characteristics all members of the class have in common. It also includes characteristics which distinguish the class from other classes. These definitional characteristics are necessary and sufficient conditions for classifying some phenomena into the class and excluding others. The data collector defines which classes of phenomena are of interest. Those classes of phenomena are called entity types, and the individual phenomena are called entity instances.

Certain attributes are identified with each class. The attributes of a class include key attributes. The key attributes are the set of attributes of a class such that the combination of values of the key attributes forms a unique identifier for each entity instance.

Conceptually, an entity instance is a spatial phenomenon of a defined type that is embedded in one or more phenomena of a different type, or that has at least one key attribute value different from the corresponding attribute values of the surrounding phenomena. Such an entity instance is not further subdivided into phenomena of the same type. Examples of entity instances include obvious physical structures (such as a bridge), fuzzy physical regions (such as a soil polygon), cultural artifacts (such as a political boundary), and modeling constructs (such as an economic region).

The distinction between class and instance of entities can be shown as:

					Class			Instance
	Class or instance				Entity Type			Entity Instance
	Characteristic				attribute			attribute value

An example of an entity type is Bridge with attributes Name and Composition. An instance of this type might be the "10th Street Bridge" composed of "steel." "10th Street Bridge" and "steel" are attribute values of Name and Composition. Another example is the entity type Farm with key attribute Owner and non-key attribute Crop_cultivated. Smith's Farm is adjacent to Jones' Farm. Both cultivate corn and soybeans. The two farms can be distinguished as entity instances because of the change in the value of the key attribute Owner. Across the road from the two farms is a shopping mall. The instances of the entity type Farm are bounded by other farms with different key attribute values, but also by instances of other types of entity such as Shopping Center and Road.

Entity instances may be aggregated into instances of a different type of entity. For example, although a lock is partly composed of walls, it is not itself a wall.

Entity types can be generalized into themes based on the definitional characteristics shared by more than one class. A theme can also have its own attributes, including Name. An example of a theme is "transportation" which is by definition a function of both Railway and Road.

The distinction between class and instance of themes can be shown as:

					Class			Instance
	Class or instance				Theme			Entity Type
	Characteristics				Theme Name			Entity Label

Associations of entity instances are defined in terms of characteristics other than those used to define an entity type. A common association is the spatial domain, which groups all entity instances having coordinates within a specified range. Another useful association is temporal domain. SDTS represents entity instances as static, without temporal dimension. However, values of a temporal attribute such as Age may be assigned to entity instances, and used to associate them into sets with a common extent in time.

A relationship is a special case of an association. A relationship exists between entity types. A relationship instance is an association between entity instances with a unique relationship value.

The distinction between class and instance of relationships can be shown as:

					Class				Instance
	Class or instance				Relationship				Relationship instance
	Characteristics				Relationship type				Relationship value

B.2 Model of Spatial Objects

Entity instances have a digital representation. That digital representation consists of one or more spatial objects. A spatial object may be an aggregation of other spatial objects, not all of which necessarily represent an entity instance. A spatial object that represents all of a single entity instance is an entity object. It may be classified into an entity object class. Entity objects have generalizations and associations as well: the representation of an entity theme is an object theme, and an entity spatial domain, an object spatial domain.

In general, the correspondence between entity instance and entity object is paralleled by all characteristics of entities and objects. Rather than creating a whole new set of terms, the characteristics' names are generally the same whether referring to phenomena or their digital representation. In this standard, whether phenomena or their digital representations are being referred to is indicated by the context. The following table gives examples:

				Phenomena				Digital Representation
	Class			Entity Type				Entity Object Class
	Characteristic			Attribute				Attribute

	Instance			Entity Instance				Entity Object
	Characteristic			Attribute Value				Attribute Value

	Generalization			Theme				Theme
	Association			Spatial Domain				Spatial Domain

Entity objects have locational attributes (spatial address), non-locational attributes, and relationships (topology). The attributes and relationships of entity objects need not be as extensive as those of their corresponding entity instances. The key attributes used to distinguish a particular entity instance may not be present in the actual transfer; instead, the entity object record identifier may be the only way to distinguish between instances.

Spatial objects may have attributes independently of whether they are entity objects or not. All objects may be classified, aggregated and associated in the same general manner that entity instances may be.

This standard defines a set of simple spatial objects. These simple spatial objects are either primitive objects (not aggregated from any other objects), or are aggregated only from spatial objects belonging to different classes (polygons are not aggregated from polygons, only from rings, chains or strings). The only exception is the composite object. Composite spatial objects may be aggregated from simple objects or from other composites.

Spatial objects are classified into module types, one of the basic building blocks of the standard. Once defined, modules may be associated into sets by spatial domain, temporal domain, data quality, security requirements, topological relationships, or any other criteria.

B.3 Model of Spatial Features

A feature type consists of an entity type and the entity object class that represents it. If a class is viewed as a set, whose members are the instances of the class, then a particular feature type is the intersection of the entity type and the entity object class. The feature type represents those entity instances that have representative entity objects.

The term feature is defined in the standard for completeness. A spatial data transfer contains entity objects and other spatial objects. Entity instances are not transferred; they exist in the real world.



[Top] [Prev] [Next] [Bottom]
| SDTS Home Page | MCMC Home | Geography | USGS | Search

U.S. Department of the Interior || U.S. Geological Survey
1400 Independence Road, Rolla, MO 65401
For general information call: (573)308-3500
URL: http://mcmcweb.er.usgs.gov/sdts/SDTS_standard_nov97/part1an1.html
Last modified: Monday, 14-Jan-2013 19:27:42 EST
Maintainer: mcmcweb@usgs.gov
Privacy Statement || Disclaimers || FOIA || Accessibility