[an error occurred while processing this directive]

Go to index.

Section 3: Mapping of Spatial Objects

3. Mapping of DLG-E Spatial Objects

This section describes the mapping of DLG-E spatial objects to the spatial objects of the SDTS. The mapping of spatial objects is based on definitions. The definitions of the DLG-E spatial objects are stated. These are then compared to the spatial object terms and definitions in the SDTS.

3.1 DLG-E Spatial Objects

The DLG-E model defines the following elemental spatial objects: point, node, chain, and polygon. These elements are used as components of aggregate spatial objects: planar graph and surface. The precise definitions of each of these DLG-E spatial object terms are in the table in section 3.2.

3.2 Mapping to SDTS Spatial Object Terms

The mapping of DLG-E terms to standard ones is done using the spatial object definitions in SDTS, Part 1, Section 2.3. The results of the mapping are stated in the table following this section. The first two columns contain the DLG-E term and definition. The last two columns contain the corresponding SDTS term and its description, including modifications made by the Topological Vector Profile.

As shown in the mapping table, all terms map one to one except polygon. The DLG-E definition of polygon includes the concept of a representative coordinate. This is modeled as a distinct spatial object in the SDTS. (It would be incorrect to encode this point as an attribute of the polygon.) DLG-E also includes the special polygon of the area outside the planar graph. The SDTS has a separate term for this to make the object more prominent.

The third column of the table lists the set of SDTS spatial objects that a DLG-E encoding will use. This set conforms to the restrictions of the Topological Vector Profile.

                                    Section 3: Mapping of Spatial Objects

DLG-E Term  DLG-E Definition                          SDTS Term        SDTS Description

  Point       A zero-dimensional spatial object that  Entity Point     A point used for identifying the location of
              is not along the path of any chain.     (NE)             point features or areal features collapsed
                                                                       to a point
  Node        A zero-dimensional spatial object that  Node (NO)        A topological junction of two or more
              is the junction of two or more chains                    chains, or an endpoint of a chain; component
              or is an end point of a chain.                           of a planar graph

  Chain       A one-dimensional spatial object that   Complete Chain   A directed nonbranching sequence of
              is a directed connection between two    (LE)             nonintersecting line segments bounded by
              nodes (not necessarily distinct) with                    nodes, not necessarily distinct, at each
              geometric location specified by a                        end; explicitly references left and right
              sequence of two or more coordinate                       polygons and start and end nodes; it is a
              pairs (or triplets.)                                     component of a two-dimensional manifold
  Polygon     A two-dimensional spatial object which  GT-Polygon (PC)  An area that is an atomic two-dimensional
              consists of the area interior to a                       component of one and only one two-
              closed circuit of chains, excluding                      dimensional manifold; its boundary is
              holes. (It may have a representative                     composed of chains
              coordinate associated with it.)         Universe         The area outside the perimeter of the area
                                                      Polygon (PW)     covered by other GT-Polygons
 
                                                      Area Point (NA)  A representative point within an area  
  Planar      A set of topologically interrelated     Planar Graph     A set of topologically interrelated zero-
  Graph       zero-dimensional (node) and one-        (GP)             dimensional (node) and one-dimensional
              dimensional (chain) spatial objects,                     (chain) objects that conform to a set of
              conforming to a set of well defined                      defined constraint rules: each chain is
              constraint rules: each chain is                          bounded by an ordered pair of nodes, not
              bounded by an ordered pair of nodes                      necessarily distinct; a node may bound one
              (not necessarily distinct); a node may                   or more chains; and, chains may only
              bound one or more chains; not more                       intersect at nodes
              than one node may exist at any given
              planar location; and chains may not
              intersect. 

3.3 From Terms to Modules

The SDTS assigns each spatial object type to a module type. The objects are identified by using representation codes in the module records. In terms of module records, the definitional equivalences for the elemental spatial objects imply that:

The SDTS spatial object records have an attribute record pointer. This field will not be used by DLG-E. In DLG-E, attributes are attached to features and not to spatial objects.

The partitioning of spatial objects into modules will be based primarily on participation in the same DLG-E surface (SDTS 2-d manifold.), and secondarily on SDTS object representation code. (This is the partitioning required by the Topological Vector Profile.) The DLG-E planar graph will not be explicitly transferred as an aggregate spatial object because it is used as a part of the DLG-E surface.

For a single DLG-E surface, there will be a Point-Node module (for E nodes), a Line module (for E chains), and a Polygon module (for E polygons). In addition, there will be a Point-Node module for DLG-E points and another for DLG-E Polygon Points.

(Pending Issue: The SDTS recognizes the "void polygon" as a distinct type of spatial object. The concept of a void area is unresolved in DLG-E. So, the first DLG-E datasets do not contain void areas.)

Go to index.

[an error occurred while processing this directive]