USGS visual identity mark and link to main Web site at

Digital Mapping Techniques '01 -- Workshop Proceedings
U.S. Geological Survey Open-File Report 01-223

Data Structure for the Arizona Geological Survey Geologic Information System: Basic Geologic Map Data

By Stephen M. Richard and Tim R. Orr

Arizona Geological Survey
416 W. Congress, #100
Tucson, AZ 85701
Telephone: (520) 770-3500
Fax: (520) 770-3505


Geoscience data are used for land-management decision-making, for engineering design, in the search for mineral resources, and for scientific research. Traditionally, geologic information has been stored and disseminated using geologic maps and written reports (Bernknopf et al., 1993). Because of the complexity of the earth, much of the information included in a geologic map is buried in several layers of abstraction. Specific applied use of geologic data typically requires preparation of a derivative map designed to show a particular aspect of the geologic data. Such maps might be designed to show rock lithology, the orientation of bedding or foliation in layered rocks, the acid buffering capacity of the rocks, or to show rocks of a particular age. Production of such derivative maps designed for a specific purpose commonly requires a geologically sophisticated analysis of the original map, as well as cartographic design and drafting of the derivative map.

Computer-based geographic information systems allow the manipulation and analysis of much larger and more sophisticated geographic data sets than was possible using paper maps and physical overlays. These systems provide tools to manipulate and integrate geologic data with other geographic data to a greater extent than ever before possible. A well designed, data-rich information system could automate much of the process of producing derivative maps designed for specific applications. This would free the data user to explore the data in more ways, and to experiment with different representations of the data. Providers of geoscience data, like the Arizona Geological Survey, must redesign information delivery systems to facilitate the integration of their geologic data resources into automated systems, and to maximize the usefulness of geologic information.

To this end, the Arizona Geological Survey is developing a computer-based geologic information system designed to meet the needs of mineral exploration geologists, researchers in search of detailed technical information, land managers or planners requiring information pertinent to regulatory, planning, and development functions, and curiosity-driven users from the general public. Many of these users may not be expert geologists, but still need to be able to query the system to obtain information. The underlying data model must be flexible enough to encompass a wide range of earth science information, storing it in such a fashion that it does not become obsolete with advances in geologic science.

Based on several years of development and discussion with other database developers (see papers in Soller, 1997; 1998; 1999; 2000; and this volume), this system has evolved into a structure with a variety of inter-related components, summarized in Table 1. This document defines a relational database implementation of the metadata, cartography, geologic map, and geoscience infrastructure parts of the Arizona geologic information system necessary to represent the basic geologic information and cartography recorded on a typical geologic map. This information includes the assignment of map units to regions on the map, the classification of boundaries between the map units as faults or contacts (here referring to depositional or intrusive contacts), the recording of basic point-referenced structural data, and the cartographic representation of these features. Subsequent documents will describe the detailed geoscience description tables (map units, lithology, age dates, stratigraphic relationships, etc.). The implementation is based on Microsoft Access (currently using the Access2000 version; datasets are distributed with Access97 tables for wider accessibility) as the relational database, and ESRI ArcInfo (v.8.0.1) and ArcView (v.3.2) as the geographic data system.

Table 1. Components of Arizona Geological Survey Information System.

Components of Arizona Geological Survey Information System

[Click HERE for a Word version.]


A geologic data set is a collection of map unit definitions, interpretations of the nature of the boundaries between the map units, locations of faults and boundaries between the map units defined, and descriptions (quantitative and qualitative) of the nature of the map units, structures, faults, and map unit boundaries. Data set as used here is independent of the format of the data -- it may be digital or analog (Richard, 2000). A geologic map image is a visual representation of a geologic data set for an area, designed to communicate information to a user. The map image is defined by the map area extent, the geologic data (both spatial location and classification) used, the choice of symbols for geologic features, the map projection and scale, a specification of the surface represented by the map, and the cultural and physiographic base map. The path from a geologic data set to a geologic map image requires selecting symbols to represent the distribution of the map units, the location and type of map unit boundaries and faults, and the location and relevant data for point observations (orientation measurements). These symbols are placed on a base map that represents the map area by means of a projection and some elevation model to represent topography on the mapped surface. The base map provides a visual reference frame to depict the spatial relationships between geologic features, and a means of physically locating the features depicted. Design of the base map is an important aspect of cartography. This definition of a map image makes no distinction between a standard geologic map (map surface = earth surface), a mine-level map (map surface = horizontal plane), or a geologic cross section (map surface = vertical plane along section line).

A digital geologic data set represents a geologic data set in a georeferenced form using a set of computer files. A digital geologic data set is defined by:

  1. The conceptual model that is the basis for the geologic data set.
  2. A logical data schema that is a mapping of the conceptual model underlying the geologic data set to data structures that can be represented by an automated system (e.g relational tables, described in this report).
  3. A physical implementation schema that defines the organization of data into files, the detailed structure of the files, and the representation of data in the files. The file format dictates the software and hardware sys-tems that are compatible with the data.
  4. A projection and map horizon specification that describes how the three-dimensional location of features on the Earth is specified.
  5. The data instances contained in the files (locations of contacts and faults, map unit definitions, classifica-tion of areas to map units....).
  6. A set of definitions that specify the meaning of attributes applied to included data instances.


This report describes the logical and physical implementation of a database system for the representation of geologic features represented on geologic maps. It is assumed that the reader is familiar with the basics of the ESRI coverage data model and the use of ESRI ArcView GIS 3.x and Microsoft Access 97-2000 software.

This database implementation is a second-generation effort, and supercedes the data structure outlined in Richard and Thieme (1997). The design is an outgrowth from a proposed North American standard data model ("NADM") for geologic maps (Johnson et al., 1998), overseen by the North American Data Model Steering Committee ( In the course of implementing this database, the Johnson et al. (1998) model was found inadequate to allow inclusion of information in existing AZGS databases and for a complete representation of geologic information. Focus then shifted to the NADM "Cordlink" variant model (Brodaric et al., 1999) as a starting point. Various aspects of this model were also found insufficient or unsatisfying. The logical model presented here was evolved to reduce the number of tables and allow greater flexibility and logical consistency. The final implementation resembles the Johnson et al. (1998) NADM 4.3 model only in very general terms. The model builds on the design philosophy laid out in Richard (1998), the conceptual model described in Richard (1999), and the recent parallel development of an object-oriented data model by Brodaric and others (Brodaric, 2000; Brodaric and Hastings, 2001; Brodaric and Gahegan, 2000).

The core components of the model are:

  1. Classification Concept table(s). At the core of the model is a table or group of tables with similar structure that define terminology. The essential elements of these ClassificationConcept tables are a unique identifier, a name, and a definition/description. The unique identifier follows the global unique identifier scheme described below. The name is a string that allows human identification of the concept (e.g "basalt"), and the definition/description is a free text field that defines the term or describes its meaning precisely.
  2. Relationship tables. These are tables that link data instances. The meaning of the link is defined by a relation-ship type attribute. The data instances that fill roles in a relationship may be any individual classification concept, description, or relationship; the kinds of valid role fillers are determined by the relationship type attribute. Three sorts of relationship tables are included with different structure and application. Hierarchy Relationship tables define parent-child relationships in hierarchies; these may be taxonomic (IsA) or meronymic (Part-Whole). Simple Relationship tables link data instances, which may have a sequence; typically these link description parts (e.g image to rock description, age date to rock description, chemical analysis to location). The most complex relationships are Attributed Relationships, which allow an attribute value to be associated with the link, along with a sequencing index, and classification confidence and classification basis attributes.
  3. Description tables. These are tables tailored to particular kinds of descriptions. Spatial objects (points, lines, polygons...) are represented in a description table. In addition, the core model includes tables for structural measurements, rock samples, text, geochronologic ages, chemical substances, lithologic description, stratigraphic time, spatial objects, images, and measured quantity. Only the spatial object, structural measurement and rock sample tables are described here. Some description tables are linked to ClassificationConcepts directly through the sharing of a unique identifier, and provide a structured description to characterize the classification concept. Others provide descriptions of "real world" instances (a particular rock sample, a particular contact, a particular fault....).
  4. Map Visualization tables. These are a set of tables used to define map visualizations. This group includes three tables:

    1. Map View Definition table - specifies a title, author, design scale, map extent, symbolization scheme and classification scheme for the map;
    2. Map Legend - relates each symbol used in the map visualization to a classification concept;
    3. Cartographic Object table - defines the symbols used on the map in implementation-independent terms.

Three modes of defining assignment of symbols to spatial objects represented on the map are used. First, in this database, all spatial objects have a default classification attribute and a default cartographic object attribute. This default symbolization corresponds to that assigned by the original author of the map visualization (default visualization). Second, symbols may be associated with spatial objects through the map legend, (symbol - classification link) and a spatial classification attributed relationship (spatial object - classification link). This approach corresponds to the NADM 4.3 and Cordlink Variant approach. Finally, a general visualization links spatial objects to symbols through an attributed relationship link (symbolization scheme) whose type is the identifier for the map view definition. This final approach corresponds most closely to how map visualizations are actually generated from spatial data. The relationship attribute is the rotation to apply to structure measurement symbols, or, in the case of purely cartographic annotation symbols, the text string to display.

Identification Scheme

Unique identification of data instances in an internationally distributed data warehouse is achieved by partitioning responsibility for maintenance of unique identifiers. The Arizona Geological Survey uses a 3-component composite key, consisting of 3 long (4 byte) integers. At the top level, each organization providing data to the system must be assigned a NameSpace by the overall system manager. Note that a NameSpace is a ClassificationConcept. The name string and an integer identifier for the NameSpace must be globally unique. Within each NameSpace, every data file must have a unique integer identifier, and should have a unique name string. The system manager for the NameSpace must assign a unique identifier number to each data table, geographic data set (coverage, shapefile, etc.), image, text file, etc. that will be used by the system. Information about each data file (called a DataSet here) is stored in a central DataSet table maintained within each NameSpace. This table is analogous to a "catalog" in the Open GIS consortium model. The DataSet table must include a physical address (url) for each DataSet so that it can be located automatically when accessed. Within each DataSet, every data instance has a unique integer identifier number. The field containing this identifier is generally named with a string in the form "DataSetName" & "ID". In summary, the unique, global identifier for any data instance is a tuple consisting of 3 integers: {NameSpaceID, DataSetID, ObjectID}. Because this system has not been adopted outside the Arizona Geological Survey at present, the NameSpaceID is not explicitly included in tables here. Because some database software cannot join on multiple fields, implementation considerations require generating a single UniqueID from the DataSetID and ObjectID under some conditions. This is done using the formula ID = (DataSetID * 10000000) + ObjectID.


Feature level metadata is implemented by linking every data instance with an origin TrackingRecord, either as an attribute of the instance, or by inheriting origin tracking from the DataSet that contains the instance. The TrackingRecord defines a person, organization, and project (an "activity") that generated the data instance, along with a link to a data processing description for how the information was obtained and introduced to the database. Each TrackingRecord may be linked (through a SimpleRelationship) to one or more bibliographic citations.

Table and Field Naming Conventions

Tables and fields are named using strings with no spaces. The first letter of separate words in the name is capitalized, and no underscore separates words in the name. Typing an underscore is error-prone, and under many display conditions, underscores may be difficult to see. Because of limitations in ArcInfo (v8.0.1) and ArcView (v.3.2) software, field names in spatial data native tables are limited to 10 characters.


Three schema at the end of this document are presented to assist in understanding the data structure. Figure 1 is a simplified schema showing the tables necessary to represent the basic geologic data and cartography contained in a typical geologic map visualization. This schema includes the three kinds of relationship tables used to implement the general relationship structure. Metadata, implementation-specific description of cartographic objects, and description representation is not expanded in this schema. Figure 2 is a simplified schema showing the description of spatial objects. Fields in the spatial object table (AAT and PAT in ESRI terminology) represent a default visualization and classification scheme, which uses the geologic symbolization and classification of the original map author. This schema also includes some representation of description -- sample locations and structural measurements are included. It does not include the correlation tables necessary for building general relationships between objects, or any metadata tables. Figure 3 is a simplified schema showing a more in-depth (but not complete) view of the feature-level metadata implementation. All the tables shown on these schemata are described in this text, and the figures should be referenced throughout the following discussion.

Simplified schema for geoscience database implementation

    Simplified schema for geoscience database implementation
Figure 1. Simplified schema for geoscience database implementation. Map visualization represented by general relationship links between cartographic, classification, and spatial objects. Metadata, implementation-specific description of cartographic objects, and description representations are not expanded in this schema. Different line patterns used to facilitate tracing links.

[Click on image to open a larger version in a new window.]

    Figure 2. Simplified schema for geoscience database implementation. Metadata links not shown.

[Click on image to open a larger version in a new window.]

Schema for basic metadata relational implementation     Figure 3. Schema for basic metadata relational implementation.

[Click on image to open a larger version in a new window.]

The geologic and cartographic information in the database is organized into several ArcInfo coverages and ESRI shapefiles. The basic geology defined in the Geo coverage requires the point-line-polygon topology implemented by an ArcInfo coverage. Other spatial data may be in coverages or ESRI shapefiles. The Geo polygon and arc coverage contains the lines that represent geologic contacts and faults, and the associated polygons based on those lines that define the outcrop area of map units. The Pnt point data set contains the field observation stations that record things such as structural measurements and collected rock samples. The GeoLines line data set contains the geologic lines that do not define boundaries between rock units, such as concealed faults and fold hinge surface traces. The CartoLines line dataset contains cartographic lines, such as text lead-in lines. Last, the CartoPnts point dataset locates the cartographic point features used in the default map layout, such as text labels.

Every spatial object (point, line, or polygon) is uniquely identified by a compound primary key consisting of a source-file identifier, named DatasetID, and a unique identifier within that file, named "DataSetName" & ID (referred to as ObjectID here). The ArcInfo-assigned User-ID field, a seemingly good candidate for unique identifiers, is not immutable under build and clean operations on the data set. Therefore, ObjectID was added as a user-defined attribute, and the uniqueness constraint must be enforced by the user. The ObjectID values in the tables in this database should not be edited unless the user fully understands the data structure and the ramifications of editing the primary key in a relational database table. All points, lines, and polygons have a TrackingID attribute that joins with the TrackingRecord table to show the source origination and tracking information for each object. Geologic points and lines also have an Accuracy attribute that defines the location uncertainty in meters for the point or line. The compound object key, ObjectID and DatasetID, and the compound source tracking key, TrackingID and TrackingDS, plus the Accuracy attribute for geologic points and lines, are the minimal set of attributes fundamental to each spatial object.

A number of other attributes are also included in the coverage and shapefile tables to facilitate visualization of the geologic data in a default layout, and to allow querying against a default classification scheme equivalent to the original source map. These default values also make simple analyses of the map possible in non-relational database environments required by some users of AZGS data. The compound classification concept attribute, ConceptID and ConceptDS, defines the default classification of every object (Fault; Bedding; Surficial Map Unit...); the classification confidence attribute, CConf, provides a subjective measure of the confidence level for the classification of the object (Low; Standard...); and the compound cartographic object attribute, CartoObjID and CartoObjDS, defines the cartographic object used to symbolize each feature in the default visualization ("0.35pt. solid black line (24K)"; "Inclined bedding symbol - color black (24K)"; "PMS-1205"...). There is also a Label attribute used to store any specific labels or names associated with an object, such as unit names for geologic polygons, and a Name attribute that contains a brief description of each object for intelligibility. Polygon features have a map unit confidence attribute, MConf, that provides a subjective measure of the identification confidence of a polygon to a particular map unit (Low; Standard...). Point features also have a Rotate attribute, measured anticlockwise, starting from a compass azimuth of 90°, that defines the degree of rotation of graphical elements used for feature symbolization in the ArcView project. The rotation magnitude is specific to the graphical environment of ArcView 3.2 using the AZGSgeo.ttf true type font. Use of these geographic data sets with a different GIS platform and/or font may require that the rotation values in the Rotate attribute be recalculated.

In the summaries that follow, each table includes a compound unique identifier and a tracking record link. These universal fields are described here rather than in the tables below. Italicized words are names of other tables in the database.

Geology Coverage

The geospatial data for a particular geologic data set are represented in a set of Arc/Info coverages or ESRI shapefiles. These files contain spatial objects that represent geologic features at corresponding locations in the physical world. A minimum of two files are required to represent a geologic map -- one that represents geologic faults and contacts, and one that represents the distribution of map units. AZGS geologic map databases include these in one Arc/Info coverage with polygon and line topology, named Geo. A third file may be included to represent point-located data (structure measurements, rock descriptions, sample locations). AZGS geologic map databases include these in an Arc/Info point coverage or an ESRI point shapefile, named GeoPnts. A fourth file may be necessary to represent those geologic lines that do not define polygon topology (concealed faults, fold hinges, dikes, marker beds...). AZGS geologic map databases include these in one Arc/Info line coverage or ESRI line shapefile, named Geolines. Fields in the data table associated with this extra line-file are equivalent to the arc attributes in the Geo coverage, and are not described separately here.

The Geo coverage is a polygon and arc coverage that contains geologic lines that bound polygons (contacts, faults, mapping boundaries...), or represent surfaces that are discontinuous within polygons (faults that become buried or die out). The polygon topology defined by the lines in this coverage identifies the distribution of geologic map units.

Polygon Attributes

Arc Attributes

Point Coverage

The GeoPnt coverage is a point coverage that represents geologic spatial features located at a distinct point (structural measurement stations, rock samples collection stations...).

Point Attributes


Cartographic elements for the default map visualization of a particular geologic data set are included in a line and a point shapefile. Because the locations of points and lines in these shapefiles are chosen to provide cartographic clarity, the Accuracy and CConf fields are irrelevant and therefore not included. Otherwise the fields are the same as those in the geologic line and point coverages (geo.aat and pnt.pat), described above. The CartoLines shapefile contains the cartographic lines (text lead-in lines...) used in the default map visualization. The CartoPnts shapefile contains the cartographic points (text labels, fault symbols, fold geometry symbols...) used in the default map visualization. Table 8 lists some kinds of points included in this shape file, and Table 9 lists some associated cartographic object examples.

Table 8. Example cartographic object codes used in the GeoPnt.pat table.
Example cartographic object codes used in the GeoPnt.pat table

[Click HERE for a Word version.]

    Table 9. Example cartographic object codes used in the CartoPnts table.
Example cartographic object codes used in the CartoPnts table.

[Click HERE for a Word version.]


The majority of geologic information is stored in thematic databases specific to particular kinds of geoscience information, and linked to the Spatial Object tables by explicit links, or through relationship links. Information in these thematic tables includes structural measurements, rock sample descriptions, text descriptions, geochronologic age data, chemical and isotopic analytical data, lithologic descriptions, stratigraphic time scales, and images. This thematic geoscience part of the database is the least developed aspect of the system at present. Only basic map unit description, structural measurement and rock sample tables are described here. The thematic tables are specific to the particular geologic data set. These tables are included in a Microsoft Access database associated with each geologic data set.

Map Unit Table

The MapUnits table defines the map units used to classify polygons in the Geo coverage. In a more complete implementation, the MapUnitID would be a link to a more complete description in a series of tables for lithology, rock volume, and geologic surface description (geoscience description component of database, see Table 1).

Database Table Fields

Samples Table

The Samples table contains location and description information for rock samples collected within the extent of the geologic data set. The inclusion of both the UTM coordinates for the sample location and a link to a spatial object representing the sample location is redundant, but both forms of location are included for reliability. If the link with the spatial object data set is corrupted, the Samples table still contains sufficient information to locate the sample. Likewise, the sample table can be exported for data exchange without including a data set with location spatial objects.

Database Table Fields

Structural Measurement Data Table

The StructureData table contains values that define the orientation of structural features. The inclusion of both the UTM coordinates for the station location and a link to a spatial object representing the station location is redundant, but both forms of location are included for reliability. If the link with the spatial object data set is corrupted, the StructureData table still contains sufficient information to locate the station. Likewise, the StructureData table can be exported for data exchange without including a data set with location spatial objects. A separate correlation table to link stations with locations is unnecessary because each station has a unique location.

Database Table Fields


The lookup tables defined below contain supporting data maintained by the Arizona Geological Survey to support all databases within the organization. These tables are included as a Microsoft Access database. By default, each data set below references a table that is included in the Arizona Geological Survey namespace.

Classification Tables

Classification Concept Table
The ClassificationConcept table is a collection of terminology definitions - a term with a definition. These terms are used to classify other objects in all parts of the database. A unique identifier (ConceptID - DatasetID pair) identifies each concept. Thus the name of the concept may be changed without updating other links. The Arizona Geological Survey geologic information system has separate classification concept tables that are specific to different components of the system (e.g. Rock Unit Lexicon, Standard lithologic terms, etc.). Each of these classification concept tables has its own data set identifier defined in the DataSetAz table.

Database Table Fields

Relationship Tables

Three sorts of relationship tables are used for representing semantic links between objects in the database (see Relationship Table Discussion, above). In the Arizona Geological Survey geologic information system, each component of the system (cartography, rock unit lexicon, standard lithology, geochronology...) has relationship tables specific to that sub-domain. A particular geologic data set may include several different relationship tables of each of the types described below, each with its own DataSetID defined in the DataSetAz table.

Attributed Relationship Table
The AttributedRelationship table is used for representing relationships between objects in the database, i.e. for linking instances of two entities in which each relationship instance is assigned one or more attributes. This table is constructed to allow up to 5 attributes: CConf (Concept Confidence), CBasis (Concept Basis), StringValue (any string), Number Value (any number), or Attribute (a link to another object in the database). The RelTypeID link defines the semantics of the relationship links. Relationship constraints on RelType specify which attributes may have values, and specify the domains of those values. Examples of attributed relationships include geologic classification of spatial objects, and various kinds of fractional analyses (e.g. chemical analysis, modal mineral analysis, grain size distribution).

Database Table Fields

Hierarchy Relationship Table
The HierarchyRelationship table represents parent-child relationships. Multiple tree hierarchies may be represented, each identified by a HierarchyType - a classification concept that defines the nature of the hierarchy. For implementation simplicity, a hierarchy is represented in this table as a set of links between each parent and all the child objects beneath it in the hierarchy tree (its transitive closure). The depth of any child object in the tree is determined by the number of parent objects linked to it. This representation makes response to queries that require all kinds (sub types) of a thing (e.g. "all spatial objects", "all map units") simple to execute. Currently, each child has only one parent.

Database Table Fields

Simple Relationship Table
This table is used to represent relationships that link instances of any two objects in which no uncertainty is involved and the relationship has no attributes. Examples include aggregations of parts, and linking SpatialObjects to Cartographic-Objects for symbolization.

Database Table Fields

Metadata Tables

Activities Table
The Activities table is a link to an activity responsible for update of, or addition to, the database. An activity is a particular person, working for a particular organization, under the auspices of a particular project.

Database Table Fields

Bibliographic Citations Table (AZgeoBibLinkTable)
The AZgeoBibLinkTable table is derived from the Arizona Geological Survey bibliographic data base (AzGeoBib, Trapp et al. (1996), DataSetID = 4 in the DataSetAZ table), and provides a mechanism for citing published literature. In this database citations are related to tracking records through the MetadataRelationship table. This derivative table is included to replace links to the full AzGeoBib database.

Database Table Fields

DataSetAZ Table
The DataSetAZ table is a catalog of the data sets within the Arizona Geological Survey namespace. A data set is any collection of data that is held in an individual file or table. Examples include individual ArcInfo coverages, ESRI shapefiles, tables in Microsoft Access databases, dBase tables in individual .dbf files, and files containing images (e.g tiff, jpeg). The contents of the DataSetAZ table define the "Arizona Geological Survey" namespace. This table is analogous to an Open GIS Consortium "Catalog." In the more complete metadata implementation, a geographic dataset is associated with map extent, projection, and map horizon objects that define the original geometry of the spatial data.

Database Table Fields

Metadata Relationship Table
The MetadataRelationship table is a SimpleRelationship table that provides a general mechanism for semantic links between metadata instances. A RelType (relationship type) identifier links to a ClassificationConcept that defines the semantics of the relationship. Constraints on kinds of objects that may play the first and second role, and the number of fillers allowed for each role, will eventually be specified by a ValidRelationshipConstraint data structure, but this part of the database is currently being revised and is not described here. This table may be used to implement a many-to-many join between tracking records and citations, project hierarchy (large project with subprojects), organization successor (when an organization changes name), organization aggregation (to represent individual departments as part of a larger organization), StartDate and EndDate links between Person-Organization affiliations and a metadata dates entity, PersonOrg-ContactInformation links to allow multiple contact addresses and types (phone, internet, surface mail...), and Object-LogEntries to allow multiple tracking records to be related to any object, to track revisions, comments, etc. See description of SimpleRelationship table for fields in this table.

Tracking Record Table
The TrackingRecord table keeps a record of the intellectual and physical sources for objects and data by defining links to tables that describe the processes and activities through which data was created. Data objects that have a TrackingID/TrackingDS link are directly linked to a TrackingRecord of TrackingRecordType "OriginTracking" that provides information on the original source of the object. The TrackingRecord data structure includes a link to an Activity (tuple of person, organization, project) responsible for the tracked event, and a link to a ProcessingMethod that describes the procedure used to represent the feature in digital form. Links to citations for publications relevant to the origin of the information are constructed through a MetaDataRelationship link (see MetadataRelationship table).

Tracking records may also be LogEntries that document updates or comments related to any data object. LogEntry tracking records are linked to data objects using a MetaDataRelationship link, allowing a many-to-many relationship between log entries and data objects.

Database Table Fields

Cartographic Tables

Cartographic Object Table
The CartographicObject table is an implementation-independent representation of symbols used to display points, lines, polygons, and text on a map visualization. This is done by defining links to tables that provide implementation-dependent descriptions of graphical objects used for symbolization. Graphical object tables in this database are designed to describe symbology for ArcView 3.2 running in a Microsoft Windows environment.

Individual cartographic objects may consist of several graphical objects stacked according to the sequence attribute in the table, with the lowest sequence symbol overlain by subsequent symbols in the sequence. A CartographicObject defines links to tables that define implementation-specific graphical objects and colors; these tables are not explained here. The DataSetID for the linked tables serves to indicate what sort of graphical element is being specified.

Geologic structure symbols present a special problem, because a standard strike-and-dip symbol is considered to be the same CartographicObject, irrespective of its orientation, and while the same symbol is used for each measurement location (SpatialObject), the symbol is rotated depending on a value (the azimuth) specific to a StructureMeasurement associated with that SpatialObject. SpatialObject must be joined with CartographicObject through an attributed relationship(s) that includes the rotation value as its attribute. The CartoObjType may be used to determine if the symbolization depends on an AttributedRelationship.

Database Table Fields

Map Legend Table
The MapLegend table contains relationship links between a ClassificationConcept and an implementation-independent CartographicObject used to symbolize objects belonging to the class. A particular map legend may contain only one instance of each symbol included, but different symbols may correspond to the same classification concept (e.g symbols for horizontal, inclined, vertical, and overturned planar bedding). The MapLegend table assigns a Name, Label, and Description for objects of that class which are used to generate the explanation to display on the map. The Sequence field orders items in the legend. Legend items may be present that have no corresponding classification concept; these typically act as headings. The compound key for the MapLegend table is the tuple {MapLegendID, DataSetID, Sequence}. Hierarchy in the legend is represented by a HierarchyRelationship with RelTypeID = MapLegendID.

Database Table Fields

Map View Definition Table
The MapViewDefinition table defines a Title, Description, Extent, Projection, DesignScale, MapHorizon, ClassificationScheme and MapLegend to use for a particular MapView. A MapView is a collection of SpatialObjects within a bounded area (the Extent), classified using a particular ClassificationScheme, and symbolized using a particular MapLegend. The MapView does not necessarily use all the items in the MapLegend, or all the SpatialObjects classified under the ClassificationScheme. Every ClassificationConcept in the ClassificationScheme that is related to a Spatial-Object included in the MapView must have a CartographicObject assigned by the MapLegend associated with the MapView.

SimpleMapView - All SpatialObjects symbolized in the MapView are entirely within the MapExtent, and the set of CartographicObjects in the MapLegend is the same as the set of Cartographic-Objects used to symbolize spatial objects in the view.

GeneralMapView - SpatialObjects may come from different DataSets that may have extents different from the MapView extent, and the MapLegend may include Cartographic-Objects not used in the MapView. SpatialObjects symbolized in the MapView must be clipped to the MapExtent, and the MapLegend must be filtered to select only the items that appear in the MapView.

The ViewSchemeType in the MapViewDefinition table determines how the MapView is constructed. In addition to specifying if the view is a GeneralMapView or SimpleMapView, the ViewSchemeType also varies along a second dimension based on how the link between Cartographic-Objects and SpatialObjects is defined, as follows:

DefaultMapView - Represents a default visualization of a geologic data set. Default ClassificationConcepts, CartographicObjects, necessary CartographicObject attributes (e.g rotation for strike-and-dip symbols) and feature-linked annotation (polygon labels, dip values) are assigned using fields embedded in the SpatialObject tables. SimpleRelationship aggregates the DataSets containing the SpatialObjects through a simple relationship of type MapViewID; sequence attribute establishes display order for DataSets. All DataSets contain data within the same MapExtent. The MapLegend can be produced through a query that returns the union of unique ClassificationObject/CartographicObject pairs included in the records for all Spatial-Objects represented in the view. MapLegendID and ClassSchemeID are not required, but a predefined MapLegend is necessary to structure the MapLegend display, display order, and explanatory name, label and text (ClassName, ClassLabel, ClassDesc) for features; otherwise the default legend layout for the particular GIS implementation will be used.
DirectMapView - MapViewID is a RelationshipType for a SymbolizationScheme Relationship linking SpatialObject with CartographicObject, and MapLegendID identifies the appropriate Map-Legend objects. All CartographicObjects used must be included in the MapLegend. The ClassificationSchemeID link in the MapViewDefinition identifies the classification scheme used as the basis for assigning symbols to spatial objects. The direct scheme is necessary for individually varying symbolization (e.g structure symbols), and also allows for map generalization in which an object classified in the same way may be symbolized differently.
NADM43MapView - SpatialObjects are linked with ClassificationConcepts through a ClassificationScheme specified by the MapViewDefinition, and ClassificationConcepts are linked with CartographicObjects through the MapLegend. Assignment of CartographicObjects to Spatial-Objects requires two joins, and the ClassificationConcepts used are conceptually equivalent to Cartographic-Objects because, in order to symbolize an object differently, it must be classified differently. Thus, in order to rotate structure symbols to the correct display azimuth, ClassificationConcepts for each azimuth must be generated, or the azimuth attribute of the data to symbolize must be propagated from the structural measurement table, through the SpatialObject, ClassificationScheme link (Spatial-Object-Classification), and MapLegend link (Classification-CartographicObject).

Database Table Fields


This database design has been developed over several years with the help of J.P. Thieme and S.M. Kneale. Discussions with Boyan Brodaric, Jon Matti, Gary Raines, and Bruce Johnson have sharpened my thinking and helped develop ideas for this design.


Bernknopf, R. L., Brookshire, D. S., Soller, D. R., McKee, M. J., Sutter, J. F., Matti, J. C., and Campbell, R. H., 1993, Societal Value of Geologic Maps: Washington, D.C., U.S. Geological Survey Circular 1111, 53 pages.

Brodaric, B., 2000, Digital Geological Knowledge: From the Field to the Map to the Internet, in Soller, David R., Ed., Digital Mapping Techniques '00 -- Workshop Proceedings, U.S. Geological Survey Open-File Report 00-325, p. 3-12,

Brodaric, B., and Gahegan, M., 2000, Geoscience Map Data Models, Open Systems GIS and Semantics, Proceedings, GeoCanada2000-The Millennium Geoscience Summit, Calgary, Alberta, p. 7. available at, Accessed June 14, 2001.

Brodaric, B. and Hastings, J., 2001, Evolution of an Object-Oriented, NADM-Based Data Model Prototype for the USGS National Geologic Map Database Project [web page, abstract]: Annual Conference of the International Association for Mathematical Geology, IAMG2001, Cancun, Mexico, Available at Conferences/IAMG/Sessions/I/brodaric.html, Accessed June 14, 2001.

Brodaric, B., Journeay, M., Talwar, S., and others, June 18, 1999, CordLink Digital Library Geologic Map Data Model Version 5.2 (Web Page), Available at English/dm52.pdf, Accessed June 13, 2001.

Johnson, B. R., Brodaric, B., and Raines, G. L., 1998, Digital Geologic Maps Data Model, V. 4.3 (Web Page): Available at, U.S. Geological Survey.

Palmer, A.R., compiler, 1983, The Decade of North American Geology 1983 Geologic Time Scale: Geology, v. 11, no. 9, p. 503-504.

Reynolds, S. J., Florence, F. P., Welty, J. W., and others, 1986, Compilation of Radiometric Age Determinations in Arizona: Tucson, Arizona Bureau of Geology and Mineral Technology Bulletin 197, 258 pages.

Richard, S. M., 2000, Proposal for Authorship and Citation Guidelines for Geologic Data Sets and Map Images in the Era of Digital Publication, in D.R. Soller, ed., Digital Mapping Techniques 2000--Workshop Proceedings: U.S. Geological Survey Open-File Report 00-325, p. 159-168,

Richard, S. M. 1999, Geologic concept modeling, with examples for lithology and some other basic geoscience features, in D.R. Soller, ed., Digital Mapping Techniques 1999--Workshop Proceedings: U.S. Geological Survey Open-File Report 99-386, p. 59-75,

Richard, S. M., 1998, Digital Geologic Database Model (Web Page): Available at, Accessed June 13, 2001.

Richard, S. M., and Thieme, J. P., 1997, Data structure for Arizona Geological Survey digital geologic maps: Tucson, Arizona Geological Survey Open-File Report 97-5, 15 p.

Soller, D. R., 1997, Proceedings of a Workshop on Digital Mapping Techniques: Methods for Geologic Map Data Capture, Management, and Publication, Lawrence, KS: U.S. Geological Survey Open-File Report 97-269, 120 p.,

Soller, D. R., 1998, Digital Mapping Techniques '98 -- Workshop Proceeding, Champaign, IL: U.S. Geological Survey Open-File Report 98-487, 134 p.,

Soller, D. R., 1999, Digital Mapping Techniques '99 -- Workshop Proceedings, Madison, WI: U.S. Geological Survey Open-File Report 99-386, 190 p.,

Soller, D. R., 2000, Digital Mapping Techniques '00 -- Workshop Proceedings, Lexington, KY: U.S. Geological Survey Open-File Report 00-325, 210 p.,

Trapp, R. A., and Reynolds, S. J., 1998, Physiographic areas in Arizona used by the Arizona Geological Survey: Tucson, Arizona Geol. Survey, Digital Information Series DI-10, 4 pages, 1 floppy disc.

Trapp, R. A., Schmidt, N., and Reynolds, S. J., 1996, AZGEOBIB, Version 2.1: A List of References on the Geology of Arizona: Arizona Geological Survey Open-File Report OFR-96-01, p. 308.

RETURN TO Contents
National Cooperative Geologic Mapping Program | Geologic Division | Open-File Reports
U.S. Department of the Interior, U.S. Geological Survey
Maintained by David R. Soller
Last modified: 18:24:54 Wed 07 Dec 2016
Privacy statement | General disclaimer | Accessibility