ArcGIS Data Models

Geodatabase Schemas: Data Models for GIS Users

ESRI has established a set of "best practices" geodatabase designs for various application domains. These database designs are intended to help GIS users rapidly become productive with the geodatabase and to share "what really works" among users and our developer communities.

What's Included in an ArcGIS Data Model?

The content of the data models can vary, but each data model should include:

• A case study implementation for a selected user site, including a small sample database
      
• A geodatabase template that enables users to import the data model as a template on which to base their system
      
• A white paper explaining the design
      
• A data model poster
      
• Tips and Tricks documents on how to employ the data model from the case study and how to employ it in your work.

Some of the data models also have important community-based Web links for particular domains to allow collaboration to develop designs and concepts.

Using Common Key Elements in Data Models

Series of thematic layers:  A GIS requires a minimum series of geographic-or thematic-layers of information. Specifying these layers is part of the conceptual design.

Common spatial representations:  Specify how each layer is to be represented as a collection of spatial datasets. Specific representations for each dataset will be defined as a collection of points, lines, polygons, map text (annotation), rasters, tabular attributes, and so on. These primarily consist of commonly agreed-upon feature and image representations with a common set of names and terms.

Minimum set of attributes:  Ideally, many of these common attributes would use similar terms and classifications. For example, agreed-upon classifications, such as road class, soil types, and land-use/land-cover classifications, could be documented and consistently applied in multiple databases. The adoption of common terms and classifications can simplify interoperability implementation.

Primary set of integrity rules and spatial relationships:  The primary set of integrity rules and spatial relationships should be specified to define the spatial behavior of features for each dataset. For example, road lines must connect at endpoints and not self-overlap (the same is true for hydrology and contour lines), connected contour lines must have consistent elevations, and county boundaries (and similar administrative units) cannot overlap one another and must nest within state boundaries.

For many users, understanding the integrity specifications of a dataset is critical in assessing its suitability for a particular application.

Map layout and map layer templates:  Data models should include the definition of common map products and their associated map scales. A common cartographic standard for map sheets, symbols, and label specifications will be important for many application domains.

Metadata templates:  A common metadata description that adheres to appropriate Federal Geographic Data Committee (FGDC) or ISO standards for key layers of information. This is essential for the consistent dissemination of information sets across hundreds of organizations. While every dataset covers its particular geography, common data layers should be consistently described among participating organizations.

Each dataset's adherence to a specific data model guideline should also be noted (e.g., hydrology data model) in the metadata. The data model can include a metadata template for each layer, with much of the general information already documented. The metadata templates can be copied, and instance-specific information can be added. This practice helps ensure that much of the metadata content for each implementation will be consistent from site to site. This will result in adoption of common terms and keywords to improve the search and discovery process.

Extraction guidelines and methods:  All datasets are a sample of reality. Common data models require a set of rules for data capture and extraction. Some considerations include  

• Is the collection of building footprints meant to capture all buildings or only prominent buildings?
• Will the stream representations be derived from a detailed digital elevation model to capture all drainage lines longer than 1,000 meters?
• What is the explicit methodology used to capture the features?
• What are the source materials and what is their vintage?

Understanding these properties for datasets is critical in determining appropriate uses of the data.

Results of case study implementation:  Case studies illustrating how an organization solved particular data model requirements are useful in providing insight to other users.

Sharing GIS Data Models

ESRI ensures that the ArcGIS data models implement relevant standards as they evolve. For example, the land records and the ArcHydro data models are based on strong standards work in these communities over the past decade. ESRI also monitors and participates in many standards-based efforts at ISO, OGC, FGDC, GeoSpatial One-Stop, ANSI, and so forth. ESRI incorporates appropriate standards into our data model work as these efforts evolve.

Each ArcGIS data model suggests using commonly adopted spatial representations (e.g., points, lines, and polygons), classifications, and map layer specifications that can be implemented in any GIS. Each data model specifies the commonly used integrity rules for key data layers and feature classes. These aspects of the ArcGIS data models can be widely adopted regardless of the user's system architecture. The ArcGIS data model templates implement what our users have found to be widely adopted best practices for building systems that really work. For individual data model descriptions and downloads, visit support.esri.com/datamodels.