Geography Reference
In-Depth Information
Specific Esri Spatial Vector Data Storage Mechanisms
Let's now move to the specifics of the storage and manipulation of spatial data based on Esri software.
Esri is decades old and both its computer programs and storage schema have evolved considerably. As
mentioned, the original, elegant data model—the coverage—that is based on using arcs to represent
linear and areal features is virtually obsolete. You should know how to convert coverages (which you
may come across because a lot of data sets are in this format) to geodatabases, but that's all you will need
to know. You may recall that making this conversion was presented in Chapter 1.
Geodatabases are the current “coin of the realm” in Esri software. All the tools being developed deal
with geodatabases. A plethora of topology rules and topology fixes accompany geodatbases. Ideally,
all coverages and shapefiles would be converted to geodatabase form. However, that is a monumental
undertaking, of which, if you continue in the GIS field, you may be a part.
Sets of spatial data in Esri are primarily stored in geodatbases, shapefiles, TINs, (and “super TINS” called
Terrains). The interactions among these types are fairly complex. Let's start with the primary ways to
store spatial data based on vectors and rasters: geodatabases.
The Geodatabase Data Structure
Esri developed the geodatabase data model for the following reasons:
To take advantage of increased computing power, data storage, and modern relational database
management systems (RDBMSs)
Because ideas of how to store geographic data have become more refined
To “umbrella-ize” the different forms of spatial data storage: vector, raster, and, terrains (but not
TINs, unless they are converted to terrains),
To permit the use of “objects” that depict real-world entities, in terms of both description (which is
still done with attributes) and behavior
Geodatabase Software
Despite the idea of a geodatabase—storing all of geometry, a spatial reference system, attributes, and
behavioral rules for data in a single relational database management system—things are still not simple.
Esri software has to “partner” with existing RDBMSs, so the conventions for dealing with spatial data must
conform to the different general conventions of these RDBMSs. For these and other reasons, there are two
Esri geodatabase flavors for single-user ArcGIS: Personal and File. For multiuser Esri software there is the
ArcSDE (Spatial Database Engine). (Multiuser ArcGIS has Desktop, Workgroup, and Enterprise versions.)
Personal Geodatabases: They depend on Microsoft Access RDBMS. The datasets are stored within a
data file, which is limited in size to 2GB. This was Esri's first version of storing both the geometry and
the attributes in a single database system. The extension on the name is .mdb, standing for “Microsoft
database.” While personal geodatabases suffer from several disadvantages compared to file geodatabases
(discussed below), they also have their strengths. If you plan to remain in the Windows operating system
environment, and want to search and work with the ArcGIS software that stores attribute tables with
Microsoft Access, personal databases can be quite satisfactory.
 
Search WWH ::




Custom Search