The Geodatabase

The geodatabase is so seductive!

And like anything that presents itself as seductive, it may require a considerable investment of time and resources. How far you can go with it depends on your license level. As with so many relationships here in Asheville, developing a "deeper understanding" may call for some sort of sequestered retreat or a seminar with like minded people on the same quest.

As with scientology, with ESRI you can get a taste of enlightenment from several free one hour online courses. Tne next steps are in the $50 to 500 range. After that, the possibilities are limitless.

The geodatabase is worth getting to know and work with.

What tools will it take to work with geodatabases?

Note on training exercises

We started the Building a Geodatabase and Editing Geodatabases exercises in the AB Tech lab using an ArcInfo license. The database we created was File Geodatabase. This created problems when I tried to continue working at home using an ArcView license. According to ESRI we can use the student ArcView license IF we work with a Personal Geodatabase.

Having access to SQL Server Express would be nice for students.

License levels for ArcGIS 9.3

Arcview student 12 month ($100 includes extensions)
Arcview student never expires, service packs but no upgrade ($250 + $150 per extension)
Arcview ($2500 + $1500 per extension)
ArcEditor ($7000 + $1500 per extension)
ArcInfo ($14,000 + $1500 per extension)
ArcInfo Hardware bundle* (>=$9,998+ $1500 per extension)

*http://gis.esri.com/partners/hw_promo/hw_promo.cfm#results includes ArcInfo license

ArcGIS Desktop 9.3 Functionality Matrix

ESRI Data Models

The geospatial data model evolved:

• as a metaphor to old analog methods
• along with database technology
• in respsponse to faster cheaper computer hardware

CAD

CAD/CAM came around before pong was a video game. It was a way of capturing manual drafting techniques in a digital format. A CAD file is a digital representation of a design or existing object.

CAD charachteristics:

• Primarily design and drawing tool
• Developed to automate drafting functions
• Features represented by graphic symbols Graphics driven
• Often stored in “local” coordinate system
• Layers often not well organized
• Attribute info not commonly stored or stored as annotation

ArcIfo Coverage

The coverage evolved out of relational database technology with spatial functions being user defined (or ESRI defined) extensions of the relational database model.

Georelational data model represents geographic features as interrelated Spatial and attribute data. In 1981, ESRI introduced ArcGIS and a georelational data model called “Coverage” as the data format for ArcInfo. The key features of the Coverage are:

• Spatial data stored as indexed binary files which are optimized for access and display of geographic data. Attributes are stored in Info tables with each row linked to a single related spatial feature by a common identifier.
• Topological relationships between vector features can be stored and updated. A line is defined as a sequence of points or nodes. By inference lines with common nodes are connected. Fields for which polygon is to the right and left of a line can define polygon boundaries.

Disadvantages to the ArcInfo Coverage:

1. The ArcInfo coverage represents features as having generic behaviors so that lines representing roads and streams have the same behaviors and unique behaviors such as flow direction and flow rates for rivers could not be added.
2. Topological features (polygons, regions and routes) are not ready to use until topology is explicitly built and must be rebuilt after editing using the “clean” and “build functions”
3. Coverage was limited to single user editing due to need to synchronize topological graph with feature geometries. Large data sets had to be broken into smaller grids or tiles which required which required extra work and time.

System generated and maintained fields - For points and polygons the (PAT and PAT) tables have Area, Perimeter, coverName#, and coverName-ID

Shape files

The shape file was created to meet a need for faster display and querying of geospatial data by users who neither needed to nor wanted to work with and maintain topological graphs. Faster cheaper hardware, especially color displays and compact cheap data storage along with GUI interface softeware began opening up GIS data to a larger audience. Shape files were introduced along with ArcView, an extended viewer with analysis and cartographic capabilities.

A shapefile is a set of at least three files (shp, shx, dbf) that store and index spatial feature geometry as a set of vector coordinates and attributes in a file with dBASE format, with optional files that can store coordinate system information, metadata, and other information. Single features are noncontiguous and can overlap.

Disadvantages of shapefiles:

• Data redundancy in shared segments between adjacent polygons
• Shapefiles do not store topology, model only generic features, do not store relationships with features
• have limited ability to enforce topological data integrity

System generated and maintained fields - FID, Shape, ID (or OID)

Geodatabase

A geodatabase is an object relational container for storing spatial data along with attribute data and the relationships that exist among them.

System generated and maintained fields - OBJECTID*, SHAPE*, SHAPE _Length, SHAPE _Area

More on the Geodatabase in other posts

Campus Buildings Layers- Endless Editing

The buildings polygons from CAD files were just a starting point for developing a campus map. Next up, buildings in 3-D. To do that we used MS Virtual Earth's bird's eye view to estimate height of buildings by number of stories.

Buildings at AB Tech are like clusters of boxes, each with a different height. While the roof of a building may be almost flat, the building is made up of sections of varying heights. Many of the buildings had to be divided into a set of smaller polygons so each section could be extruded to a relative height.

This required a combination of:

• splitting existing polygons
  
• digitizing sub-polygons within existing building polygons


• digitizing new buildings polygons where a building was incomplete.
  

The aerial images were not quite in registration with the buildings polygons. They were clearly at the same scale, and were probably the source for the CAD files. This was tedious............This was occassion for reviewing editing techniques.

After working on the buildings layer, I brought in data on impervious surfaces from the City of Asheville geodatabase. This provided a good start to a paved roads and paved parking lots layer. On top of the campus aerial photo, I overlaid the impervious surfaces layer with transparency set at 66% so I could see everything. This made selecting roads and parking lots fairly easy. Some of the lots and roads had been changed by new construction and needed editing or digitizing.

The map looks good with buildings, roads, and parking lots along with utility poles and sidewalks fromt the CAD layers. Some decent sidewalks polygons could come out of the impervious surfaces. Oops outa time.

Georeferencing CAD data ArcGIS

August 27, 2008






We worked with a CAD dataset fo the AB-Tech campus that was not in real world coordinates. To set the coordinates for the data frame, as the first layer, we added campus buildings layer (this layer is 'defined' as State Plane NAD 83 feet - therefore the data frame will be defined as SP NAD 83 Feet).

1. first layer added to data frame defines coordinate system, campus buildings (CS = NC State Plane NAD83 feet)


2. open CAD file ABTmap_full.dwg(CS = not real world)


3. turn on Georeferencing toolbar


4. Zoom to campus buildings layer


5. Make sure CAD polyline layer is selected in georeferencing toolbar


6. Georeferencing ==> 'Fit to Display'


7. two point transformation on CAD layer


8. select from CAD polyline layer using field "Layer" for valeu "BUILDING"


9. Export selected data using coordinate system of data frame, to a geodatabase


10. Repeat export to create layers for

• utility poles (also from points layer)


• Roads paved
  
• Roads unpaved
  
• Sidewalks

The CAD data is organized by attributes related to graphic display.

Where ever you go, there you are

Monday, 18 - august

Here we are again at the start of a GIS course. This time it is GIS Data Models. This feels familiar. It's like a continuation of GIS 121 with Steve, Tammy, Andrea and myself from spring semester, plus Willie and Cameron.

It feels like I never left after finals in May. That may be because I took CIS 115 this summer and spent more than a few hours with Pete installing ArcGIS 9.3 with ArcSDE Server.

As with everything in life that is worthwhile, and especially GIS, persistence is the key to getting ArcSDE up and running.

There were several hurdles:

• setting up a GIS lab computer with ArcGIS 9.3 and getting the network logions right
• intalling SQL Server 2005 on a server in the closet downstairs
• installing ArcSDE on the server
• debugging the ArcSDE installation
• debugging to logon process for SQL Server, Windows and ArcSDE

A special thanks to Tom in IT network support. His patience and assistance were essential.

When working with ArcSDE, working well with very capable persons who provide network support is important.

I am excited about learning to work with ArcSDE geodatabases. Support for multiuser editing, versioning and using SQL Server 2005 as an RBMS makes this a "real world" learning lab. This satisfies two of the major reasons why I returned to school:

1. Learn most current version of ArcGIS 9.x
2. Work with client/server model