Database Design and Automation

Data Conversion and Data Quality Control

Brian Ragland, Long Beach Water Department

Albert Lin

Long Beach Water Department is converting hard-copy maps to ArcInfo digital data for its potable water, reclaimed water, and wastewater systems. The conversion involved proper interpretation of maps; data standardization; database design and construction; data transformation among ArcInfo, AutoCAD, and Oracle software; as well as data quality control checking and establishment of realistic acceptance criteria.

Spatial Data Coordinate Quality and the ArcInfo Data Model

David Hansen, U.S. Bureau of Reclamation

The Content Standards for Digital Geospatial Metadata calls for an assessment of the accuracy of the coordinate values of the features stored in a GIS system in real-world coordinates. The focus of this paper is on the identification of factors affecting the accuracy of features represented in GIS for the development of a report on the accuracy of those coordinate values. Primary factors affecting the accuracy of the stored coordinate values include:

Methods used in locating the feature in a coordinate system.
Methods used in representing the feature on source material used to digitally capture the feature location.
The method used to digitally capture the feature location.
The map projection system used to store and display the feature.
Coordinate control (TICs) used for the projection system.
Machine processing precision used in the manipulation and storage of the coordinate values.

Errors associated with the development of digital data are described in Esri Proceedings on Map Projections and Coordinate Management by Kopp and Kneler. Estimating the relative magnitude of the errors introduced into a GIS data theme or layer is very valuable in the development and application of that data. Some of these errors are often not uniform across the spatial extent of the GIS data theme. In addition, more information is often available about the coordinate values of some features that are within a theme than about all features within a theme. This information represents key elements in the lineage of the data set. It represents information about the coordinate system and processing history of:

The source of the feature (map source, remote sensing source, GPS, etc.).
The base coordinate system used for control on the project.
The initial capture, manipulation, and storage of the features for the GIS data theme.

Although data elements exist in the data quality section of the Content Standards for Digital Geospatial Metadata, the relationship of these elements to estimates of coordinate accuracy is not described. Lineage information is captured for the development and initial manipulation of the GIS data theme, but usually not at the feature level. In addition, lineage for the base coordinate system used for control and often for the source material for the feature is not part of the metadata for the digital data set.This paper describes information which is useful for developing estimates of coordinate accuracy of a GIS data layer at the feature level within the context of the Content Standards for Digital Geospatial Metadata. As the example, it use digital data themes captured from base cartographic maps, aerial photography, and GPS.

Digital Elevation Models: Issues of Data Accuracy and Applications

Jay Lee, Kent State University

Digital elevation models (DEMs) have been used as the basis for many GIS applications. Two recent trends have prompted for the discussion to be presented in this paper. First, there are more DEMs being created and made available through Internet and WWW. Secondly, GIS software, such as ArcInfo GRID, now possesses a wide variety of functions for processing and analyzing DEM data with very easy-to-use interface/commands. As a result, even novice users are able to manipulate DEMs without much training. Most alarming is that users are often using DEMs without knowing the impact that data quality might have on the results of their analysis. In this paper, we present results of our investigation on the impact that data accuracy has on the results of applying DEM data. We used a 1:24,000 DEM released by USGS as the test dataset. We conducted a series of simulations in which errors of elevation measurements were added to the original DEM to materialize the accuracy statement given by USGS on their DEM products. We then applied procedures for extracting drainage networks with the simulated DEM data by using ArcInfo GRID. The results were used to construct a set of quantitative parameters to give drainage densities, network densities and others. The comparison among these parameters indicates that results of terrain analysis are very sensitive to the accuracy of DEM data.

QA/QC on a Shoestring

Mark A. Tuttle, Oak Ridge National Laboratory; Ed Bright, Yang Cheng, Lockheed Martin Energy Systems; Phil Colemann; Rick Goeltz; Marsha Henley

Over the past two years, the GIS and Computer Modeling Group at Oak Ridge National Laboratory, has been engaged in treating planimetric, topographic, digital ortho photograph, digital elevation, land use/land cover, and flood plain databases for the Oak Ridge Reservation (ORR), Portsmouth Gaseous Diffusion Plant, and Paducah Gaseous Diffusion Plant. These sites are located in Tennessee, Ohio, and Kentucky respectively. In total, the spatial databases cover 165, 907 acres. The production process is compounded by the delivery requirement of multiple formats for both the raster data and the vector data. The project is further complicated by the number of participants: two subcontractors, three groups within one federal agency, and the receiving agency. Ensuring the highest possible quality product within the constraints of financial limitations and limited time frame is more challenging than ever before. This paper examines the implementation of QA/QC techniques to a very diverse spatial data conversion project, with particular emphasis on the technical aspects. In addition, administrative aspects of data tracking and work flow are discussed. Only through careful consideration of all facets of both the technical requirements and administrative logistics of data tracking and work flow, can success be achieved.

Back to Paper Presentation Abstracts