1999-08
Introduction
This report outlines the efforts of the SHPO offices of New Mexico and Wyoming to implement a common cultural resource database design. This collaboration is developing a common spatial data structure and implemented a prototype Geographic Information System based on the ESRI Spatial Database Engine and other supporting ESRI Technologies.
This report documents NCPTT-funded efforts aimed at the advancing State Historic Preservation Office (SHPO) Geographic Information Systems (GIS) in the western United States. Many largescale cultural resource inventories are operated by agencies other than SHPOs for example, Tribal HPOs, historical societies, universities, and museums, so our focus has been more on cultural resource information systems (CRIS) and particularly on GIS components of these systems — rather than on SHPO-specific data needs.
This project was prompted by the common experiences of several SHPOs in implementing GIS technology in large, transaction-heavy CRIS environments, combined with their desire to pool resources. Because many GIS implementation problems stem from the magnitude of both the archeological record and the demand for information common to these two states, project findings will be most relevant to other similarly situated Western states. Our solutions are undoubtedly scaleable to less “intensive” CRISs, but we suspect the Cost effectiveness of benefits may be
diminished.
Although the applicants for this grant were the New Mexico and Wyoming SHPOs, many other entities were involved. Most of the initial data-modeling tasks were conducted as part of a U.S. Geological Survey, Federal Geographic Data Committee (FGDC), grant to develop metadata and data content standards specific to cultural resources in the western United States (Appendix 1). Representatives from most western states and federal land managing agencies participated in one or both of the FGDC-sponsored workshops. Their contribution to this project has been significant. In addition, the University of New Mexico, Earth Data Analysis Center (EDAC), conducted comprehensive training on the FGDC metadata standards and documentation tools during the first meeting.
Project personnel consisted of the following individuals:
Report Organization
This is a technology implementation project. Presentation of all of the many technical details relating to our effort is exceedingly difficult in a report format. We have decided to present much of the technical information as appendices to this report, but many more details exist in application
code, database dictionaries, manuals, and the like. The main products of this grant are computer applications These too are difficult to appreciate from a report. All project materials and applications are, of course, available for review and demonstration, and we stand ready to assist
any agency that may wish to benefit from our experience.
The report first describes the problems we set out to solve in this project, our major objectives, and our strategy for reaching these objectives. Second, we provide a task-by-task summary of our efforts to date. Finally, we attempt to assess the cost effectiveness of what we have done and to
make recommendations on technology transfer.
GIS and SHPO Information Systems: “The Awful Truth”
Location is central to the management of cultural resources. If the location of a building, district, site, or object is unknown, then no action can be taken to manage, to preserve, to reconstruct, or to protect it. In spite of the central importance of location, spatial information technologies such as GIS are rarely integrated into everyday SHPO decision-making. As revealed in recent surveys of historic preservation archives in the US (Wood 1990, Ebert et al 1994, COAHP 1994), database management technologies are well established, but the transition to GIS technology has been very
slow in coming in spite of a very high user demand for geospatial data on cultural resources. Why?
Three reasons for slow adoption of GIS based on Wyoming and New Mexico SHPO’s experiences
are:
- GIS is not optimized for transaction-based computing environments typical among Western SHPOs. GIS is optimized for data analysis not data management. Current GIS technology provides efficient and robust tools for manipulating and storing spatial data, but the tools provided for data management are designed mainly to build and manage large, static, analytical datasets. Most SHPO information systems are transaction-based. That is, these systems are built and continuously updated through database transactions, consisting of property and investigation records generated by management and research activities. In western states, the number of annual database transactions generated by these activities ranges in the
tens of thousands. Compared to a modern RDBMS, GIS transaction processing controls–those functions that maintain the security and integrity of a multi-user database during simultaneous data entry, validation, and query events — are primitive. - Spatial relationships are difficult to create and maintain at the statewide level. Most GIS data models are organized in terms of space rather than the mapped features. This requires that feature topology be established at the time of data capture and then pre-stored in a proprietary file format (a coverage). Topology refers to the spatial relationships among connecting, adjacent, or overlapping elements (i.e., nodes, arcs, polygons, etc.) used, in this case, to rep resent cultural features (i.e., sites, buildings, districts, trails, etc) GIS is designed to efficiently store, manage, and manipulate these spatial relationships but, in GIS products like Arclnfo, the topological model necessitates special processing and storage in a proprietary file format (a coverage). Even more problematic, the topological model requires large databases to be partitioned into smaller spatial units (tiles) for processing and storage efficiency. Coverage tiles must be rebuilt every time new features are added or existing features are edited. Features crossing partition boundaries must also be edge-matched prior to storage. The creation and maintenance of topology requires considerable skill and training and cannot be performed effectively by untrained staff. Topology is responsible for the power and efficiency of GIS in manipulating spatial data, but it also introduces complexity and processing overhead to the basic data collection and transaction-processing functions that dominate SHPO information systems.
- Broad distribution of spatial data through GIS is difficult and expensive. Access to applications has been a major obstacle to widespread use of GIS technology. Paper maps are commonly used to distribute the results of GIS-based analysis widely, but real-time use of geographic data has been precluded by the requirement to have access to expensive and complex hardware and software. Even when GIS or desktop mapping applications are obtained, the use of spatial data beyond the local work group usually requires replication of the database or access to very high speed wide area network connections. If data replication is performed then highly structured procedures must be developed and followed religiously in order to safely manage multiple database copies.
During NMSHPO’s pilot GIS project, the ESRI Arclnfo environment was found to be very cumbersome. Data capture proceeded in batch mode, one map sheet at a time, while new in formation was placed in a backlog until the appropriate map sheet was processed. After multiple layers for each map sheet were digitized or scanned, extreme care had to be taken to associate database records with the appropriate manuscripted features, requiring considerable collaboration between the archeologist who created the manuscript and the GIS specialist. The entire data collection process was so procedurally complex and labor-intensive that we began questioning the cost-effectiveness of GIS. We found that GIS tended to intensify, rather than resolve, problems related to the three fundamental information system objectives: data capture, data management and data delivery.
We proposed a technical solution to NCPTT based on a promising new GIS technology: The Environmental Systems Research Institute (ESRI) Spatial Database Engine (SDE). The remainder of this report describes our efforts to implement and evaluate this technology. The objectives of the project are twofold:
- Develop a common logical spatial model for cultural resources among New Mexico, Wyoming, and other interested Western states.
- Develop a spatial database prototype, using SDE, in New Mexico based on that model.
In out proposal, several tasks and deliverable products were defined under each of these objectives. These are discussed below.
