from the published application-specific datasets. This separation serves to reduce redundant updates that tend to cause inconsistent representations. Al provides detailed instructions for how to structure and process data so that it can be used to support multiple applications without having to separately and duplicatively maintain the data.
Al Butler shows how to address the many problems unique to transportation data and related business processes. Transportation geodatabases have been difficult to construct in the past, due in large part to a lack of basic guidance. This book fills that void, explaining how to construct transportation geodatabases in a manner that fulfills our vision for flexibility to meet multiple needs. I am impressed by both its breadth and depth. Designing Geodatabases for Transportation will prove to be an important and powerful template in the series.
I greatly appreciate the vision of ESRI, particularly Jack Dangermond, for supporting development of this transportation template addition to the geodatabase design series. I have known Jack for many years and continue to be impressed by his deep commitment to the development of GIS.
Kenneth J. Dueker, AICP
Emeritus professor of urban studies and planning
Portland State University
Portland, Oregon
Preface
Designing Geodatabases for Transportation is the first book published on transportation spatial database design, a fact you may find quite surprising. To be sure, much work has been done to develop transportation data models, but the result of these efforts has remained fairly theoretical and conceptual. Geographic information systems (GIS) have included spatial databases for transportation since the vector data structure was first used. Indeed, one of the oldest and most famous vector data structures is the TIGER/Line file, which uses transportation features as a principal organizing element. As with many transportation datasets that are used as a reference layer, the TIGER/Line file was designed to convey something else: census data. But it has become the de facto source of transportation data for most local governments and several commercial vendors. Yet, as old and common as transportation datasets are, there has been virtually no organized presentation of how they might be broadly constructed. This book is intended to provide such a presentation by offering a set of geodatabase design problems and alternative solutions that cover the range of travel modes and GIS users.
Although the context is transportation and its many modes of travel, this book is about data maintenance, emphasizing a “measure once, cut twice” design philosophy that offers guidance for all types of spatial databases. Designing Geodatabases for Transportation is a useful guide for all data themes, not just transportation. The book’s editing geodatabase design approach, which supports record-level metadata and the ability to recover the state of the database at any point in time, has widespread applicability. Part 1 also includes fundamental guidance on the principles of agile design, database normalization, key geodatabase classes, and ArcGIS functionality. It takes you “under the hood” so you can understand better why the ideas offered in parts 2 and 3 can work for you.
The foundation established in part 1 shifts to a purely transportation orientation in part 2, concentrating on the fundamental needs of the GIS for transportation (GIS-T) field. The original industry data models developed in NCHRP (National Cooperative Highway Research Program) 20-27, UNETRANS (Unified Network for Transportation), and other efforts are presented and critiqued. Part 2 illustrates a variety of linear facility segmentation schemes, with design approaches offered for each. Most importantly, part 2 shows how to integrate multiple segmentation schemes in a single geodatabase.
In part 3, useful ideas from the historical references and various design approaches of part 2 are incorporated into a revised UNETRANS enterprise GIS-T data model and in a number of design approaches directed at specific functional requirements. The revised UNETRANS data model drops the original’s geometric network foundation for the new street network capabilities of ArcGIS. There is no need for the user to adopt the entire data model, which is offered only as a general guide for how a large organization might assemble all the pieces.
The multimodal UNETRANS model leads to mode-specific chapters. Two chapters provide guidance for building a highway inventory geodatabase at a state DOT. Chapter 14 covers the editing environment, showing how one model can be implemented in different ways depending on the agency’s business rules. Chapter 15 deals with the publishing process, showing how to use ArcGIS functions or to structure the SQL statements that extract data from the normalized editing geodatabase and publish it in denormalized tables and fully attributed feature classes.
The transit chapter expands on the earlier discussion of transportation networks and the capabilities of a street network foundation for the new UNETRANS model. The chapter shows real-world examples for readers interested in the transit mode. The closing chapters on navigable waterways and railroads are intended to serve as primers for the generalist to understand transportation features and their attributes in these modes of travel.
Throughout the book, design concepts presented in a problem-solution approach enable you to quickly identify the parts you need. Each discussion begins with the simplest approach and grows increasingly complex by adding functional requirements to the initial problem. In general, the simpler solutions are most appropriate for smaller agencies with limited transportation feature attributes, while the more complex solutions will likely be applicable at larger organizations that must meet a variety of application requirements.
For all the recent technological advances in the GIS field, it is really just getting started. In a significant evolutionary step, GIS now is adopting concepts and practices more commonly found in mainstream information technology (IT). Such adoptions are consistent with the growing importance of GIS to the organization. As the technology moves from the workgroup to the enterprise, it presents both greater benefits and higher risks. IT practices are designed to limit risk, not only from hardware and software issues, but also from data defects.
Many of the ideas presented in Designing Geodatabases for Transportation are derived from common IT practices. Concepts like normalization and record-level metadata may be fairly unknown in geodatabase design, but they are the bread and butter of traditional transactional database design. So, while this book may be targeted to the transportation theme, it serves as a general guide for designing geodatabases of all themes.
The GIS-T field has been held back by two common issues. First, the field has sought to develop an industry data model when such a model really has no application. This is because almost all organizations that need a transportation spatial database already have one, which is the second issue. Exacerbating these two issues is the general rule that the larger the organization, the more likely it has developed homegrown solutions to meet its information needs and these solutions tend to be poorly documented and difficult to change. As a result, it is almost impossible for an organization to abandon its long-held data structures with a wide variety of homegrown applications constructed to use them. The larger the organization, the more difficult it is to change. Designing Geodatabases for Transportation shows how to solve the many problems existing today as a result of haphazard database design over many years. Technical problems
