Research project with funding from European Science Foundation (ESF) and Deutsche Forschungsgemeinschaft (DFG) on the conceptualization, storing, analysis and visualization of temporal data in Geographic Information Systems.

The project results were summarized in a book:
Ott, Thomas & Swiaczny, Frank (2001): Time-Integrative Geographic Information Systems. Management and Analysis of Spatio-Temporal Data. Berlin / Heidelberg / New York: Springer. xiv + 232 pages, 65 illustrations, CD-ROM, ISBN 3540410163

Contents

The book deals with the integration of temporal information in Geographic Information Systems. The main purpose of an historical or time-integrative GIS is to reproduce spatio- temporal processes or sequents of events in the real world in the form of a model. The model thus making them accessible for spatial query, analysis and visualization. This volume reflects both theoretical thoughts on the interrelations of space and time, as well as practical examples taken from various fields of application. Table of contents:

  • Introduction
  • Conceptualising real world entities in spatio-temporal GIS
  • Integrating time in Geographic Information Systems
  • Implementation of time in GIS
  • Processing and analysis of temporal data inside a GIS
  • The visualisation of spatio-temporal processes
  • Examples
  • Conclusion
  • Appendix

Reviews

Environment and Planning b vol 30, no. 3, 2003, pp. 465-466
“In keeping with their promise of a solutions-oriented approach to presenting time in GIS, the authors give a wealth of interesting examples.”

Kartografisch Tijdschrift vol. XXVII, no. 4, 2001, S. 59-60
“Mensen die nog nooit met GIS hebben gewerkt, maar dit overwegen te doen, hebben hier echter een prima GIS-leeboek mee in handen. In de breedte worden beginnende gebruikers goed op weg geholpen, terwijl er ruim voldoende verwijzingen (onder meer een literatuurlist van 53 pagina’s) zijn om diepere kennis op te doen.”

From the Introduction

New GIS users are often surprised to discover that the software has great difficulties in answering questions such as, “When did the river flood more than 20% of the agricultural land?”, “Were there ever any parcels of land with industrial zoning within the polluted area?”, “Are there any significant connections between the extension of railway lines and the population development in the 19th century” or responding to requests such as, “Display the forest boundary changes between 1850 and 1950.” These setbacks are due to the fact that commercial GIS packages mostly lack the ability to perform temporal analysis of spatial data. Yet, not only historical research is dependent on the ability to represent and analyse temporal processes in a GIS. Today, the widespread application of GIS in urban, regional and environmental planning, in marketing and logistics, or scientific disciplines like archaeology requires the integration of historical data in order to perform time-change analyses or to deal with temporally variable data. Another crucial motivation for the development of time-integrative GIS techniques is the fact, that enormous amounts of data have been collected in the last decades by census offices, research projects and not the least companies, which can be revaluated and reused.

Main target of this book is, to contribute to the ongoing discussion on temporal GIS and to present practical solutions for the conception and implementation of time-integrative geographic information systems. Main purpose of a temporal or time-integrative GIS is, to reproduce temporal processes or sequences of events of the real world in a model in such a way, to make them accessible for spatial query, analysis and visualization. For that it is necessary to link the spatial information with the corresponding spatial objects (geocoding). In order to be able to represent and analyse both, the temporal and spatial interrelations, the modelling of spatio-temporal data (i.e. the process to derive useful information and “knowledge” from a set of raw data) requires extraordinary assiduousness.

The specific problems of a historical or time-integrative GIS are based on the necessity of simultaneous spatial and temporal reliability. If the individual information layers of the GIS are not to be explored independently, it is of fundamental importance, that on the level of the logical data model, identical points (even as element of a line or polygon) are represented by exactly one pair of coordinates. This holds true for identical points in different thematic layers as well as identical points in different temporal layers or different scales. In regard of the practice of data capturing the latter forms the greatest challenge. Historical information has to be digitised as time sections from diverse maps of differing precision and scale. In opposition to different thematic layers in a specific time section, which usually can be composed from base maps with coherent accuracy, the digitisation of historic sequences is impeded by differing map projections and scales as well as varying levels of precision. Although these problems are to be mastered, both in a fundamental and methodical way as well as in practice with an appropriate expenditure on programming, smaller projects are usually confronted with unsolvable difficulties.

The demand for spatial and temporal reliability leads to the most prominent problem with the construction of a historical GIS: the representation of temporal information in a conceptual data model and the reproduction of temporal data in a logical data model. The structured proceeding described in the book is not to be understood as an end in itself. It rather permits a first overview of the system to be analysed. The more exact the elements and interactions of the system are studied and transformed into a conceptual model against the background of theoretical knowledge, the easier and better it will be possible to form the logical entity-relationship model in the next step. Another advantage of this procedure results from the fact that missing links and relations can already be uncovered, before the actual data acquisition has started. Since the construction of the data objects exerts strong influence on the possibilities of analysis, their feasibility can already be checked at an early stage and modified if necessary. Furthermore the entity-relationship model forms a basis of decision for the selection of data sources, as it permits to identify inappropriate or redundant information and thus may save operating time and costs. However, the significance of these theoretical considerations is qualified by the characteristics of the available sources, since these are not available or accessible in every desired temporal and/or spatial resolution.

Test chapter

Download the first pages of chapter 8 as a PDF.