Using GIS to map rail network history

Journal of Transport History, The, Mar 2004 by Siebert, Loren

Research on the history of transport systems often involves the creation of maps to represent the conditions of a transport network at a given point in time, or perhaps a series of maps showing the development of the transport network. Traditionally, such maps have been prepared by manual cartographic techniques. However, the development of geographic information systems (GIS) over the past decades has now made it possible to generate such maps by computer. Besides offering a more efficient method of preparing maps, geographic information systems can be used as a repository of a variety of data about transport history and as a powerful tool for analysing and visualising historical patterns and geographic temporal relationships.

In principle, use of GIS to produce a history of a transport system would appear to be quite straightforward. Fundamentally most transport systems consist of linear links between places, so it should be possible to use basic GIS features such as lines and points to represent the transport system. However, there are inherent complexities in transport networks and their histories that make development of such a GIS database more difficult. In this article, I will introduce my project to develop a GIS-based history of the rail network in the Kanto region of Japan, around Tokyo, as an example of how GIS can be used to record the history of a complex transport network.

The first railroad in Japan opened in 1872, four years after the Meiji restoration of 1868. It connected Tokyo (at Shiodome, near modern Shinbashi) with the new port of Yokohama to the south, along Tokyo Bay. The Tokyo area and the surrounding Kanto region now have one of the world's densest rail networks, serving Tokyo metropolitan prefecture, the three adjacent prefectures of Kanagawa, Saitama, and Chiba, as well as parts of Gunma, Tochigi, Ibaraki, Yamanashi, and Shizuoka.

To record the history of this vast rail network, I obtained detailed source materials, designed the necessary geographic and database structures in the GIS, and recorded numerous aspects of the system's historical development. My goal was to be comprehensive in space, theme, and time in order to identify as many different types of historical conditions as possible and ensure that the GIS would be able to handle them.

In this article, I describe the technical issues involved so that other researchers who want to document the rail history of another region in a GIS will be able to assess what they need and how they can use a GIS to do it. Most of the examples I give are meant to illustrate the nature of the data, the way they are recorded in the GIS, and how they can be mapped, rather than specifically to describe or analyse the history of the Greater Tokyo rail system. Separate articles dealing with historical mapping and interpretation of the Kanto region's rail network with this GIS historical spatial database have been published or are in preparation.1

This GIS-based rail history is part of a larger project2 to produce a GIS spatial history of Tokyo and its surroundings during the modern era. The project currently includes shoreline changes in Tokyo Bay; changes in rivers and canals; administrative boundary changes and consolidations from village, to town, to city, to city ward; and population changes. It also includes continuing work on characterising land cover changes.

Tokyo has one of the world's densest rail networks, and thus is both an ideal location for producing a GIS rail history and a very difficult one, owing to the complexity of the system and the vast amounts of information available. By choosing this extensive system, I have been able to discover historical situations requiring development of special recording methods in the GIS. It would now be quite easy to apply and adapt the methods I have developed for mapping the history of another urban region's rail network. I will soon begin expanding my project to include all of Japan.

Besides its rail network, Tokyo also had an extensive streetcar system, as did Yokohama and some other cities in the Kanto region. In general, streetcar networks are even more complex than rail networks, because individual streetcars can be routed in many ways from point to point, and streetcar stations can be more easily and more often relocated. Also, obtaining detailed maps showing the many changes in the streetcar network is a more challenging task than finding sources for the rail network. Therefore, I decided to treat input of the Kanto area's streetcar systems as a completely separate project for the future. Adding such information would be essential, however, for portraying the full picture of transport possibilities in Tokyo, because streetcars served as the initial links between the peripheral commuter line stations before commuter lines and subways were extended into and through the urban core.

Information sources for rail network history

A significant amount of information exists on the development of railroads in Japan. Most rail companies have sought to provide information to the public about the history of their company, lines, and stations. In addition, various publishers have responded to, and helped foster, a high level of interest in railroads, trains, and related historical facts. This is not at all surprising in a country where railroads play such a vital role in the lives of most people. In many senses, therefore, this project has been data-rich.