Geographic information systems (GIS) were developed in the 1960s and 1970s, primarily in the public and academic sectors. In the 1980s, the GIS industry developed under governmental leadership and grew by about 20% each year. The Federal Geographic Data Committee in the U. S. was organized in 1990 under Office of Management and Budget Circular A-16 as the interagency coordinating body for the promotion of sharing, use, and dissemination of the National Spatial Data Infrastructure. The maturing of the GIS industry supported by Federal and state governments and the development of “geographic information science” surrounding the technology have had a major influence on our society and the methodological framework for geographical research.
This GIS revolution of the early 1990s allowed us to develop new ways of understanding the real world and new technologies such as virtual reality using GIS. GIS has also had a major influence on the discipline of geography. The purpose of this paper is to describe some issues concerning the nature of the GIS influence on geographical methodology.
The first issue is the concept of a scaleless GIS database. This means that the database, unlike a paper map, has no apparent scale and is limited by the unique precision of the digitized original data. The second issue concerns the geospatial data model of the real world. Geographic information is extremely complex and not captured well by the commonly used vector and raster data models in GIS. An object-oriented and feature-based approach provides better strategies for modeling the real world as close to a user's perspective as possible. A fundamental unit of the Geographic Information Standard in ISO/TC211 is called a feature. A Geographic Information Standard provides a standard framework for the classification of features in the real world. Figure 1 shows the modeling process from the universe of discourse to the geographic dataset using UML. A feature catalogue means a dictionary of feature types. This methodology of geospatial modeling is based on the concepts of the general feature model (GFM). The GFM is a model based on an object-oriented approach and is similar to the chorographic model of traditional geography. Feature-based GIS provides a new methodology for geography under the traditional chorographic conceptualization of the real world.

View full abstract

As studies of geographic information systems (GIS) have shifted the emphasis from “system” to “science” since the 1990s, a close relationship between GIS and spatial cognition research has emerged. The aim of this paper is to review the findings and issues of the previous studies on spatial cognition in terms of GIS, by classifying the research topics as follows: 1) GIS as a tool for analyzing and representing spatial cognition; 2) GIS as a model of spatial cognition; 3) GIS as a source of spatial knowledge; and (4) improvement of GIS by applying the findings of spatial cognition research.
Although there have been few studies of spatial cognition employing GIS as a tool, researchers have recently directed attention to the use of virtual environments constructed by connecting virtual reality (VR) with GIS for experiments on spatial cognition. VR technologies can be also applicable to the development of GIS user interfaces.
As Golledge (1992) has aptly noted that a cognitive map is an “internalized GIS, ” GIS should become a conceptual model of cognitive mapping and elaborate the model of cognitive processing of geographic information. In particular, ideas of “spatial primitives” and “naive geography” will not only be useful for understanding cognitive mapping by comparison with GIS but also provide many suggestions for designing GIS.
On the other hand, the role of GIS as a source of spatial knowledge will become important when GItechnologies spread among people. Specifically, advanced techniques for communicating geographic information through three-dimensional views, animation, sound, and VR enabled by GIS will lead to reconsideration of the dichotomy between direct (or primary) and indirect (or secondary) sources of spatial information. Such a variation in the user interface of GIS will provide a flexible method for presenting geographic information according to the spatial abilities of users.
The results obtained by previous research on spatial cognition can be also applicable to the development of GIS. Since the output of GIS is usually represented by maps, studies on cartographic communication and geographic visualization should provide numerous suggestions for the improvement of its user interface. Language also plays an important role in user interfaces of GIS. In particular, it becomes important to study verbal descriptions of space as well as tactual maps in considering the communication of geographic information for the visually impaired.
The above-mentioned issues in spatial cognition research related to GIS must be tackled in collaboration with researchers in psychology and information science. In this interdisciplinary field of research, the important role of geographers probably lies in examination of the findings obtained by other disciplines in terms of ecological validity, paying attention to the context of using geographic information.

View full abstract

This paper analyzes the stability of accessibility to ophthalmic hospitals in Matsudo City, Chiba Prefecture, by changing measurement methods and aspects using geographic information systems. Especially the number of spatial choice and area unit are changed, and consulting hour is considered to analyze the stability of accessibility.
First the accessibility is measured as the shortest road distance from the center of a neighborhood districts (Chouchou-aza) to the nearest hospital on the road network. The accessibility to ophthalmic hospitals in Matsudo is classified into four levels: good, normal, bad, and very bad. This accessibility is compared with the accessibility classified into nine types of four levels in terms of the shortest road distances from Chouchou-aza not only to the nearest but also the second nearest hospitals. The reason why the second nearest is considered is that a second choice is important in the spatial choice of an urban facility. The results of comparison are as follows.
(1) Most of Chouchou-aza evaluated as having a high level of accessibility to the nearest facility are lowered by considering the second nearest one. This means weakness of their stabilities. On the other hand, the Chouchou-aza at the bad and very bad levels are stable at the same levels even if considering the second nearest. It is noteworthy in changing the method of measurement that about 80% of Chouchou-aza evaluated as good are weak in stability.
(2) In Chouchou-aza with good accessibility where the agglomeration of facilities is low, stability is weak. Therefore the second nearest facility should be considered in the measurement of accessibility.
Next the accessibility is classified into nine types in terms of the shortest road distances from the center of a 100-meter mesh to the nearest and second nearest hospitals and is compared with the accessibility at Chouchou-aza level.
(3) In 20% of Chouchou-aza, the accessibility differs in half the area in comparison with the 100-meter mesh level.
(4) A modifiable unit area problem in the measurement of accessibility will occur in the city center where the evaluation of accessibility drops when changing the unit area of measurement from Chouchou-aza to 100-meter mesh.
Finally, accessibility is measured by considering consulting days and hours to understand the temporal dimension of accessibility.
(5) All hospitals are not open at the same time, but the accessibility is nearly same as at 10 a.m. on Monday.
(6) Saturday is the lowest provision level in hospital service except for Sundays and holidays. The number of hospitals opened is the highest at 3 p.m. on Saturday and accessibility around the largest center of Matsudo is highly stable, but the second and third largest centers are unstable with accessibility falling drastically.
This paper analyzes the stability of accessibility by changing measurement methods and considering time the dimension. It was clarified in this paper that the stability of accessibility is dependent on the location pattern and density of urban facilities.

View full abstract

This paper proposes a method of school redistricting using geographic information system, with a focus on evaluation of the school family system. Considering the physical environment of education, the method treats school redistricting as a spatial optimization problem, or to be more exact, a spatial allocation problem. The physical environment is measured by the total distance to school, the number of students who cross arterial roads on their way to school, the number of students who belong to different schools after redistricting, and the area of regions that belong to different school districts after redistricting: These measures are used as objective functions in optimization. Constraints include the number of schools, the maximum distance between student and school, and the maximum number of students who cross arterial roads on their way to school. The results of optimization are evaluated in terms of the average and maximum distances to school, the number of students who cross arterial roads on their way to school, the number of students who belong to different schools. after redistricting, etc.
Under the above conditions, the school district system is optimized in two cases. One is the case where the school family system is not introduced, that is, the school district is determined separately for elementary and junior high schools. The other considers the school family system explicitly, where any elementary school district belongs to exactly one junior high school district. The method is applied to school redistricting in Kita-ku, Tokyo.
The average and maximum distances between student and school slightly increased with the introduction of the system. This problem, however, can almost be resolved by adding the maximum distance between student and school as a constraint. Consequently, it is reasonable to say that the school family system is a possible option in terms of the physical environment of education.

View full abstract