B. Pieke, A. Krüger 261 Flow Maps- Automatic Generation and Visualization in GIS

Abstract. Flow maps are used in order to indicate movements by persons or objects in a given time period. The major problem concerning the production of flow maps is the appropriate placement of movement lines in the map. This paper presents a method of placing lines automatically, based on a modification of an exisiting algorithm. The algorithm has been implemented as an ArcMap extension and was subsequently tested with a data set representing pupil school transitions. 1 BASIC PRINCIPLES Flow maps are of use in many applications: whenever movements by people or objects take place, they can be shown on a flow map. The data visualized in these maps exhibit a high complexity, due to the high number of transition edges present in such a map. Drawing a flow map is difficult, because every flowline has to be placed very carefully depending on the other egdes and nodes. In order to arrange the map as effectively as possible, a special emphasis in its generation has to be put on a clever placement of the edges within the map. A study on the cross-linking of schools (in terms of transition of students) within a defined region served (us) as a use case to analyse the problem of flow maps. The effectiveness of a map is at its highest if the designer and the map user use the same concepts for the organization of the representation, see MacEachren (1995). Following MacEachren (1982), the effectiveness of a map is determined by a variety of aspects, of which map complexity and symbolization are the most important ones. The symbolization specifies how data are represented on a map, whereas the complexity specifies the degree to which the symbols on a map appear intricate (MacEachren 1982). In this use case the important factors are: 1. The complexity, because of its negative impact on effectiveness. Thus it is important to reduce the number of objects depicted/visualized in the map to such an extent that no visual clutters are arising