Space can be classified in different ways. The most familiar way is the classification in zero, one, two and tree dimensions representing point, line, area and volume. Animals are often represented as point objects in GIS. For the purpose in consideration here this is being considered adequate, so in the following I will focus on point objects. The term 'animal' used throughout the text can most of the times be considered as an example for a 'moving point object' in general. I assume that the readability and understanding is made easier when using a concrete example, so I will mostly use the term animal instead of 'moving point object'.
Geographical Information Science has a variety of facets. Database issues are fundamental to any GIS. A lot of research in this area is currently being conducted on object-oriented systems, interoperability and the integration of time (e.g., Vckovski, 1997; Yearsley and Worboys, 1995). Temporal GIS research can be grouped into three topics: database, query and analysis. In contrast to the first two topics, the last one has not received much attention yet. Chapter 3.4 will give a short overview of temporal GIS research.
Spatial analysis, on which this research is concentrating, is at the core of GIS applications. Basic functionality can be found in all 'off-the-shelf' systems, but as the diffusion of GIS into other research disciplines increases more analysis functions are needed. Especially in biological research a lot of the acquired data is in the form of point data, e.g. locations of plants or animals. In spite of the relatively large literature on the analysis of point data (not to be confounded with point measurements of continuous phenomena), GIS provide little analysis methods for this type of data. Chapter 3.3 will give an overview over the currently available analysis methods for point objects. In the past few years there has been a lively discussion going on about the integration of statistical analysis of spatial and non-spatial attributes. Several ways of coupling GIS with statistical software have been provided in the literature as a proof of concept, mostly using GIS as a visualization front end to the statistical package, but true integration is still lacking (MacLennan, 1991; Lippert-Stephan, 1996).
In chapter 2 the characteristics of points and time specific to wildlife research are elaborated. After that an overview of the current analysis methods is provided in chapter 3.
In part II concepts and new analytical methods are developed. The main focus is set on the integration of various temporal aspects in the analysis of point objects.
A relatively new research field in GIScience is addressing problems of how to handle uncertainty in GIS with its different aspects from boundary location accuracy to classification uncertainty (Aspinall and Pearson, 1995). It would be tempting to include such aspects in the development of new analysis methods. Unfortunately, as this is still a very difficult aspect to handle for geographic information, it would not be realistic to try including it in a first approach for analysing moving point objects. This will need to be addressed in later studies.
In the following I will examine some issues of the main data type for this study, the point object.