Purpose

The aim of this chapter is to take the first of three steps towards a more natural and accurate handling of animal data in spatial analysis. It is a conceptual shift from statics to dynamics and will set the basis for the following chapters. It is a relatively technical but easy to read chapter which will introduce powerful instruments for the visualization and fast analysis of complex spatio-temporal patterns.

At a first glance temporal aspects might seem to be relatively simple to handle. But going a bit more into detail they become quite complex. In section 2.2 some of these problems were mentioned. Calculations of the beginning and end of dawn, for example, depend on the season, latitude and longitude and require a lot of geographical, astronomical and temporal computations. In the case of the tides calculations become even more complex. The complexity increases even more when combinations of such aspects are considered.

Shepherd (1995) provides a recent classification scheme for dynamic visualizations of geographical data. He distinguishes the following 6 classes according to the source of variation (table 4.2). I will introduce the Temporal Data Frames (TDF) concept as an extension to the first category of converting real time to display time. It is needed as a basic concept in the next chapter. It will provide fast methods for accessing large volumes of data.

Table 4.2: Classification scheme for dynamic visualizations from Shepherd (1995)

Class	Example/Explanation
Data	converting 'real time' to 'display time'
Representation	changing symbolism
Observer	moving observer
Agents	e.g. particles flowing
Entities	e.g. objects trigger symbol change
Designers	e.g. esthetic symbols

In this chapter I shall try to elaborate a methodological framework for the following objectives:

• Refinement of the temporal perception of biological data. As I stated in the introduction temporal aspects of spatial phenomena have not received enough attention in the past. Time is handled in a very coarse way which probably hides a lot of information inherent to the data. Finer methods should be developed to analyze the data to their full extent.
• Integration of different time aspects. Normally researchers in wildlife biology only include seasons as a temporal dimension in their analysis. As I have shown in section 2.2 there are several temporal aspects that are interlinked in complex manners. It is the aim to make these aspects accessible to the research community and provide the necessary methods for it.
• Recognition of changes. In the area of fast changes in the environment it is important to be able to detect such changes in a very efficient way. The following section shall develop mechanisms for the detection of changes in relation to different temporal aspects at a very fine level.
• Recognition of biologically reasonable phases. Preventing the application of 'unnatural' phase definitions in the analysis of biological data, methods should be developed to enable the researcher to perceive and define biologically reasonable phases in the data. It should provide the means to move away from the human-oriented definitions based on cultural calendars.