|Hauhs, M; Trancón y Widemann, B; Archner, O: Theoretical and Practical Challenges of Integrating Ecosystem Data, Lecture Notes in Computer Science, 5333, 64 (2008), doi:10.1007/978-3-540-88875-8_23|
The challenges of data integration in the field of ecological and environmental research are exemplified by the widespread mixture of time series and maps. Typically ecosystem data sets originate from interdisciplinary studies organised around pragmatic, small scale or short-term interests. Integration of such data, e.g. in geographic information systems or time series analysis has often been based on ad hoc solutions. The theoretical basis of the field has remained largely unsolved. Here, we propose a categorical basis for the task of organising and integrating ecosystem data. We focus on time series (data streams) and on maps (spatial configurations of objects) as the two pure limiting cases illustrating typical data evaluation tasks. The core idea of our approach is to use the duality between algebra and coalgebra as the mathematical basis for representing and relating these two data types. This results in two dual modelling paradigms that relate the formal basis to testable models and experiments in different ways.