Article Preview
TopSpatial Uncertainty Analysis
Interactions between biological cells at different scales are characterized by their local dynamics and the emergent spatial patterns are the outcome of different processes. The development of specific new, applied statistical techniques can be explained by the emerging field of specific regions of human body, which focuses on spatial processes operating over various spatial extents biologists are trying to collect quantified information about spatial pattern in order to answer questions regarding the underlying processes (e.g. competition) (Turner, 1989; Wiens, 1989). Although different processes could be responsible in generating the same spatial pattern, its quantification may help to identify these processes. Complementary to answering causal questions about biological processes, quantification of spatial pattern can be used to analyze spatial dependences. However, the sometimes hidden spatial dependence in data can lead to violations of the assumption (Legendre, 1993). It must be noted that there is considerable variety of statistical methods that have been applied in the analysis of spatial variation in biological data. These include dispersal analysis, spectral analysis, wavelet analysis, kriging, spatial Monte Carlo simulations and many geostatistics methods (Zimeras & Matsinos, 2011).
Uncertainty in models can be divided in a similar way by statistical and systematic uncertainty. The statistical uncertainties arise from the variability of input variables and parameters where the variability is known. This variability can be described by probability density functions (PDFs) describing the variability of the input variables and the parameters (Zimeras & Matsinos, 2011). Systematic uncertainties arise from variability in input variables and parameters when variability is unknown. Also unknown processes in the model e.g. incorrect model structure contribute to the systematic uncertainties.