Bayesian Spatial Uncertainty Analysis

摘要

The term uncertainty originates from imperfect knowledge of the processes under question. These errors are sometimes associated with questionable data quality and scarcity; complexity of the phenomena that are treated by models as simplified systems. Important components of uncertainty analysis include (i) qualitative analysis that identifies the uncertainties, (ii) quantitative analysis of the effects of the uncertainties on the decision process, and (iii) communication of the uncertainty (Funtowwicz and Ravetz 1990, Petersen 2000, Regan et a 1.2002, Katz 2002). In landscape simulations (which are widely used in landscape ecology) a generation of landscape patterns is taking place for the investigation of the local or global connectivity between regions. Neighbourhood structures are used for that purpose to explain and analyse the spatial connectivity between smaller to bigger regions including investigation of the spatial homogeneity. Modelling of that spatiality involves conditional probabilities which are explained by Markov random fields models. Estimation of these particular models could be introduce a Bayesian statistics analysis based on these conditional probabilities which are explained the spatial variability into the regions especially when hidden information's are involved (Zimeras and Matsinos, 2011). In this work a spatial analysis methodology based on Bayesian analysis was introduced and procedures to solve the problem with spatial variability are described based on point estimation techniques.