Paper detail

Derivation of Correlation Dimension from Spatial Autocorrelation Functions

Spatial autocorrelation coefficients such as Moran's index proved to be an eigenvalue of the spatial correlation matrixes. An eigenvalue represents a kind of characteristic length for quantitative analysis. However, if a spatial correlation is based on self-organized evolution, complex structure, and the distributions without characteristic scale, the eigenvalue will be ineffective. In this case, the single Moran index cannot lead to reliable statistic inferences. This paper is devoted to finding advisable approach to measure spatial autocorrelation for the scale-free processes of complex systems by means of mathematical reasoning and empirical analysis. Based on relative step function as spatial contiguity function, a series of ordered spatial autocorrelation coefficients are converted into the corresponding spatial autocorrelation functions. Then the mathematical relation between spatial correlation dimension and spatial autocorrelation functions is derived by decomposition of spatial autocorrelation functions. As results, a set of useful mathematical models are constructed for spatial analysis. Using these models, we can utilize spatial correlation dimension to make simple spatial autocorrelation analysis, and use spatial autocorrelation functions to make complex spatial autocorrelation analysis for geographical phenomena. This study reveals the inherent association of fractal patterns with spatial autocorrelation processes in nature and society. The work may inspire new ideas of spatial modeling and analysis for complex geographical systems.