Explore chapters and articles related to this topic
Data Analysis Methods for Assessing Eutrophication
Published in Michael Karydis, Dimitra Kitsiou, Marine Eutrophication A Global Perspective, 2019
Michael Karydis, Dimitra Kitsiou
The spatial correlogram is a graph with the autocorrelation values plotted in y-axis against distance classes among sampling sites. The Moran’s I or Geary’s c spatial autocorrelation statistics can be calculated to measure spatial autocorrelation (Cliff and Ord, 1973). The variogram is the basic tool of geostatistics and provides the means for the detection of anisotropies and the calculation of the degree of homogeneity in data sets (Mabit and Bernard, 2007). It is expressed as a graph with the semi-variance γ(h) plotted in y-axis against distance classes among sampling sites (h, lag). When anisotropy is present in data, the autocorrelation function is not the same for any geographic direction. The maximum value of γ(h) is called sill and indicates the absence of spatial dependence in data. The semi-variance value γ(h) when h = 0 is the nugget variance and represents the local variation occurring at scales finer than the sampling interval as well as the measurement and sampling error.
Examining the spatial mode in the early market for electric vehicles adoption: evidence from 41 cities in China
Published in Transportation Letters, 2022
Zhengxia He, Yanqing Zhou, Xin Chen, Jianming Wang, Wenxing Shen, Meiling Wang, Wenbo Li
Spatial autocorrelation (local and global) analysis is generally carried out in three steps (Cliff and Ord 1981a): (1) sampling, (2) calculating the spatial autocorrelation coefficient, and (3) autocorrelation significance testing. Global autocorrelation describes the overall spatial distribution of a geographical phenomenon(aggregation, dispersion, or randomness), while the local autocorrelation is calculated between each spatial unit and its adjacent spatial units on a certain attribute(Goodchild 1986).The commonly used measurement approaches of spatial autocorrelation are Moran’s I, Geary’s C, and Getis’s G, among others(; Geary 1954; Getis, 1992). Moran’s I, developed by Patrick Alfred Pierce Moran (Moran 1950), is often used to judge whether there is a correlation between spatial entities within a certain range and was chosen as our research tool, calculated as:
The spatial effect of factor market distortion on green agriculture development in China
Published in Energy Sources, Part A: Recovery, Utilization, and Environmental Effects, 2022
Xinming Wang, Chao Hua, Jianjun Miao
Spatial autocorrelation means that for a certain variable, a region has a positive or negative relationship with its surrounding regions. At present, the testing methods for spatial autocorrelation are relatively mature, such as Moran’s I method and Geary’s C method. The Moran’s I method (Moran 1948) is widely used in the research of spatial geography and regional economics, and is most common in academic circles. The value of Moran’s I is between [−1,1]. A negative Moran’s I value implies a negative spatial autocorrelation of the variable. A positive Moran’s I value implies a positive spatial autocorrelation of the variable. When Moran’s I value is 0, the variable has no spatial correlation.
Comparing four regression techniques to explore factors governing the number of forest fires in Southeast, China
Published in Geomatics, Natural Hazards and Risk, 2021
Qianqian Cao, Lianjun Zhang, Zhangwen Su, Guangyu Wang, Shuaichao Sun, Futao Guo
Model fitting was evaluated using AIC and mean squared errors (MSE) (Burnham and Anderson 2004). Smaller AIC and/or MSE implies better model fitting performance. To evaluate the spatial autocorrelation of the residuals, the Geary’s contiguity ratio (Geary’s C) was calculated, as well as the Z-test for testing H0: C = 1 and associated p-value. The closer to 1 the Geary’s C is, the lower the spatial dependence of the residual is, and hence, the model accounts for more spatial structure problems.