Locational Error Impacts on Local Spatial Autocorrelation Indices: A Syracuse Soil Sample Pb-level Data Case Study

This paper focuses on propagation of location errors in spatial data analysis. Specifically, it investigates how location errors impacts local spatial autocorrelation indices that often are used to identify spatial clusters. Results of a simulation experiment using heavy metal soil sample points in Syracuse, NY, are summarized. In the simulation experiment, artificial location errors were introduced to perturb points, and then local Moran’s I and Getis-Ord local statistics were calculated. The results show that location errors have an impact on the identification of spatial clusters. Some significant spatial clusters with no location error become insignificant ones with location errors, and some insignificant ones with no location error become significant ones with location errors. More severe deviations from the true results are observed with larger location error, as expected.