ABSTRACT
Spatial data analysis has been a salient feature of quantitative geography and a foundation for GIS. However, the temporal aspect of spatial data should not be ignored when studying geographic objects or events (from here onward simply events). This is because spatial analysis of a set of geographic events often fails to capture the underlying process when time is not considered. In order to understand how the events evolve in space and over time, spatiotemporal (ST) analysis is often needed in order to identify and measure the level of spatiotemporal clustering or dispersion so that the ST processes can be characterized as a random process. Similar to the research paradigm of spatial analysis, measuring the level of ST clustering in a set of ST points (that represent geographic events) is generally the very first step in ST analysis. If the ST points are found to be a non-random process with statistical significance, the analysis then proceeds to trying to find if there are any underlying socio-economic or environmental factors that help form such a non-random process.
