A quantitative measure of the linear relationship among values in a dataset and the values (from the same dataset) at nearby locations.
Spatial autocorrelation statistics measure and analyze the degree of [linear] dependency among observations in a geographic space.
"Autocorrelation" is often used more loosely to refer simply to the presence of significantly nonzero autocorrelation at one or more distances. The term is also sometimes used, albeit incorrectly, to refer to any measure of similarity among values and neighboring values. For example,
Temporal datais said to exhibit
serial autocorrelationwhen values measured close together in time are more similar than values measured far apart in time.
Spatial datais said to exhibit
spatial autocorrelationwhen values measured nearby in space are more similar than values measured farther away from each other.