The VARIOGRAM Procedure

Autocorrelation Statistics

Spatial autocorrelation measures offer you additional insight into the interdependence of spatial data. These measures quantify the correlation of an SRF $Z(\bm {s})$ with itself at different locations, and they can be very useful whether you have information at exact locations (point-referenced data) or measurements that characterize an area type such as counties, census tracts, zip codes, and so on (areal data).

As in the semivariogram computation, a key issue for the autocorrelation statistics is that you work with a set $z_ i$ of measurements, $i=1,\dots ,n$, that are free of nonrandom surface trends and have a constant mean.