ABSTRACT
Chapter 4 described geostatistical modeling of scale based on variograms, which assume continuous indexing of locations over space. As mentioned toward the end of Chapter 4, unlike geostatistical data, lattice data result from aggregation of field variables over finite areal units, between which there is no space where additional sample points could be obtained. This chapter will discuss lattice data and their scale models, based on measures of spatial autocorrelation, for which neighborhood relationships play a central role in data analysis and scale.
