ABSTRACT
The complex spatial nature of the landscape presents a number of problems when designing optimal sampling schemes for surveys of land use. First, there is a lack of information on the spatial characteristics of the landscape. Second, little attention has been given to extensions of traditional sampling theory for two-dimensional sampling of land use. Here we present a methodology which uses a digital representation of the landscape derived from classified Landsat thematic mapper (TM) imagery to provide spatial data on land-use characteristics, which in turn provides the basis for a series of simulation experiments which tested the efficiency of different sampling approaches.
