Advanced methods to design samples for land use/land cover surveys

Abstract

The particular characteristics of geographically distributed data should be taken into account in designing land use/land cover survey. The traditional sampling designs might not address the specificity of this survey. In fact, in the presence of spatial homogeneity of the phenomenon to be sampled, it is desirable to make use of this information in the sampling design. This paper discusses several methods for sampling spatial units that have been recently introduced in literature. The main assumption is to consider the geographical space as a finite population. The methodological framework is of design-based typology. The techniques outlined are: the GRTS, the cube, the SPCS, the LPMs, and the PPDs. These methods will be verified on data deriving from LUCAS 2012.