ABSTRACT
Georeferenced data usually require considerable manipulation and checking before they can be subjected to a statistical analysis. Point sample data, whether they are manually sampled (e.g., soil core data) or automatically sampled (e.g., LiDAR data), must often be converted from a spreadsheet or database format into a geographic information system (GIS) data –le format such as the ESRI shape–le. Automatically sampled data often contain numerous outliers that must be detected and dealt with. Image data must often be georegistered to the earth’s surface. Moreover, spatial data are often misaligned, that is, they are not recorded at the same location and, in a sense that will be made clearer later in this chapter, they are often measured at different scales.
