Spatial sampling is very used in environmental survey but can also be
interesting for business or population surveys. We refer to spatial surveys
when a distance is defined between the statistical units. This distance can be
based on geographical coordinates but can also be defined on auxiliary
variables available from a business or population register. In general, two
neighboring units are similar. It is then interesting to spread the sample in
the space of the units in order to make the design more efficient. New spatial
methods with equal and unequal inclusion probabilities will be presented: the
GRTS, the local pivotal method and the local cube method will be covered. Some
examples will be presented using the R package Balanced Sampling developed by
Anton Grafström.