Spatially Balanced Sampling Jennifer Brown 1 , Blair Robertson 1 , Trent McDonald 2 1 University of Canterbury, New Zealand 2 Western Ecosystems Technology Inc, USA E-mail for correspondence: Jennifer.Brown@canterbury.ac.nz Abstract: Spatially balanced sampling is an emerging area in statistical sam- pling. These designs are popular because they are one way to ensure the selected sample has spatial coverage over the entire survey area. This feature of spatial coverage aids in the resultant sample being representative of the population of interest. One of the first and the most commonly used spatially balanced design is called GRTS (Generalized Random Tessellation Stratified sampling) where sample effort is spread evenly over the target region. The term spread evenly in this context means having coverage of survey effort over the region. The coverage from GRTS has a stochastic component rather than a fixed interval, regularly spaced coverage as in a systematic sampling design. We have extended the idea of GRTS to a new design called Balances Acceptance Sampling (BAS). The BAS design allows surveys to be balanced in dimensions higher then two (n - dimensional space). Until now, most designs have considered balance in 2-D geographic space. With BAS we can achieve balance in 3-D space, or in higher dimensions. In some applications these dimensions can be features other than the spatial measures of geographic location, and the design allows aspects such as time for repeat surveys to be incorporated into sample balance. Keywords: Sample; Environmental Sampling; Spatial Sampling. 1 Introduction Spatial sampling is a broad category referring to designs that incorporate spatial reference in site selection. The purpose of sampling is usually to estimate a population parameter, such as the mean or total of some char- acteristic of interest. For example, in environmental applications surveys may be to estimate the density or abundance of a plant species of inter- est. In social science applications the survey may be to estimate household This paper was published as a part of the proceedings of the 32nd Interna- tional Workshop on Statistical Modelling (IWSM), Johann Bernoulli Institute, Rijksuniversiteit Groningen, Netherlands, 3–7 July 2017. The copyright remains with the author(s). Permission to reproduce or extract any parts of this abstract should be requested from the author(s).