Spatial sampling schemes for remote sensing

It is discussed in detail how an efficient and effective sampling scheme can be designed in view of the aims and constraints of the project at large.

Abstract

The objective of this chapter is to communicate to researchers some basic knowledge of sampling for applications in remote sensing projects. As a consequence, all theory, formulae and terminology not essential for application in remote sensing or for basic understanding, is omitted. Reference is made to the sampling literature where possible. However, in one way this chapter is broader than usual texts on sampling, because full attention is paid to how sampling theory can be employed and embedded in real-life research projects. Thus it is discussed in detail how an efficient and effective sampling scheme can be designed in view of the aims and constraints of the project at large. The chapter also discusses the differences and the choice between the designbased and the model-based approach to sampling, as much confusion around this issue still exists in the applied literature. More text is devoted to design-based strategies than to model-based strategies, not because they are more important but because there are more of them, and reference to existing literature is often problematic. In the general sampling literature design-based strategies are mostly presented in a non-spatial finite population framework, and ’translation’ into the spatial context is in order.