Review of environmental monitoring methods: survey designs

Abstract

During the past decade and a half, environmental monitoring programs have increased in number and importance. Large scale environmental monitoring programs often present design difficulties because they tend to measure many (sometimes hundreds) of parameters through space and time. This paper reviewed and summarized one important component of environmental monitoring programs, the statistical survey design. Survey designs used for long-term monitoring programs lasting multiple (> or = 3) occasions were reviewed, paying special attention to those published after 1985. During this review, two key components of the overall survey design were identified. The first key component was the membership design. Groups of population units sampled the same occasion were called panels here, and the membership design specified which units were members of which panels. The second component was the revisit design that specified when panels were to be revisited. Membership designs varied, but some form of simple random or systematic design was popular. Among revisit designs, four basic patterns were found in the literature and their relative strengths and weaknesses were summarized. To efficiently discuss revisit designs, a new unified short-hand notation was proposed and adopted.