Does environmental data collection need statistics?

Abstract

The term 'statistics' with reference to environmental science and policymaking might mean different things: the development of statistical methodology, the methodology developed by statisticians to interpret and analyse such data, or the statistical data that are needed to understand environmental pollution and to identify possible policy options. This may and indeed has led to confusion on the part of the users of 'statistics' in environmental science and policymaking. This paper focuses on some of the needs of environmental scientists for statistical methodologies to help us do our jobs better. Statisticians and statistics can contribute to environmental science and the environmental policymaking process in three main ways. First, in describing phenomena: this may in some cases be a non-trivial application of statistical methods. Second, by assessing causal relations: (multivariate) analysis techniques can be applied in establishing the (causal) relations between pressures, state and impact. Third, by setting and checking of standards: uncertainties between the dose-effect relations and the formulation of the standard, and the standard and the state of the environment should be carefully dealt with. Statisticians may help in formulating the standards in such a way that standard checking is as straightforward as possible.