Statistical Applications for Environmental Analysis and Risk Assessment

Preface

Although the subject of environmental statistics has been around for decades, the average environmental professional is far from comfortable with statistics, in my experience. This should be concerning because the protection of public health and ecological well-being falls largely on these professionals, and statistics should be a prominent part of their arsenal. It is fair to say that the environmental profession revolves around data. The environmental engineer, geologist, or scientist routinely collects data from soil, sediment, water, ambient air or other environmental media, for analysis and interpretation to determine the potential presence and concentrations of environmental contaminants. Based on the results of the data analysis and professional judgment, the practitioner makes a recommendation to the appropriate environmental protection authority that remedial action is necessary to reduce contaminant concentrations and minimize exposure risks, or that the exposure risks are minimal, warranting no further action. More often than not, the environmental regulatory authority concurs.

As it turns out, environmental data samples are rarely well-behaved, with nondetects, outliers, skewness, sustained and/or cyclical trend as habitual offenders in many data samples. Without functional familiarity with at least the basic statistical analysis principles and practices, how can we make sense of data such as these? Even more worrying, risk management decisions are often ...