Statistical Methods for Environmental Pollution Monitoring.

Abstract

Environmental pollution studies may be divided into the following broad and somewhat overlapping types.1. Monitoring.Data may be collected (a) to monitor or to characterize ambient concentrations in environmental media (air, water, soil, biota) or (b) to monitor concentrations in air and water effluents.The purpose may be to assess the adequacy of controls on the release or containment of pollutants, Overview of the Design and Analysis Process 3give misleading results when estimating the variance of estimated means, computing confidence limits on means, or deternlining the number of measurements needed to estimate a mean.Other problems that plague environmental data sets are large measurement errors (both random and systematic, discussed in Chapter 2), data near or below measurement detection limits (Chapter 14), missing and/or suspect data values (Chapter 15), complex trends and patterns in mean concentration levels over time and/or space, complicated cause-and-effect relationships, and the frequent need to measure more than one variable at a time.Berthouex, Hunter, and Pallesen (1981) review these types of problems in the context of wastewater treatment plants.They stress the need for graphical methods to display data, for considering the effect of serial correlation on frequency of sampling, and for conducting designed experiments to study cause-and-effect relationships.Schweitzer and Black (1985) discuss several statistical methods that may be useful for pollution data.Many routine monitoring programs generate very large data bases.In this situation it is important to develop efficient computer storage, retrieval, and data analysis and graphical software systems so that the data are fully utilized and interpreted.This point is emphasized by Langford (1978).The development of interactive graphics terminals, minicomputers, and personal computers greatly increases the potential for the investigator to view, plot, and statistically analyze data.In contrast to monitoring programs, some environmental pollution research studies may generate data sets that contain insufficient information to achieve study objectives.Here the challenge is to look carefully at study objectives, the resources available to collect data, and the anticipated variability in the data so that a cost-effective study design can be developed.Whether the study is large or small, it is important to specify the accuracy and precision required of estimated quantities, and the probabilities that can be tolerated of making wrong decisions when using a statistical test.These specifications in conjunction with information on variability can be used to help determine the amount of data needed.These aspects are discussed in Chapters 4-10 and 13. OVERVIEW OF THE DESIGN AND ANALYSIS PROCESSWhen planning an environmental sampling study, one must plan the major tasks required to conduct a successful study.The following steps give an overview of the process.Schweitzer (1982) gives additional discussion relative to monitoring uncontrolled hazardous waste sites.5. Develop a qual ity assurance program pertaining to all aspects of the study, including sample collection, handling, laboratory analysis, data coding and manipulation, statistical analysis, and presenting and reporting results.6. Examine data from prior studies or conduct pilot or base-line studies to approximate the variability, trends, cycles, and correlations likely to be present in the data.7. Develop field sampling designs and sample measurement procedures that will yield representative data from the defined population.S. Determine required statistical data plots, summaries, and statistical analyses, and obtain necessary computer software and personnel for these needs.9. Conduct the study according to a written protocol that will implement the sampling and quality assurance plans.10.Summarize, plot, and statistically analyze the data to extract relevant information and to evaluate hypotheses.11.Assess the uncertainty in estimated quantities such as means, trends, and average maximums.12. Evaluate whether study objectives have been met, and use the newly acquired information to develop more cost-effective studies in the future.1.4