The Philosophy of Monitoring [and Discussion]

The paper is concerned with the efficient design of monitoring studies on a logical basis to meet stated objectives. A series of questions are posed: why monitor, what to monitor, where to monitor, when to monitor and how to monitor? Three broad categories of monitoring can be discerned: observation, explanation and control. The first tends to be orientated towards the receptor of pollution and the last is generally the most source-orientated. The paper discusses the requirements of different types of study. It emphasizes the desirability of monitoring the specific polluting agent having a particular effect but recognizes that sometimes a surrogate may have to be accepted. When one should measure depends on the nature of the effect produced by the pollutant in question, e.g. acute effects related to short-term peaks or cumulative effects related to long-term mean exposure. Statistical sampling should be considered but may be incompatible with some basic objectives. The siting of measuring stations is considered from the global to the local scale of monitoring and the relative attractions of mobile and stationary measuring stations are discussed. The problems of measurement of gases and particles are compared and a check list is put forward for use in selecting monitoring methods or instruments. Methods of presenting and interpreting monitoring data are briefly discussed. Modelling of the physical dispersion of emissions is recognized as crucial to the use of monitoring data in air quality control. Much more difficult is the interpretation of biological effects with respect to pollutant exposure and concern is expressed about the tendency to attribute cause and effect relations on the basis of correlation analysis. Monitoring is not to be confused with mindless measurement.