Background Concentration of Chemical Substances in Aquatic Systems

It is from the discipline of Exploration Geochemistry that the notion of “geochemical background” first arose and where it has been more deeply discussed. This discussion is relevant because the same contradictions, limitations, and ambiguities are shared by aquatic systems, with the additional difficulty of their transient nature. Some of these aspects have been acknowledged (e.g., 9, 10) for example, in proposals aiming at meeting the EU Water Framework Directive. The term background has been defined with multiple and ambiguous meanings, but its most important concept is that it should not be viewed as a fixed and constant value, rather as a range of values.

The reason seems obvious because there is always some sort of random natural variability that affects all measured parameters. Even the notion of baseline concentration so widely used in groundwater studies by researchers is in the sense of a range of values as well. Nevertheless, it is necessary to define criteria on where to put a boundary on a range of concentrations that are to be considered as background.

This upper boundary is usually defined as the threshold. It should be stressed that for water quality assessment the notion of threshold is fundamentally different and is usually associated with the maximum concentration of a chemical substance in a water resource such that it can be considered safe for its defined use (drinking, agriculture, etc.). As previously observed, this may be totally unrelated to any background concentration of substances in the water system.

Statistical techniques, such as moving averages and probability plots, have been used to define background values quantitatively, meaning that a statistically representative number of samples is required. Current practice advises against the use of the most common approach which is “mean + 2 standard deviations’!

The reasons are multiple and include, for example, the fact that most concentration distribution of substances do not follow a normal (or even log-normal) distribution, instead they are multimodal, or that the mean and standard deviation should be population parameters, not retrieved from the sample. Reimann et al. suggest using the “median + 2 median absolute deviation” instead, which is more robust against the effect of data outliers. Nevertheless, by means of probability plots, the 95th percentile has been chosen for baseline studies as a starting point, such as in USA and in the UK, or the 97.5th percentile which is fairly equivalent to the “mean + 2 standard deviations” for normally distributed data.

For the purpose of responding to the European Water Framework Directive, the 90th percentile was used for small datasets (<60 samples) and where human impacts cannot be excluded, while the 97.7th percentile was used for larger datasets (>60 samples) of pristine groundwaters. Other approaches include using the upper inner fence of Turkey boxplot boundaries.

While statistical techniques might be helpful in providing criteria to define background concentration values, they bring little added information on the (bio)geochemical processes that take place in aqueous environments that ultimately control the chemistry of natural waters. As such, the current approaches tend to include historical data, down-gradient profiles, extrapolation from adjacent, geologically similar areas, and geochemical modeling additionally to the statistical approaches.

The large majority of the U.S. Environmental Protection Agency (EPA) list of priority pollutants refers to organic compounds that are solely due to human activities, and most of the inorganic pollutants occur in the natural waters at trace element concentrations (<1 mg L-1). So, to define the background concentration of pollutants this is an aspect to bear in mind, because inorganic pollutants occur naturally, with or without anthropogenic-induced disturbances, while the overwhelming majority of organic pollutants are due exclusively to anthropogenic activities.

 






Date added: 2025-01-04; views: 41;


Studedu.org - Studedu - 2022-2025 year. The material is provided for informational and educational purposes. | Privacy Policy
Page generation: 0.014 sec.