Aarhus University Seal / Aarhus Universitets segl

No. 414: Possibilities for modeling of environmentally hazardous pollutants in surface water

Sørensen, P.B., Andersen, H.E., Fauser, P., Bjerg, P.L., Bro, R., Holm. P.E. & Abrahamsen 20201 Muligheder for modellering af miljøfarlige forurenende stoffer i overfladevand. Aarhus Universitet, DCE – Nationalt Center for Miljø og Energi, 60 s. - Videnskabelig rapport nr. 414.http://dce2.au.dk/pub/SR414.pdf



As with the biological quality elements, the state of streams, lakes and coastal waters must be assessed in relation to the contamination conditions of environmentally hazardous pollutants (EHP) in the aquatic environment. To date, such condition assessments have been carried out by comparing measured concentrations in the water areas obtained under the auspices of the NOVANA programme with set environmental quality requirements. However, it is a resource-intensive task to carry out surveillance in all target water areas based solely on measured concentrations. A possible response to this challenge may be to develop models that can generate or assess substance concentrations that are sufficiently robust to supplement the measured values in, for example, condition assessments for use in the water plan work. Based on this, the Danish Environmental Protection Agency wants to clarify the possibilities of development and use of models for use in the condition assessments of EHP. This report intends to explore the possibilities for such use of models, with the proviso that the EHP span a very large area of a large number of substances with very different properties, and therefore the report should be seen as the first step of a major task.

The concept for modelling EHP in the aquatic environment can be established around four different domains:

  • Source domain: The source of a substance may cause emission of the substance directly into the water, soil or air environment.
  • Soil domain: In the soil domain, there are two main pools of material, one of which is material in the water phase, while the other is placed in the soil phase. The two phenomena, degradation and retention, are thus the properties that govern the release of material that is found in the aquatic environment.
  • Water domain: Typically, there will be three main phases in the water domain. The free water phase transports dissolved material and substance adsorbed to suspended substance. The sediment phase is in contact with the water phase, thereby absorbing material by diffusion and sedimentation of suspended material. In addition, biota phasemay be of great importance.
  • Atmosphere domain: In the atmosphere, the material is degraded chemically/photochemically before being redisposed either to the soil or water domain. In reality, for the water domain the atmosphere domain will typically act as a boundary condition that provides input/output of the material to the water domain.

Model-wise, these four domains can advantageously be handled in different ways, depending on the specific material in focus and the amount of available knowledge. It is therefore operational to understand modelling based on two types:

  • Source Models: Emission of material into the environment must be quantified using a model that determines the source's strength and the spatial variation.
  • Fate Models. When a substance is released from a source, it will undergo a decay, detention and transport process. Models that attempt to calculate the consequences of this fate and, thus, predict concentration levels are collectively referred to as fate models.

The best approach to modelling is strongly dependent on knowledge and therefore varies. Therefore, in this section different model archetypes with varying complexity are defined:

  • Empirical fate models: Also called statistical fate models, can be summarised as belonging to machine learning or artificial intelligence, which is also known under many other terms, such as multivariated analysis, chemometrics, etc. They vary from very simple models, such as linear and log-linear regression, to very complex and non-linear models.
  • Semi-empirical fate models: Incorporate the use of empirical data in deterministic contexts based, for example, on mass balances and relatively simple theoretical contexts.
  • Deterministic fate models: are primarily based on theoretical rules, which can be formulated mathematically. The theoretical rules are based on physico-chemical legalities as well as mass and power balances.

When considering something as complex as EHP in the aquatic environment, it is important to see the models as a tool box with many different types of tools, where different tools are used in different combinations, depending on the specific problem. Thus, source models can be used for large areas to identify areas where it would be advantageous to use more complex models or to collect measurements with the greatest benefit. Conversely, complex models may support the use of simpler models. For instance, this can be implemented by relatively complex deterministic models being used to systematically calculate a large number of realistic scenarios which, again, as a supplement to measurements can be used to calibrate empirical or semi-empirical models. Such a calibrated empirical or semi-empirical model can then be used to calculate concentration levels for the entire country. 

Relevant data sources are each described and separated according to the measured concentration in the aquatic environment and the descriptive data that is central to producing models. In addition, an overview of the relevant international models, which the project participants are familiar with, is also shown.

It is a challenge to both provide a structured and comprehensive description of the potential and possibilities of modelling the incidence of EHP in surface water and, at the same time, make specific instructions for the next step in the process, as the big picture is easily lost in the details. This challenge has been dealt with by first providing a general description of the potential for modelling based on key substances and substance groups and then selecting some of the individual substances and substance groups as proposals for the most likely first step to be used in the efforts to develop EHP concentration levels in the aquatic environment.

In general, it is a challenge to set up models with the general aim to describe the EHP in the aquatic environment at a national level. The working group behind this report therefore sees it as a start of strategic initiative and not as a complete description of all the elements necessary for modelling EHP in Denmark. The sequence of events of model development should always begin by getting a good grasp of the sources of the substance and consider the task as independent modelling task. Only when the source model is in place should the further development process of the model commence. It is also important that model development take place in close dialogue with the authorities, e.g. The Danish Environmental Protection Agency, and preferably in a cross-disciplinary environment, ideally in an interdisciplinary working group, but at minimum with a multidisciplinary advisory group.

The report describes specific activities with a three-year horizon, covering pesticides, metals and non-polar organic compounds.