Bathing water standards in real-time

A review of: Stidson, R. T., Gray, C. A. & Mcphail, C. D. (2012). Development and use of modelling techniques for real-time bathing water quality predictions. Water and Environment Journal 26, 7-18.

Real-time predictions are advocated as an appropriate strategy in assessing bathing water quality in response to the recent change in approach to maintaining bathing water standards, prompted by a new European Union (EU) directive and current World Health Organisation (WHO) guidelines. The approach relies on the prerequisite of suitable information being made available to the general public, whilst ensuring an accurate description of water quality and effective risk management strategies based upon the successful monitoring of results. Increased access to a comprehensive store of information is advocated to permit beach users to make an informed decision on how they may choose to use bathing waters at any given time.

The new directive has ignited debate which questions its integrity in assessing water quality in the face of short-term pollution incidents, due to outlining a practice which allows samples taken during transient pollution events to be disregarded in order to formulate annual compliance calculations. Due to slow experimental methods and the close correlation between periods of high faecal contamination and antecedent rainfall and river flow, predictive tools united with rapid molecular methods are advocated to be a potentially effective strategy to providing adequate management measures.

Since 2004, the first real-time bathing water quality predictions have been developed by the Scottish Environmental Protection Agency (SEPA) using 10 sites located throughout Scotland. The paper focuses on how models can be constructed whilst accommodating key requirements and complying with regulatory framework. The paper considers two stages in the advancement of
modelling tools [the basis of these primitive real-time predictions] founded upon antecedent rainfall and river flow, using data from four of the ten sites.

Modelling tools coupled with related advisory systems are shown in the study to demonstrate a possible reduction in public exposure to sites of undesirable water quality, associated with discrete short-term pollution incidents derived from rainfall events. Additionally, these tools could be utilised in predicting and describing water quality, permitting the development
of alternative management strategies complying with the new EU directive. Supplementary
approaches could be useful in compensating for contentious sampling guidelines which disregard samples taken during periods when bathing is not advised (off-season) and samples affected by transient pollution events.

From the assessment provided in this paper, modelling tools show promise as a flexible and powerful method in maintaining bathing water by predicting standards. Through successfully representing relationships between variables, predictions can be formulated in consideration of data affected by multiple factors, showing that models of this nature hold the potential to improve and evolve when applied to new datasets.

Nevertheless, a comprehensive appraisal of predictive models should be conducted to better assess their performance. To successfully evaluate this approach, current modelling techniques can be improved and the data predicted should be tested over a broad period of time. Additionally, I think that further environmental variables which are acknowledged to influence bathing water alongside antecedent rainfall and river flow (e.g. UV radiation, tidal patterns
and wind), should be integrated into a more holistic review. All the same, the evolution of modelling tools in the future could lead to their inclusion as a strong supplementary
strategy, assisting in the successful management of bathing waters in the UK.