Time, Science and the Water Framework Directive
If Europe is to move closer to achieving the clean rivers and lakes that are the goals of the Water Framework Directive, future investments must be chosen taking the relevant science into account, together with its uncertainties - writes Keith Beven, professor of hydrology at Lancaster University.
It is now 2006; there are nine years left for all countries in the EU to decide on how to designate water bodies at risk, to decide how to define good ecological and chemical status, and to decide what sustainable use might mean, particularly in those parts of Europe that are particularly vulnerable to climate fluctuations or potential climate change. The Directive does allow for some freedom in interpreting these phrases at national levels.
The majority of water bodies in Europe are not in a pristine state. They have been modified by man, either directly though the construction of dams or inputs of effluents, or indirectly through, for example, upwind air pollution or the wash-off of sediments and nutrients from agricultural land.
Good ecological and chemical status clearly therefore does not mean pristine. It has been suggested that even to get lake nutrient concentrations back to pre-1940 levels in the UK would require stopping nearly all agriculture and waiting, in some catchments, for a very long time.
In very many rivers of Europe, good ecological and chemical status clearly also do not mean the current status. Despite widespread improvements in water quality (partly as a result of previous EU Directives), there remain many rivers and lakes where quality is poor. So investments will be needed, either directly (further improvements to effluent treatments) or indirectly (more widespread agricultural subsidies aimed at encouraging better environmental stewardship), to make improvements. Decisions about investment priorities require prediction of the impact of those investments into the future. Accurate prediction of such impact requires models of the processes that properly reflect the scientific understanding of the catchment system.
Unfortunately, that understanding and the prediction of the models based on it are inherently uncertain. The catchment system is an open system subject to uncertain precipitation forcing that responds in quite non-linear ways dependent on the nature of a precipitation event and the antecedent wetness of the soil prior to the event.
The moisture status of the soil controls how much of the rainfall runs off over the surface of the ground with the potential to transport sediments and nutrients, but depends on how water moves below ground - which is precisely where it is most difficult to study.
We have relatively little knowledge, for example, of the long-term residence times of water within the catchment system. Thus understanding of how nutrients or other pollutants might move through the subsurface system is relatively poor. It is also complicated by the effects of people. Catchment properties are changing as a result of changes in land management, installation of artificial drainage, and urbanisation.
This leads to problems in prediction. The understanding on which our models are based is necessarily approximate, the data to drive the models are subject to uncertainty, the characteristics of catchments contributing to different water bodies are uncertain, and the character of the system and its external forcing might be changing over time.
The predictions on which decisions must be based will therefore also be uncertain, and it might be that taking account of that uncertainty within a risk-based decision framework might affect the decision made.
The implementation of the Water Framework Directive has provided some interesting examples of this issue. It is known, for example, that in the chalk aquifers of southeast England, an annual peak of nitrate concentration can be distinguished in the unsaturated layers above the water table. The structure of the chalk rock means that the nitrate is moving downwards at rates of 50 to 100 cm per year without the annual signal being smoothed too much.
The water table is, in places, tens of metres below the ground but is a critical source of drinking water in an area where the demand for water is currently increasing. Simple arithmetic suggests that much of the nitrate arising from the intensification of the agriculture on the chalk since 1950 has yet to reach the water table and therefore has yet to contaminate the water supply.
The fact that more nitrate will reach the water table is not uncertain. The fact that there will be a supply of nitrate from the unsaturated chalk that will last for decades is not uncertain. Predictions of the spread and duration of the impact of this nitrate, however, remain highly uncertain.
There is a similar issue with respect to phosphorus reaching rivers. In freshwaters, phosphorus is a very important control on eutrophication of rivers and lakes with consequent ecological impacts. Availability of phosphorus plays an important role in the algal blooms that sometimes blight lakes and reservoirs in summer. The principle sources of phosphorus are from agriculture, domestic sources including septic tanks, and sewage treatment works.
There have been significant reductions in phosphorus inputs to rivers in the last 20 years as a result, for example, of the elimination of phosphorus from household detergents and washing powders and improvements in sewage treatment. This reduction can be seen in the water quality data collected by the Environment Agency.
There is still too much phosphorus reaching or retained in the sediments of rivers and lakes, however, and in some sensitive locations further measures have been taken such as the introduction of phosphorus stripping plants in sewage treatment works. This adds additional costs to water bills but may not be sufficient if there are continuing supplies of phosphorus from diffuse agricultural sources where the problem may be very much more difficult to deal with.
Here, the science is much less certain. Most of the phosphorus added to the soil by farmers in the form of animal slurries or artificial fertilisers is either used by the crop or retained in the soil. It is only a very small proportion of what is added that is transported to streams by water, either in a dissolved form or attached to soil particles and fine colloidal particles. Measured concentrations of phosphorus vary significantly in both space and time and assessing how much of what is measured might reach stream channels (and what might happen to it in streams and lakes) is really rather uncertain. Thus predicting the relative importance of these diffuse sources of phosphorus and the point sources from sewage treatment works and septic tanks is difficult.
In terms of implementing the Water Framework Directive, it might be easier to deal with the point sources than the diffuse sources but this will not solve the problem - even dealing with the diffuse sources may not solve the problem unless there is parallel attention to sediment stores of P in rivers and lakes.
The science may be uncertain but what we can be sure of is that if we pay farmers to ensure their farms are managed so as to be in nutrient balance (the nutrients added are equivalent to those used by the crop; this is already a requirement in some European countries). The storage of phosphorus in catchments is already such that there will be outputs to streams far beyond the 2015 date for the implementation of the Water Framework Directive.
Thus, even these simple (and uncertain) evaluations of the problem suggest that in the UK the achievement of good ecological and chemical status for sustainable use will not be possible by 2015. However, as with other European Union Directives, the initiative will have value in identifying those catchment areas that are most vulnerable and in providing the impetus to move towards better environmental quality.
There is a real question, however, as to whether adequate science exists to provide good predictions of the impact of different investment strategies in making those improvements. Defra has funded significant research into understanding of how nutrients are delivered to stream channels. Understanding has undoubtedly improved but has also proven difficult to extend beyond the small catchments that have been studied in detail. Much remains to be done, and in making decisions about how to spend tax and water industry revenues, decision makers are going to have to grasp the nettle of uncertainty in the science.