Project: Funded Project › Research
1/07/08 → 31/12/13
Phosphorus Export and Delivery from Agricultural Land (PEDAL2)
To meet Water Framework Directive (WFD) responsibilities there is a need to estimate the likely effect of current and future policies on agricultural land-use and land management and subsequently on the quality of receiving waters in agricultural catchments. Additionally, where targeted mitigation measures are required to reduce diffuse water pollution from agriculture there is a need to estimate the most effective strategies. These strategies include establishing and implementing programs of measures (POMs) within river basin districts and identification of Water Protection Zones where restrictions on activities resulting in pollution of watercourses may be restricted by law. Programs of measures aimed at reducing pollution may be site-specific or local/regional, more generally applied through cross-compliance, meeting codes of good agricultural practice or through environmental stewardship schemes via the Entry Level Scheme or in some circumstances through the restricted number of Higher Level Schemes in important areas such as Sites of Special Scientific Interest. In all of these schemes it is important to assess likely efficacy and practicality. Owing to uncertainties in our ability to predict the effects of the many interacting processes at appropriate scales (e.g. farm and small catchment), these estimates are not an exact science. This has led organisations such as the Environment Agency to employ risk-based decision making strategies where confidence in predictions is taken into account (http://www.environment-agency.gov.uk/aboutus/512398/516810/516839/517298/?lang=_e). These types of approach recognise that making decisions using single-value estimates, that are known to be inaccurate, does not help policy makers understand the likelihood of being wrong and how the risk of being wrong may vary spatially. The framework proposed here will be established so that uncertainties in our knowledge and available information are propagated through the system to give more robust estimates. This will be achieved by using fuzzy information and fuzzy rules of catchment and farming system function (fuzzy mathematical techniques are useful for dealing with imprecise data and knowledge where a range of outcomes are possible). Without consideration of our current abilities to estimate the behaviour of diffuse pollutants, determination of diffuse pollution mitigation measure effectiveness, and hence cost-effectiveness, can be misleading.
This project will advance and expand a methodology developed in an earlier project (PEDAL: Phosphorus Export and Delivery from Agricultural Land - PE0113) which considered the delivery of phosphorus (P) in headwater catchments. In addition to considering P delivery, the proposed model will use similar fuzzy-rule based methodologies to estimate DESPRAL P mobilisation (PE0106) for unmonitored catchments, and hence P available for delivery. Expansion of the model in this way will allow predictive estimates of mobilisation and delivery to be made at unmonitored locations. Superimposed upon this structure will be a set of expert-derived rules which modify model output to simulate the likely effect of given mitigation such that estimates of their effectiveness can be made. This will allow environmental managers to target the implementation of mitigation measures (e.g. the programmes of measures to be made operational in WFD River Basins by December 2012).
In addition, a fuzzy-rule based model for faecal indicator organisms (FIOs) will be developed in a similar manner to the improved PEDAL model. The philosophy behind PEDAL is to make estimates that reflect our confidence given the data available for model evaluation using model inputs and parameters that are functionally significant in describing our observations. The resultant fuzzy ranges can, however, be updated when additional information becomes available, thus providing a framework where new information can be incorporated. In this way, the PEDAL approach gives realistic estimates which explicitly take into account factors which make the prediction of diffuse pollution uncertain. Failure to account for uncertainty may lead to ineffective policy decisions made without all available relevant information.
This project will deliver:
• an improved PEDAL model which will provide location-specific estimates of P mobilisation and delivery at unmonitored catchments together with a quantitative assessment of the confidence associated with predictions;
• an improved catchment visual assessment (VA) methodology and a formal protocol for combining VA information with estimates of P delivery coefficient such that a refined P delivery estimate is provided;
• an FIO model based upon similar concepts to the improved PEDAL model;
• location-specific estimates of the likely effect of mitigation measures on predictions of P and FIO fluxes, which will allowing targeting of POMs