Coupling system model with fuzzy logic rules for use in runoff and total phosphorus load prediction in a catchment

Tackling the problem of eutrophication in fresh waters is at the top of the agenda for the implementation of the Water Framework Directive (WFD) in Europe. The problem is caused primarily by an increase in phosphorus loading from diffuse sources. Therefore there is a need to apply appropriate measures, which are able to reduce the phosphorus diffuse pollution, at a catchment scale in each River Basin District (RBD). As the implementation of such measures disturbs the existing system in the catchment it is important to be able to predict their impact and this requires a reliable mathematical model to represent the system. In this study, a new, lumped catchment, methodology to improve on an existing diffuse phosphorus pollution model, the Grid Oriented Phosphorus Component (GOPC) model, is proposed. This methodology consists of two elements; (i) the Soil Moisture Accounting and Routing (SMAR) hydrological model was used to provide the required hydrological variables to the GOPC model; and (ii) fuzzy logic rules were formulated with the notion that each rule corresponds to a sub-model representing a particular hydrological behaviour in the catchment and the combined results of all rules give the total response. Sixteen modelling cases, each of which uses different numbers of fuzzy sub-sets for the rainfall and the evaporation, were compared for their discharge and total phosphorus (TP) simulations in a catchment in Northern Ireland. The comparison was based on the validation results as they allow testing the applicability of the models for conditions different from those used in the calibration period. Using 2 fuzzy sub-sets for the rainfall and a single fuzzy sub-set for the evaporation produced the best simulation for the discharge whereas the best TP simulation was obtained from the case of 4 rainfall fuzzy sub-sets and 3 evaporation fuzzy sub-sets.