Modelling hydrology and water quality in a Mediterranean catchment

In this study the SWAT model has been used in order to analyse and quantify pollution dynamics at basin scaledepending on concentrated and diffuse sources. Nowadays, the receiving water bodies quality safeguarding isof growing importance due to the promulgation of recent laws as well as the growing sensitivity regarding theenvironment issues by the scientific and practitioner committee. Recently the EU 2000/60 (Water FrameworkDirective) makes the analysis of receiving water bodies even more complex by integrating the pollution in urbanareas in a framework of the pollution sources at catchment scale. and making necessary further integration ofenvironmental impacts associated with discharges concentrates civilian and productive with the widespreadpollution linked mainly to agriculture and zoo-technical activities. The complexity of natural systems and the largenumber of polluting sources and variables to be monitored requires the adoption of models able to get a betterview of the whole system in a simplified way without neglecting the most important physical phenomena.Particularly, in this study the SWAT model was considered since it is an integrated hydrological model that are,nowadays, needed to support the implementation of integrated water management plans and to comply with thecurrent requirements of the WFD. In addition, the SWAT model is interfaced with the ARC-VIEW software whichallows easy pre-and post processing of the spatially distributed input data, driving the rainfall-runoff process.The model has been applied to the experimental Nocella catchment located in Sicily (Italy), with an area of about50 km2. The river receives wastewater and stormwater from two urban areas drained by combined sewers. Thestudy demonstrates that the analysis of water quality in partially urbanised natural basins is complex depending onvariable polluting contributions of the different parts of the system depending on specific polluting compounds.The model was calibrated and then validated, obtaining satisfactory performance. The estimation of loads fromdiffuse sources was difficult due to limited data availability. Thus, it was only possible to include constant diffusepollution concentrations at present. In spite of these limitations, the model captured rather well the dynamic offlow generation and was able to predict the range of nutrient concentrations in surface water. The contribution ofurban areas to the polluting loads at catchment scale is relevant especially during the dry season.