A simple model to assess N and P contamination in ungauged surface drainage networks. Application to the Lake of Massaciuccoli catchment, Italy

Modeling is a common practice to evaluate factors affecting water quality in environmental systems impaired by point and nonpoint losses of N and P. Nevertheless, in situations with inadequate information, such as ungauged basins, a balance between model complexity and data availability is necessary. In this paper, we applied a simplified analytical model to an artificially drained floodplain in central-western Italy to evaluate the importance of different nutrient sources and in-stream retention processes and to identify critical source areas. We first considered only a set of chemical concentrations in water measured from February through May 2008 and from November 2008 through February 2009. We then broadened available data to include water discharge and hydraulic-head measurements to construct a hydrogeological model using MODFLOW-2000 and to evaluate the reliability of the simplified method. The simplified model provided acceptable estimates of discharge (ranging from 0.03–0.75 m3 s-1) and diffuse nutrient inputs from water table discharge and in-stream retention phenomena. Estimates of PO4–P and total P retention (ranging from 1.0 to 0.6 mg m-2 s-1 and from 1.18 to 0.95 mg m-2 s-1 for PO4–P and total P, respectively) were consistent with the range of variability in literature data. In contrast, the higher temporal variability of nitrate concentrations decreased model accuracy, suggesting the need for more intensive monitoring. The model also separated the dynamics of different reaches of the drainage network and identified zones considered critical source areas and buffer zones where pollutant transport is reduced.