Sensitivity Analysis of VELFEEM –Venice Lagoon Finite Element Ecological Model- to the Macro-Nutrient Input Regime

A sensitivity analysis of VELFEEM, the Finite Element Ecological Model for the lagoon of Venice, has been performed in order to test the responses of the model to changes in external input regimes. The model is obtained by internally coupling a hydrodynamic model, an energetic model, and an ecological model. The hydrodynamics are simulated using SHYFEM, a barotropic bi-dimensional model based on a finite element discretization of the spatial domain, which allows for a very good spatial resolution of the lagoon morphology while keeping at a low level the computational demand. Using a standard heat fluxes formulation, the energetic module computes the water temperature of each element starting from meteorological daily measurement. The ecological processes are simulated by the evolution of nine state variables, namely by phytoplankton, zooplankton, nutrients (ammonia, nitrate and phosphate) organic detritus (organic nitrogen organic phosphorous and CBOD) and dissolved oxygen. In a previous work, the role played by physical forcings in the definition of the water quality level has already been investigated, and the conclusion was that a proper parameterisation of these processes would increase the accuracy of our model prediction. Here we aim to see whether the same consideration could be extended at the macronutrient input regime. Assuming that the total amount of macronutrients and of freshwater entering the basin is the same in our scenarios, we vary the river input regime comparing an idealised scenario with a realistic scenario. The sensitivity to the variation in input regime and variations in the macro nutrient exchange with the sea has been investigated, by comparing model predictions of spatial and temporal evolution of major state variables and of an aggregate index of water quality (TRIX). In the idealised scenario the water discharge is represented by an idealised function and the concentration in macronutrient is constant during the whole year, and it is the same for each river, during the one-year simulation. In the real scenario we use monthly data field of river flow and the concentration values sampled at each river.