Camillo Dejak, Roberto Pastres, Giovanni Pecenik
Dipartimento di Chimica Fisica
Fax : 5298594
Segreteria Amministrativa :
Phone: 5298535 5298539
Stanza Dottorandi sez. Chim. Fis. Ambientale:
Phone: 5298631 5298632
This finite- difference model considers a small three dimensional grid (min 4 x 4 x 2 cells) where macronutrients and heat are introduced at constant rate in the upper corner cell and dispersed by means of a turbulent diffusive process. Eight state variables are followed: Phyto- and zooplankton densities, reduced and oxidized nitrogen concentrations, reactive phosphorus concentration, dissolved organic detritus, dissolved oxygen and sedimented organic detritus. The forcing fuctions are the input rates, the water temerature and the sunlight intensity. The last two can be calculated as indicated in the 1-D vertical model proposed by the same authors, or introduced by the user. The system can reach a steady state condition if the forcing functions are kept constant because the two horizontal walls opposite the input corner are open to the outward fluxes. A smooth behaviour of the numerical solution of the reaction diffusion equation which connect the state variables is assured by purposely studied boundary conditions. According with this approach, external values are extrapolated by assuming that the three last grid point follow a gaussian profile, whose asymptotic behaviour can be set by the user or computed using the 1-D model. The model can represent with sufficient approximation a small lake or pond not completely mixed or a shallow water basin strongly influenced by the tide, where the advection does not, on average, contribute significantly to the dispersion process. It can also been used as a quick tool for testing some improvements for an already existing combined transport-water quality model, as the computational request of the program can be conveniently satisfied by a 486 PC.
The simulation of the yearly dynamic of a two trophic level aquatic ecosystem under continuous flow of nutrients, dispersed by turbulent diffusion through the open boundaries, is achieved by using a very samll grid size. This makes the model particularly suitable for testing possible improvements for larger models of the same type, at very low computational costs.
Source of the Abstract:
Joergensen S.E., B. Halling-Soerensen and S.N Nielsen (Edts.) 1996: Handbook of Environmental and Ecological Modelling. CRC Press Boca Raton et al. 672 pp.