USING FIELD DATA AND SIMULATION MODELING TO DETERMINE NITROGEN-CYCLE PARAMETERS

The identification of representative parameter values is an important aspect of N modeling because many parameters are unavailable or diffic ult to evaluate, Reversed modeling was performed using a conceptual si mulation model, AgriFlux, and NO3- concentrations from a held near Que bec City, Canada, to identify N cycle parameters, A sensitivity analys is of the model with regard to the unavailable N cycle parameters iden tified three parameters related to the organic N transformations that largely control soil N dynamics (efficiency of the internal synthesis of the microbial biomass, litter humification fraction, and humus mine ralization rate), The available parameters and those of lesser influen ce were kept constant while various combinations of the three influent ial parameters were used for the simulations, An optimal parameter set was identified using the root mean square error procedure, and its un iqueness was confirmed by the isocontouring technique, For the site st udied, low values of humus mineralization combined with low values for the efficiency of the microbial biomass and small transfers from the litter pool to the humus pool generated the NO3- concentrations that c ompared the most favorably with the measured values, However, a fittin g difference was observed between simulated and measured concentration s that can be imputed to nonoptimized parameters or processes not diff erentiated in the model.