Integrated modelling of suspended matter in the North Sea

The topic of this paper is the modelling of suspended particulate matter (S PM) patterns and their variation on North Sea scale over the time scale of several months. The modelling results are assessed by a formalized comparis on with concentrations derived from NOAA/AVHRR reflectance imagery. The nat ure of the study is a sensitivity analysis of the various modules forming t he integrated SPM model: the hydrodynamic model, the wave model and the SPM transport and erosion-sedimentation model. A Goodness-of-Fit criterion is introduced as a single parameter to quantify the difference between SPM mod el results and data. Its specification reflects the aim of the model, which in this study is the representation of the seasonal behaviour of SPM patte rns, and the determination of the sensitivity to uncertainties in model par ameters of the transport model, its inputs and forcing or boundary conditio ns. This Goodness-of-Fit method is applied to a North Sea wide scale SPM mo del application in both 2-D and 3-D using a curvilinear grid schematization . Results are presented for the period March-September 1994. The results co nfirm the suitability of NOAA/AVHRR based data to assess the modelling and the importance of the net volume flux which is input to the SPM model. The uncertainties in the present estimates of inputs from e.g. erosion and dump ing, and the erosion-sedimentation model parameters are shown to also have a major impact on the modelling results. The approach using the Goodness-of -Fit criterion as such is shown to lead to an objective and reproducible qu antification of the model quality. (C) 2001 Academic Press.