Integrated data-modelling approach for suspended sediment transport on a regional scale

This paper discusses an integrated data-modelling concept to monitor the se asonal variability of suspended particulate matter (SPM) patterns in the No rth Sea. It covers two aspects. First, the use of SPM transport model data to retrieve SPM concentrations from NOAA/AVHRR reflectance imagery by impro ving the algorithm to convert the reflectance data to SPM concentrations an d to generate synoptic SPM images which are consistent in time. Second, the use of these observed SPM concentrations as model output targets to assess the sensitivity of the model performance for various model input parameter s in some initial model set-ups, for example, the loads and dumping, the cr itical shear stress for erosion and sedimentation and settling velocity. The sensitivity analysis is based on the definition of a so-called Goodness -of-Fit (GoF) criterion (also denoted as cost-function) being a measure to quantify the difference bt tween the model output and the model output targ ets, which is derived from both synoptic NOAA/AVHRR imagery and in situ con centration data. Key element in this approach is the requirement that a GoF criterion is defined that mimics the main features of the end-user require ments (i.e. the modelling objective) and the associated characteristic leng th and time scales. The sensitivity analysis is carried out by means of the adjoint model which is shown to provide a detailed, that is fully spatially and temporally dis tributed, insight into the model sensitivities, The objective of this chapter is to describe the components in the integrat ed use of observations and models as outlined above. This approach is demon strated in a number of case studies of SPM transport in the Dutch Coastal Z one and in the North Sea. From the case studies, it can be concluded that l oads and dumping are a major source of error. Due to the absence of observa tions over the vertical, the errors in the erosion/sedimentation processes that govern the vertical exchange and the bed sediment load are difficult t o assess. As such, concentration profile observations and synoptic remote s ensing imagery are considered to provide an ideal and necessary combination to monitor the SPM transport on a regional scale. (C) 2000 Elsevier Scienc e B.V. All rights reserved.