Organic Pollutants in the River Rhine Network
- Modeling, Simulation, Visualisation -

Intermediates are produced in large quantities by chemical companies located along the River Rhine. Production volumes of 50 chemical intermediates and annual estimated release, location and capacity of the German companies were taken from ECETOC (Environmental Exposure Assessment, Technical Report No. 61, 1994). Because of the large quantities and the substance character as intermediates continuous point discharges were assumed.

The Rhine and its main tributaries were divided into stretches with one gauging station in each stretch. The boundaries of the stretches were chosen by the mouths of tributaries or barriage weirs, so homogenous hydrologic properties can be assumed within one stretch. Each stretch is subdivided into boxes with one discharge point in each box. The mean flow and mean water level of the summer and the winter half-year were taken from 1989; based on long-term statistics flows can be chosen from a log-normal distribution. Flow velocities in each section were estimated by regression equations for the Rhine or from rectangular cross-sections for the tributaries.

A deterministic fate model assuming advection, dilution, volatilisation, sedimentation and degradation was used to calculate downstream concentrations. The model was implemented as a stand-alone programme with a specified data interface. Using a tree-walking algorithm the model calculates concentrations on a river network defined by a special data structure representing the network topology. All model inputs and internally calculated values are stored to ensure quality assurance. The model results are written into files to be analysed by special tools and/or GIS.

All spatial and chemical data were stored in a database coupled with the desktop GIS ARCView. A User Interface was implemented in Avenue providing full model and data control: going through the dialougs the user is assisted in selecting and editing all necessary data. The Interface extracts the data from the database and runs the model. The results of a simulation run are visualised for the whole river network by colouring the river boxes using the ARCInfo/ARCView methodology of Dynamic Segmentation and by plotting the concentration profile. Stochastic calculations provide frequency distributions of the predicted concentration to be used in risk characterisation.

Volatilisation, depending on water flow velocity and wind speed, was seen as the most important form of loss for many intermediates. Calculated concentrations were in the same order of magnitude as measured values for volatile intermediates. It was also shown that chemical discharges of more than one site are accumulated downstream. They cause higher concentrations than when considered separately.