Heavy metal contamination in the Geul River: Assessment of the metal inventory in and export from the river valley.

Abstract

One of the most important environmental problems in many fluvial systems is that of heavy metal contamination. Metal mining has caused spreading and accumulation of heavy metals in many river basins during the last century. As it has been discussed in previous studies (Dennis et al., 2008) floodplains can act as a metal storage for hundreds or thousands of years and become a source of heavy metals during floods. Meandering rivers, such as the Geul River, are very dynamic systems where erosion and deposition processes can produce remobilization and redistribution of heavy metals over the catchment altering the quality of the water and disturbing the flora and fauna of the area. Soil sample collection in the Dutch part of the Geul catchment was carried out during this project with the purpose of studying the distribution of zinc and lead within the Geul catchment. A total of nine cross-sections were performed in the upstream direction with a 4 km interval. The analysis of the samples was performed using a hand-held X-Ray fluorescence spectrometer (Thermo Fisher Scientific Niton® XL3t-600 hand-held XRF analyzer) and the results were statistically analysed. The statistical analysis of the samples shows a decrease in heavy metal content in the downstream direction. A general decrease in zinc and lead content was also found when the distance from the river increases. Results of the statistical analysis show by higher significance that downstream distance from the pollution source is a useful variable to explain the distribution of the heavy metals along the Geul catchment. This study aimed to assess the annual export of zinc and lead from the Geul floodplain obtained after a one year simulation using the CAESAR model. These calculations were based on the average values of zinc and lead provided by the analysis of the samples previously mentioned together with the sediment results after a one year simulation of the CAESAR model. The assessed amounts of lead and zinc deposition for the year 2008 were 14.5 and 57.9 tons respectively. By comparing total and suspended sediment results of the model with other studies (Leenaers, 1989; De Moor, 2006) the results of the CAESAR model seem to be between 3-5 times higher. The reason for this can be that the model uses a 25 meter resolution, which makes the river wider than in reality and consequently more sediment is generated. Other irregularities in the Digital Elevation Map and the ‘spin up’ process where the model starts assuming uniform bed material over the whole catchment and generates high amounts of sediment during the first years of the simulation, can influence the final sediment results and therefore, the calculated amounts of zinc and lead can be overestimated.