Quantifying excess heavy metal concentrations in drainage basins using conservative mixing models

This is a Preprint and has not been peer reviewed. The published version of this Preprint is available: https://doi.org/10.1016/j.gexplo.2023.107178. This is version 4 of this Preprint.


Download Preprint


Jonas Alexander Eschenfelder , Alex G. Lipp, Gareth G Roberts


High concentrations of heavy metals and other pollutants in river sediments can have detrimental effects on the ecosystem and humans. The composition of river sediments throughout drainage basins therefore provides important information for environmental monitoring. An obvious first step for using river sediment compositions for monitoring is to quantify natural baseline concentrations. Once baselines have been quantified, is it straightforward to compare them to observations to identify excesses generated by, for example, anthropogenic inputs. In this study a new strategy for mapping element concentrations along rivers from discrete geochemical observations upstream is presented. We demonstrate our approach in a case study of the Clyde drainage basin in western Scotland, UK. First, continuous baselines are generated using simple forward models that conservatively mix source region concentrations along drainage networks. 1185 measurements of elemental concentrations from first-order streams are used to parameterise the source region. The calculated baselines are then compared to concentrations measured at 60 localities along the main channel of the Clyde river. For a range of major and trace elements (e.g., Mg, Sr, K, Mn), the downstream observations are in close agreement with baseline concentrations predicted by conservative mixing models. However, some heavy metal concentrations (Pb, Cu, Zn) tend to exceed predicted baseline concentrations. Therefore, the second part of our approach calculates element concentrations in source areas required to match the observed Pb, Cu and Zn concentrations measured along the river. An inverse approach is used to `unmix' the observed concentrations utilising, again, a conservative mixing model. Model resolution is determined by the spatial distribution of the data. Resultant calculated natural baselines and heavy metal concentrations along the river can easily be compared to estimate excesses. We tentatively suggest that anthropogenic input to sediment composition along the river is equivalent to annual fluxes of 9.7, 1.5 and 5.7 tonnes (10^6 g) of Pb, Cu and Zn, respectively.




Earth Sciences, Environmental Sciences, Geochemistry, Geomorphology, Physical Sciences and Mathematics, Sustainability, Water Resource Management


River pollution \and Clyde, Scotland \and Geochemical modelling \and Heavy metalMixing, River Pollution, Clyde Scotland, Geochemical Modelling, Heavy metal, mixing


Published: 2022-05-14 04:24

Last Updated: 2023-03-14 06:42

Older Versions

CC BY Attribution 4.0 International

Additional Metadata

Conflict of interest statement:

Data Availability (Reason not available):
The data used was provided under licensce but can be requested from the BGS

Add a Comment

You must log in to post a comment.


There are no comments or no comments have been made public for this article.