Skip to main content

Inverse modelling of transport distance to reduce ambiguities of microbial and chemical source tracking in karst catchments

The identification of contamination sources is vital for water protection, especially in highly vulnerable karst aquifers. Contamination sources might be distinguished by host-specific DNA markers of bacteria (Microbial Source Tracking, MST) or source-specific indicator compounds (Chemical Source Tracking, CST). These methods can help to identify a type of contamination source but fail to distinguish similar contaminant signals from different origins, e.g. multiple points of wastewater infiltration. Transport modelling can reduce these ambiguities by considering the time course of contaminant concentration, thereby allowing for a better allocation of the input source. However, flow in karst aquifers is highly heterogeneous and very dynamic. Hence, distributed numerical transport models on catchment scale are complex, difficult to parameterise and suffer from manifold ambiguities. Here, an approach is presented, which aims at improving identification of contamination sources by combining MST/CST with transport modelling. Fast (conduit) transport is represented by a 1-D problem and a maximum transport distance for contamination events is modelled. The model is based on (semi-)analytical solutions of transport models, well-established in tracer test analysis to estimate apparent tracer velocities. In this study, a-priori knowledge about velocities and input times is used to inversely model transport lengths from contaminant breakthrough curves. The inverse transport model (implemented in GNU Octave) was validated and parameter sensitivities were analysed. The maximum transport distance approach was shown to perform well during periods of flow recession. It was applied successfully to a contamination event at a karst spring and allowed for assigning its input to a stormwater tank.

Details

Author
Johannes Zirlewagen1, Ferry Schiperski1, Tobias Licha2, Traugott Scheytt3
Institutionen
1Technische Universität Berlin, Germany; 2Ruhr-University Bochum, Germany; 3TU Bergakademie Freiberg, Germany
Veranstaltung
GeoKarlsruhe 2021
Datum
2021
DOI
10.48380/dggv-01ev-9c24