Dresden, Germany Stormwater runoff transports particles and contaminants, which are highly mobile in the urban water system. Their export shows significant temporal variability described by pollutant flush types. Understanding this variability is essential for improving monitoring and proposing stormwater pollution control strategies at the urban catchment scale. Hence, we characterised the sediment export and element patterns from a stormwater outlet in Dresden (Germany) using both grab samples and high-resolution monitoring data during rainfall events. Our results showed that the stormwater discharge consisted mainly of fine (< 63 µm) and inorganic sediments, representing ∼80 % of suspended sediments. Pairwise associations and a hierarchical cluster analysis revealed strong Kendall correlations among fine and coarse suspended sediments, their organic content, and elements (i.e., Al, Ba, Cu, Fe, Mg, Mn, Zn), indicating similar transport mechanisms. These variables clustered with turbidity, emphasizing its potential as an easily measurable proxy for evaluating the dynamics of particle-bound contaminants through continuous monitoring. Hydrological descriptors may explain the variability of flush types. In the analysed catchment, second flush events could be linked to preceding higher-intensity rainfall, highlighting the influence of antecedent conditions on transport dynamics. The occurrence of two pollutant flush types through the year and the existence of both anti-clockwise and clockwise hysteresis patterns provide insights into delayed transport mechanisms, highlighting the need for flexible infrastructure in stormwater management. • Fine and inorganic particles prevail in the stormwater from an urban catchment. • A middle flush mostly transports fine particles, suggesting temporal storage. • Particle-bound element transport differs from the traditional first flush model. • Turbidity serves as a proxy for particles, organic, and elemental composition. • Continuous monitoring allows capturing stormwater sediment dynamics more accurately.
Rojas-Gómez et al. (Thu,) studied this question.