Simulating microplastics temporal dynamics, driving mechanisms and giving insights on sources

At the scale of a watershed, the microplastic (MP) fluxes within and between compartments contribute to the contamination of the air-soil-water continuum. The sources and fluxes of microplastics vary greatly depending on land use and human activities within the watershed. The quantifying of MP – sampling and analysis – is a time-consuming task which limits the monitoring of the fine temporal dynamics of MP. The objective of this work is to simulate the MP dynamics in the outlet of a peri-urban watershed and to gain insight into the sources. The study site is the Avenelles sub-catchment, 50 km east of Paris (France), and surface of ~50 km2, and it is an experimental site highly instrumented for physico-chemical parameters of the hydrographic network and meteorology. The subcatchment is dominated by agricultural activities, which account for 81% of its surface area, while 18% is forested, and 1% is urbanized. The arable land in the catchment is drained and its influence on the flow rate regime is characterized by flash floods. Samples of MP were collected at the watershed outlet during 2023 using a Universal Filtration Object and analysed using the micro Fourrier Transform Infra-Red. The main sources contributing to the MP dynamics are: loss of MP stocks in the soil via drainage system; remobilization of MP stocks in the sediments; the effluent of the water treatment plant (WWTP) and the stormwater overflows.The MP modelling approach is based on a multilinear model using hydrological variables. The hydrological variables used were (i) the baseflow and the quickflow, estimated from a conceptual automated process, and, (ii) the filling rate of routing reservoir and the production reservoir simulated by the hydrological model GR4H. Simulations evaluated at the MP dynamics at hourly timestep at annual and rainy season time scales. Besides the streamflow characteristics and storages, the precipitation, via the index of previous precipitation, the water conductivity and the total suspended solids are input variables as well. On an annual scale, the most significant variables in the regression appear to be the TSS, the filling rate of the routing reservoir and the water conductivity (R2=0.89). Isolating the hydrological variables, the baseflow and the total flow presented significancy (R2=0.64). For the rainy season, fast flow and total flow are the variables contributing to the MP dynamics (R2=0.86). These results can indicate the MP sources : at annual simulation, denoting TSS and base flow as significant variables, the WWTP discharge might be the main source as it is constant throughout the year; During the rainy season, presents quick flow as major contribution highlighting the drainage system and thus the MP stocks in the soil as major source. In conclusion, this simple model provides a better understanding of the sources of MP at a catchment and a better estimate of the dynamics and contamination of MP.