Investigating River-Marine Interactions and Plastic Pollution Dynamics Using a Multimodel Approach to Support Policy Development in Environmentally Sensitive Areas

The majority of marine plastic pollution originates from land-based sources that reach the sea via rivers (Jambeck et al., 2015; Lebreton et al., 2017). In dispersion models, this process is often oversimplified as a single-point discharge at the river mouth. However, the input of macroplastics into the sea and subsequent dispersion or deposition processes —strongly influenced by the characteristics of the plastics— are significantly affected by small-scale phenomena. In coastal areas, wave-induced hydrodynamics, wind, and the complex interactions between marine currents and river inflow play a predominant role (Guo et al., 2020). These complexities are not adequately represented in Lagrangian circulation and dispersion models typically applied at the coastal, shelf and basin scale.

In this study, we focus on integrating observational data, acquired through drone surveys of stranded plastics on beaches, and simulation data obtained via a multimodel approach. This approach incorporates both phase-averaged and phase-resolving wave models. The latter are particularly suitable for small-scale processes (e.g., at the river mouth) to describe dispersion and stranding dynamics.

Our pilot study area is the mouth of the Arno River, located within the San Rossore Natural Park, an environmentally valuable area where macroplastic pollution is notably evident, as highlighted by numerous surveys (e.g. Merlino et al., 2020). The observed patterns are replicable using the hydrodynamic dispersion models employed. Observational data were collected using drone flights at various altitudes with differing levels of detail, some of which were processed through machine learning algorithms (Liu et al., 2021).

Dispersion and deposition processes at the river-mouth scale, analyzed using these two modeling approaches, reveal distinct advantages. Large-scale coastal deposition processes (spanning kilometers) are better described using the phase-averaged approach, while small-scale fluvio-marine interactions (hundreds of meters) and stranding processes are more accurately captured by the phase-resolving approach. Furthermore, this dual approach allows for the identification of the "signature" of the river on its pollution pattern at both coastal and littoral scales, highlighting the specific spatial footprint and dynamics of plastic dispersion associated with the river's outflow.

This detailed understanding provides essential guidance for policy-making and monitoring in environmentally sensitive areas, facilitating the design and implementation of more targeted strategies to reduce plastic pollution.