Microplastics (MP), which are particles smaller than 5 mm, can be found in various aquatic environments, including the oceans, seas, estuaries and rivers (Cole et al., 2011; Jambeck et al., 2015; Setälä et al., 2016; Mishra et al., 2022; Matjašič et al., 2023). The significant increase in plastic production since the early 1970s has raised numerous concerns about plastic pollution in aquatic systems. It has been estimated that over 170 trillion plastic particles are floating in the world’s oceans (Eriksen et al., 2023), and their presence is also increasing in the seabed, coastlines, and marine biota (Barnes et al., 2009; Suaria and Aliani, 2014; Llorca et al., 2020; Matjašič et al., 2023). Despite substantial efforts and initiatives to reduce plastic usage, global annual plastic waste production is projected to continue rising in the coming years. By 2025, the United Nations Sustainable Development Goal 14.1 aims to reduce marine pollution, including plastics.

In Europe, The Marine Strategy Framework Directive (2008/56/EC, European Commission, 2008) (MFSD) identified anthropogenic litter as a dominant pressure and a main source of impact on coastal habitats. The MSFD establishes requirements for the EU (European Union) member states to achieve and maintain a good environmental status in their marine environments, as well as to prevent any future deterioration including the MSFD descriptor D10. In addition, the European Chemical Agency (EGCHA) has proposed restriction of MP in many products within the EU/EEA (European Economic Area) region, with the goal of preventing or minimizing their discharge into the environment (European Chemicals Agency, 2019). In 2021, the European Union also banned single-use plastics within its member states (Harvey and Watts, 2018).

According to GESAMP (2019), the marine environment can be infiltrated by plastic through multiple entry points, including riverine systems, shoreline activities, shipping, and atmospheric deposition. Various studies (Ziajahromi et al., 2016; Mintenig et al., 2017; Kay et al., 2018; Prata, 2018; Schernewski et al., 2020) have highlighted the significant influence of human activities on MP deposition. Among these activities, Wastewater Treatment Plants (WWTPs) are recognized as a significant emission pathway. For example, Municipal WWTPs have shown high efficiency in removing MP (Carr et al., 2016; Talvitie et al., 2017; Gies et al., 2018); however, untreated WWTP effluents exhibit elevated MP concentrations (Sun et al., 2019; Schernewski et al., 2020). Baresel and Olshammar (2019) proposed that MP retention in WWTPs based on their respective treatment stages ranges from 85% to 98% in the Baltic Sea region. Despite this relatively high overall removal efficiency, WWTPs are still considered a significant MP emission pathway in the Baltic Sea region due to the substantial volumes of wastewater they process (Baresel and Olshammar, 2019). In the Baltic Sea, wastewater and stormwater plants are typically separated (Schernewski et al., 2020). Sewer overflows, comprising stormwater and untreated wastewater can substantially contribute to the MP load in the environment (Magnusson, 2016; Dris et al., 2018). During periods of heavy precipitation, stormwater serves as a critical entry point for MP into the aquatic environment. Baresel and Olshammar (2019) suggest that the yearly discharge from sewer overflows is comparable in magnitude to that of treated wastewater.

Several studies have indicated that rivers are a primary source of MP and play a crucial role in transporting plastic waste into oceans (Jambeck et al., 2015; Siegfried et al., 2017; Schrank et al., 2022). Rivers flowing through highly populated cities with significant industrial activity along their banks may serve as an important source of MP in the estuarine bays such as the Gulf of Finland (GoF) in the Baltic Sea (Martyanov et al., 2021). It has been estimated that between 1.15 and 2.41 million tons or more of plastics are deposited annually into oceans via rivers (Lebreton et al., 2017; Schmidt et al., 2017). Numerous studies have discussed the pollution patterns of large rivers and provided insights into the regional and global factors responsible for MP pollution in the water column and sediments (Matjašič et al., 2023). High variability in MP concentration can be seen in both the water column and sediments depending on factors such as sampling methodology, anthropogenic activities and the size of the catchment area (She et al., 2022; Matjašič et al., 2023). Additionally, a significant proportion of marine beach litter is attributed to the input of plastic waste into rivers (Veerasingam et al., 2016). However, it is important to note that this study does not consider river retention in its analysis.

The Baltic Sea, located in northern Europe, is known as one of the largest brackish water bodies in the world (HELCOM, 2023). With a catchment area four time larger than its surface area (372,858 km²) (Marko and Urs, 2013) and an average depth of 55m, the Baltic Sea faces significant challenges related to marine litter (HELCOM, 2023). Coastal areas along the Baltic Sea exhibit significant concentration of beach litter (HELCOM, 2023). Plastic materials make up the most frequently encountered marine litter in the Baltic Sea (HELCOM, 2023). As of 2023, HELCOM’s aim to substantially decrease plastic waste and mitigate its harmful effects on coastal and marine ecosystems remains unfulfilled (HELCOM, 2023). The Baltic Sea receives a substantial volume of water from various rivers, with an average combined flow rate of approximately 14,085 m³/s (Meier and Kauker, 2003). Due to the extended residence time of pollutants in the Baltic Sea during the water renewal period, which can last up to 30 years (Leppäranta and Myrberg, 2009), the pollutants present in the Baltic Sea have a significant impact on the aquatic environment. Consequently, it is reasonable to assume that the Baltic Sea serves as a major hotspot for plastics, primarily through river discharge. Large impacts of riverine inputs as critical pathways for plastics into marine environments has also been identified in other parts of the world (Vianello et al., 2018; Uaciquete et al., 2024).

The GoF is an elongated estuarine basin situated in the northeastern region of the Baltic Sea with an average depth of 37 m and a maximum depth of 123 m (Leppäranta and Myrberg, 2009). The gulf stretches approximately 400 km in length, with a width that varies between 48 and 135 km (Alenius et al., 1998). There is a free water exchange between the GoF and BP at the western border, and fresh water is discharged mostly to the eastern part of the GoF. Several studies have reported the presence of MP in the GoF (Lips et al., 2020; Setälä et al., 2016; Uurasjärvi et al., 2021; Mishra et al., 2022). However, the knowledge about the spatial and temporal variation of MPs in the Baltic Sea is limited (Aigars et al., 2021). In addition, the methodology for acquiring information about MPs can vary based on the instruments utilized, mesh size, sampling depth, and the extent of the sampling area (Mishra et al., 2022; She et al., 2022).

Modeling the movement and fate of MP is particularly relevant in semi-enclosed systems like the Gof, where limited exchange and localized inputs contribute to accumulation of marine debris (Tsiaras et al., 2021). Eulerian and Lagrangian models are commonly used in such simulations (Bigdeli et al., 2022). Lagrangian modeling, also known as particle tracking modeling, tracks individual particles (Siht et al., 2025, in press), while an Eulerian approach considers advection and diffusion at specific locations (Bigdeli et al., 2022). Pärn et al. (2023) employed a combination of a hydrodynamic model and a particle tracking model to understand the transport and fate of marine litter including accumulation areas in the Baltic Sea. Martyanov et al. (2023) considered different initial fall velocities of suspended MP to study their distribution in the eastern GoF. Schernewski et al. (2021) incorporated emission scenarios from WWTPs and combined sewerage plants into their model to estimate the fate of plastics in the Baltic Sea environment. The GETM (General Estuarine Transport Model) ocean circulation model has been utilized in several studies, including those conducted by Schernewski et al. (2021) and Osinski et al. (2020), to analyze the transportation of MP in the Baltic Sea. However, these studies did not include the impact of biofouling on the buoyancy of floating MP and their removal process through sinking and sedimentation (Osinski et al., 2020; Schernewski et al., 2021). Modeling studies in the Baltic Sea (Martyanov et al., 2021; Frishfelds et al., 2022; Murawski et al., 2022), North Sea (Cuttat, 2018), and Mediterranean Sea (Tsiaras et al., 2021) have incorporated biofouling of MP particles that is important to simulate their fate in the marine environment accurately (Murawski et al., 2022).

The objective of this study is to provide an overview of the pathways and accumulation areas of MP in the GoF using a multi-year high-resolution model simulation and realistic loads from the rivers and WWTPs. We chose Lagrangian particle tracking model approach, describing MP as Super-Individuals (SI; (Scheffer et al., 1995) to improve computational efficiency, with each SI representing a group of particles. We have also conducted a series of sensitivity experiments aimed at gaining a deeper insight into the impact of various processes, such as mixing, beaching, resuspension, and biofouling (Siht et al., 2025, in press). In the present study, a 3-year model simulation was conducted to identify potential MP accumulation patterns in the surface layer, water column and sediments.

The paper is organized as follows: it begins with a description of the hydrodynamic model, biogeochemistry model, and Lagrangian particle tracking model, along with MP input data sets. It is followed by an analysis of the model results, aiming to uncover the MP pathways and accumulation areas in the GoF. Finally, the results are discussed, and conclusions are derived.

In this study, results from a multi-year (2018-2020) high-resolution model experiment were employed. This model system incorporated hydrodynamic, biogeochemical and Lagrangian particle tracking models to identify the MP pollution’s pathways and accumulation zones within the GoF in the surface layer, water column, coastline and bottom layer. Our research utilized the existing datasets for MP distribution estimates and employed advanced modeling techniques.

We considered two distinct particle types: (1) PET particles with density greater than sea water and (2) PP/PE particles with density less than sea water. In addition, particles within a size range of 20 to 500 um were considered and further categorized into small (20 - 200) and large (200 - 500) particles. Generally, this is well supported by literature. PP, PE, PET, PVC, and PS are the most common polymers found worldwide (Vermeiren et al., 2016; Geyer et al., 2017; Kooi and Koelmans, 2019). In China, Lv et al. (2019) found the following polymer shares: PP (15%), PE (18%), PET (47%) and PS (20%) in the raw wastewater. In the Mediterranean, out of the total MP observed, Pedrotti et al. (2016) reported that around 86-97% share accounted for the following polymers: PP, PE and polyamides. In our study, we used PP, PE and PET due to their abundance in the environment and as they cover a wide range of densities from 900 to 1300 kg/m³. We used the previous load estimates from Schernewski et al. (2020), which included WWTP locations and emissions exclusively used within the GoF catchment area. Based on the available literature (Schrank et al., 2022; Constant et al., 2020), we have considered mean MP particles found in the rivers. Nevertheless, substantial uncertainties still exist regarding emissions from the WWTPs and MP presence in rivers.

The Baltic Sea receives land-based MPs from rivers and coastal catchment areas (Murawski et al., 2022). The GoF is under considerable anthropogenic pressure, and as a result, the levels of pollutants, including the MP in water and biota are higher compared to the neighboring basins such as the Gulf of Riga or Baltic Proper (Mishra et al., 2022; Kuprijanov et al., 2024). In addition, our results indicate that most of the MP particles entering the GoF do not spread to the Baltic Proper, but instead accumulate within the GoF.

Due to prevailing cyclonic surface circulation in the GoF, floating litter tends, in general, to drift towards the Baltic Proper in higher abundances along the northern coast (Pärn et al., 2023). Our study reveals that the easternmost part of the GoF exhibits the highest levels of MP pollution in the surface layer. Similar tendencies for the GoF have been shown by other modelling studies (e.g. Murawski et al., 2022; Pärn et al., 2023) due to the large freshwater input from the Neva River. Pollution levels were notably reduced in the central gulf, with concentrations at least an order of magnitude smaller. Rivers discharge freshwater and substances in amounts not typically found in seawater (e.g. Hetland and Hsu, 2013) and, as a result, river plumes, small or large, with high concentrations of tracers and plume fronts in the vicinity of sources are formed. Five river plume fronts were noticeable in the study area associated with the Neva River, Luga River, Narva River and Kymi River (Suursaar et al., 2021). Nevertheless, none of these fronts are stationary as they undergo spatial excursions and other dynamic transformations (Suursaar et al., 2021), and consequently, we could not detect regions with persistently high MP concentrations at these frontal regions. However, such convergent density fronts are characterized by relatively large vertical velocities (D’Asaro et al., 2020) that likely restrict MP transport over long horizontal distances. Additionally, quasi-persistent eddy activity may contribute to such a high number of plastics near the emission areas (Andrejev et al., 2004; Pärn et al., 2023).

Our simulations revealed that both the PET and PP/PE particles predominantly accumulated in the seabed in close proximity to coastal regions and emission points. The accumulation areas of lighter PP/PE particles extended further offshore. Due to their negative buoyancy, PET particles consistently descend into the water column, ultimately settling on the seabed, particularly near coastal areas. Recent research, using an Eulerian modeling approach, also reported the accumulation of PET particles in shallow coastal waters of the Baltic Sea (Schernewski et al., 2020). In contrast, PP/PE particles, initially buoyant, remained suspended longer before sinking. This coastal accumulation was particularly evident near riverine sources such as the Neva River estuary and the Narva Bay. The shallow nature of the GoF, with an average depth of 37 m (Leppäranta and Myrberg, 2009), contributes to higher deposition rates in these areas. Kuprijanov et al. (2021) reported similar patterns of hazardous substance accumulation in shallow areas, such as Neva Bay and Finnish coastal inlets, though MP accumulation may persist longer in deeper areas.

Resuspension events, driven by near-bottom currents > 30 cm s^-1, occasionally > 50 cm s^-1 (Liblik et al., 2013; Rasmus et al., 2015; Suhhova et al., 2018), or wave-induced shear stress (Jönsson et al., 2005), can remobilize MPs from sediments back into the water column gulf. This dynamic interplay of deposition and resuspension highlights the importance of hydrodynamic forces in redistributing MPs in the current study.

The beaching of particles occurred almost throughout the entire GoF. The hotspots were mostly in bays with limited access to the open sea, e.g. Neva Bay, Tallinn Bay, and multiple locations on the northern coast. Previous studies have highlighted similar trends, with higher particle accumulation near the major rivers and urban areas, such as St. Petersburg and the Gulf of Riga (Schernewski et al., 2021). Our results show that light particles had substantially higher beaching rates compared to heavy particles, likely due to differences in their buoyancy and settling dynamics. Likely, a lot of the heavy PET particles settled in the shallow areas quicker than in 10 days required for a particle to be considered beached as defined in the simulations in the present study. Therefore, a significantly smaller amount of PET was found on beaches compared to light particles. These findings underline the importance of focused monitoring and sampling near key pollution sources, such as the Bay of Tallinn, to validate the simulation results and improve understanding of beaching dynamics.

The model effectively replicates the seasonal variability of MP concentrations observed in the GoF. Spring and summer are periods of heightened biological activity in the Baltic Sea (e.g. Lips et al., 2014). MP concentrations in the surface layer are lower in summer than winter were consistent with observations in the northern Baltic Proper and the GoF (Mishra et al., 2022). Furthermore, a recent modeling study conducted in Neva Bay, which used a different modeling approach compared to our study, reported a 10-fold decrease in the surface concentration during the summer months compared to the winter (Martyanov et al., 2021). This seasonal pattern reflects the combined influence of biofouling, hydrodynamics and stratification. In summer, phytoplankton blooms promote biofouling, enhancing the sinking rates of MPs and further leading to their accumulation on the seabed. Vertical stratification, while necessary for biofouling, also acts as a barrier, potentially trapping MPs withing the thermocline (Uurasjärvi et al., 2021). By late summer, the decay of thermocline and reduced primary production lead to less biofilm formation and reduced organic matter (Almroth-Rosell et al., 2011; Liblik and Lips, 2011) which may slow MP sinking.

In contrast, during winter, reduced biofouling, weaker stratification and stronger hydrodynamic activity contribute to higher MP concentrations at the surface. Stronger near-bottom currents can also resuspend particles from the seabed, maintaining elevated MP levels in the water column.

A wind-induced coupled coastal upwelling-downwelling event, which is frequent in the GoF (e.g. Lips et al., 2009; Uiboupin and Laanemets, 2009; Laanemets et al., 2011; Väli et al., 2011; Liblik and Lips, 2017) was selected as a case study. Previous studies have proposed that coastal upwelling may result in relatively low concentrations of MPs in coastal waters (de Lucia et al., 2014; Desforges et al., 2014; Mishra et al., 2022) while high concentration patches can form in the convergence zone of coastal downwelling (Mishra et al., 2022). In contrast, La Daana et al. (2017) found no significant difference in MP concentrations between Benguela upwelling sites and other non-upwelling sites.

During the upwelling events, the distribution in the surface layers indicated large particle concentrations converged in the sub-mesoscale stripes or small eddy-like features in the upwelling frontal zone, where the Rossby numbers (not shown) and temperature gradients were high ( Figure 10 ). Previously, Väli et al. (2017, 2018) showed the convergence of particles in the GoF at the locations of the high Rossby number and sub-mesoscale activity. A recent modelling study (Väli et al., 2024) indicated the frequent occurrence of sub-mesoscale activity in the GoF. Observations of the temporal changes of MP concentrations are very challenging in such estuarine systems, where, on the one hand, the MP input from land is high, and, on the other hand, strong thermohaline gradients exist.

Previous studies have shown that episodic events, such as storms or wind-driven upwelling, can generate bottom currents, which can disturb bottom sediments, resuspending MPs into the overlying water column (Osinski et al., 2020; Zhou et al., 2021). In areas with high sedimentation rates, where MPs accumulate on the seabed, even short-term increases in current velocity have been shown to result in the release of particles back into the water column (Kane and Clare, 2019). Our results mirror this process, where we observed a marked decrease in sedimented MP concentrations and an increase in water column particles following periods of intensified bottom currents.

We applied a combination of hydrodynamic, biogeochemical, and Lagrangian particle tracking models to trace MP pathways and identify potential accumulation zones in the GOF. This three-year simulation, with sub-mesoscale permitting horizontal resolution, revealed MP concentrations at the surface, within the water column, and accumulation on the seabed. The results provide critical insights into MP distribution pathways, spatial heterogeneity and the influence of hydrodynamics on MP transport and accumulation.

The study revealed that approximately 75% of MP particles settle on the seabed, making it the primary accumulation area in the GoF. Around 10% of particles were beached, with notable accumulation in Neva Bay, Narva Bay, and parts of the Finnish coastline. Only 1% of MP particles exited the gulf through the western boundary, while 14% remained suspended in the water column, influenced by episodic resuspension events. These findings highlight the GoF’s role as a significant retention zone for MPs due to its semi-enclosed geography and hydrodynamic conditions.

MP concentrations were highest near major coastal and riverine sources, particularly in the eastern part of the gulf. Surface concentrations of light particles exceeded those of heavy particles, especially in proximity to WWTPs and river mouths. The zonally integrated concentrations demonstrated higher values in the shallower eastern areas and a marked decrease west of 28°E. This spatial variability underscores the influence of anthropogenic inputs and local hydrodynamics on MP distribution.

Hydrodynamic processes played a key role in shaping MP transport and redistribution in the GoF. Upwelling and downwelling events, as well as episodic intensification of bottom currents, significantly influenced the redistribution of MPs. Strong bottom currents resuspended settled MPs into the water column, temporarily increasing their concentrations. These dynamic processes emphasize the importance of monitoring MPs across all layers of the marine environment – not just the surface layer but also the water column and seabed, to fully understand their transport and fate.

This study provides a scientific basis for policymakers to regulate MP emissions from WWTPs, manage riverine inputs, and address urban coastal pollution. By identifying hotspot areas, such as Neva Bay and Narva Bay, the findings can help prioritize resources for pollution control. Future monitoring efforts in the GoF and Baltic Sea should extend beyond the surface layer to include the water column and seabed, incorporating multipoint sampling at various depths, especially near emission sources. Additionally, mitigation measures should focus on the eastern gulf, where MP concentrations are consistently elevated.