Machine learning-based detection and mapping of riverine litter utilizing Sentinel-2 imagery

Abstract

Despite the substantial impact of rivers on the global marine litter problem, riverine litter has been accorded inadequate consideration. Therefore, our objective was to detect riverine litter by utilizing middle-scale multispectral satellite images and machine learning (ML), with the Tisza River (Hungary) as a study area. The Very High Resolution (VHR) images obtained from the Google Earth database were employed to recognize some riverine litter spots (a blend of anthropogenic and natural substances). These litter spots served as the basis for training and validating five supervised machine-learning algorithms based on Sentinel-2 images [Artificial Neural Network (ANN), Support Vector Classifier (SVC), Random Forest (RF), Naïve Bays (NB) and Decision Tree (DT)]. To evaluate the generalization capability of the developed models, they were tested on larger unseen data under varying hydrological conditions and with different litter sizes. Besides the best-performing model was used to investigate the spatio-temporal variations of riverine litter in the Middel Tisza. According to the results, almost all the developed models showed favorable metrics based on the validation dataset (e.g., F1-score; SVC: 0.94, ANN: 0.93, RF: 0.91, DT: 0.90, and NB: 0.83); however, during the testing process, they showed medium (e.g., F1-score; RF:0.69, SVC: 0.62; ANN: 0.62) to poor performance (e.g., F1-score; NB: 0.48; DT: 0.45). The capability of all models to detect litter was bounded to the pixel size of the Sentinel-2 images. Based on the spatio-temporal investigation, hydraulic structures (e.g., Kisköre Dam) are the greatest litter accumulation spots. Although the highest transport rate of litter occurs during floods, the largest litter spot area upstream of the Kisköre Dam was observed at low stages in summer. This study represents a preliminary step in the automatic detection of riverine litter; therefore, additional research incorporating a larger dataset with more representative small litter spots, as well as finer spatial resolution images is necessary.

Comparative Analysis of Riverine Plastic Pollution Combining Citizen Science, Remote Sensing and Water Quality Monitoring Techniques

“…According to the annually repeated (since 2019) monitoring along the river [30][31][32][33][34], the microplastic pollution of the sediments was the highest in 2019 (3177±1970 items/kg). In the same year, the largest MaP accumulation (18.8 thousand m 2 ) was detected upstream of the Kisköre Dam [22]. Still, the microplastic fragments contribute just to 2-33%, as micro-fibers dominate in the sediments within the water system of the Tisza [30,34,35].…”

“…The MaP pollution of the Tisza and the mountainous areas of the catchment were documented by Liro et al [16], Mohsen et al [22] and Magyar et al [27]. Besides, one of the longest and most detailed microplastic surveys of the world was performed in the river system of the Tisza, revealing the MiP transport in the river and its accumulation in the sediments [30][31][32][33][34][35].…”

“…The dense canopy of floodplain forests limits the aerial surveys (e.g. UAV, remote sensing data) of the stranded litter under the canopy; besides, the resolution of the remotely sensed images is often a challenge [21,22]. Therefore, a different approach is needed.…”

“…The plastic waste could enter the fluvial system during heavy rainfalls when the run-off and mass movements mobilised the waste along the banks and valley slopes [26]. Along the Tisza River, litter accumulations (containing waste and natural items) were identified based on Sentinel-2 images and machine-learning algorithms [22,27]. Based on the results, the greatest riverine litter accumulations develop upstream of water engineering structures (e.g., dams, bridges, hydroelectric power stations), and the highest litter transport rate occurs during floods.…”

“…Based on the results, the greatest riverine litter accumulations develop upstream of water engineering structures (e.g., dams, bridges, hydroelectric power stations), and the highest litter transport rate occurs during floods. The largest litter spot area was observed at low stages upstream of the Kisköre Dam [22,27]. During floods, the transport rate of plastic litter could be as much as 500 plastic bottles per minute [6]; thus, it is called the "plastic flood" [28].…”