We can't find the internet
Attempting to reconnect
Something went wrong!
Hang in there while we get back on track
Machine learning-based detection and mapping of riverine litter utilizing Sentinel-2 imagery
AI summary Read the abstract
Researchers used satellite imagery and machine learning to detect and map litter accumulation in the Tisza River, Hungary, finding that dams are major hotspots and that flood events drive litter transport. While models performed well in controlled tests, real-world accuracy was moderate, highlighting the challenge of using satellite data to monitor riverine plastic pollution at scale. This matters because rivers are a primary pathway for plastic and microplastic debris reaching the ocean.
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.
More Papers Like This
Insights into suspended sediment and microplastic budget of a lowland river: integrating in-situ measurements, Sentinel-2 imagery, and machine learning
AI summary Read the abstract
Researchers combined river measurements, satellite imagery, and machine learning to track how much microplastic the Tisza River in Hungary carries downstream each day. They found that flood events spike microplastic transport five-fold, meaning a small number of high-water days drive the majority of plastic particles reaching the sea.
Comparative Analysis of Riverine Plastic Pollution Combining Citizen Science, Remote Sensing and Water Quality Monitoring Techniques
AI summary Read the abstract
A multi-method study combining citizen science litter surveys, remote sensing, and water quality monitoring characterized plastic pollution along the Tisza River, one of Europe's most plastic-polluted rivers, spanning five countries.
High spatiotemporal resolution analysis on suspended sediment and microplastic transport of a lowland river
AI summary Read the abstract
Researchers conducted dense spatial and time-series monitoring of microplastic and suspended sediment transport along Hungary's Tisza River, finding that microplastic concentrations closely tracked sediment levels during floods but diverged during low water. Most detected particles were fibers likely from wastewater, and the data showed that floods, tributaries, and river dams all influence how microplastics move through river systems.
Comparative Analysis of Riverine Plastic Pollution Combining Citizen Science, Remote Sensing and Water Quality Monitoring Techniques
AI summary Read the abstract
Researchers combined citizen science reports, remote sensing imagery, and water quality monitoring to compare riverine plastic pollution along the Tisza River across five countries from 2016 to 2022, finding over 3,200 litter accumulation zones. The study demonstrates the value of integrating multiple data sources for large-scale plastic pollution assessment.
A learning approach for garbage and debris identification
AI summary Read the abstract
This study developed machine learning approaches for detecting garbage and plastic debris in river and ocean environments from imagery. The authors present detection models and discuss the challenges of building training datasets for environmental plastic monitoring systems.
Research digests by email
When a large batch of papers lands in the Atlas, we read through it and send a short write-up of what stood out.