Researchers are increasingly concerned about microplastics (MPs) since they have proven to be detrimental to the environment and humans. They not only contribute to the accumulation of plastics in the environment, but they can also help spread micropollutants through absorption. MPs have entered the environment through a variety of channels, including direct discharge through human activities, the textile industry, and wastewater treatment systems. Wastewater treatment plants (WWTPs) are seen as the final line of defense, forming a barrier between MPs and the environment. The most common MPs found in wastewater are polyethylene terephthalate (PET), polyethylene (PE), polypropylene (PP), and polystyrene. The purpose of this study is to investigate and measure the efficacy of conventional and advanced technologies in removing microplastic. The fate of MPs in WWTPs and their impact on various wastewater treatment processes will be thoroughly examined. Both conventional and advanced methods of treatment will be discussed. Conventional methods include coagulation/flocculation, sedimentation, and filtration. Advanced treatment methods that will be discussed include membrane bioreactors, rapid sand filtration, electrocoagulation, and photocatalytic degradation. Despite the high removal of MPs, WWTPs remain an important entry point into aquatic and terrestrial systems due to the large volumes involved and the quantity of sludge reused through land application. Also, the complex interactions between MPs and other environmental pollutants may broaden MPs' effects on wastewater treatment processes, which have yet to be fully researched. Furthermore, potential issues with MP removal from treatment procedures and microplastic remediation techniques in drinking water was discussed.
Per- and polyfluoroalkyl substances (PFAS) are arguably the most common water contaminants in the world today. While several research experiments have been done to understand and remove PFAS from the environment, there is still a lot of unknown. Little has been known about the use of Machine learning (ML) to understand PFAS. This work hence reviews some leading ML approaches and applications in PFAS studies in the distribution, transport, removal, and occurrence predictions of PFAS. Several evaluation matrices were examined and used to perform this function. There are still a lot of areas whereby ML can be used to improve our PFAS knowledge base, some of these were briefly stated in this review.
Anthropogenic activities such as the damming of rivers have resulted in the formation of depositional systems for sediments that are naturally transported along river channels all over the world. The steady accumulation of sediments has profound effects on how reservoirs function. As the reservoir ages, the impact of sedimentation becomes more pronounced. In an alluvial channel, the dynamics of sediment movement, scour, and deposition are exceedingly complicated. The quantitative and qualitative management of river engineering heavily relies on sediment motion. One of the key concerns affecting reservoir storage capacity is reservoir sedimentation. Understanding the form and behavior of the river or reservoir is necessary for a scientific approach to sedimentation. An overview of reservoir sedimentation processes, and models of reservoir and shallow water bodies is provided in this article. More research in the developing fields is advised for continuing studies on reservoir sedimentation due to climate change effects
Data-driven streamflow prediction models have achieved high predictive accuracy, but the relative contributions of meteorological forcing, hydrologic memory, and spatial watershed connectivity remain poorly understood, particularly in cold-region watersheds. This study proposes a six-tier diagnostic framework to quantify the contributions of meteorological forcing, storage memory, and network connectivity to streamflow prediction in the Goose River watershed. Six progressively more complex modelling tiers were developed, ranging from temporal deep learning and static machine-learning models to graph neural networks (GNNs) and spatio-temporal GNNs. Meteorological forcing dominated short-term discharge prediction, whereas static physiographic attributes alone showed limited predictive capability. Incorporating antecedent discharge memory produced the highest numerical performance (NSE = 0.989), while the spatio-temporal GNN achieved the most physically consistent representation by integrating dynamic forcing with upstream–downstream connectivity. These findings demonstrate a trade-off between statistical accuracy and hydrologic interpretability, showing that graph-based models provide greater physical realism and transferability despite slightly lower predictive accuracy than autoregressive approaches.
Reliable streamflow estimation at data-limited locations requires hydrologic relationships learned from gauged rivers to remain transferable beyond the river systems used for model development. This study tested whole-river transferability of upstream–downstream streamflow relationships across 36 gauges in eight tributary systems of the Red River of the North. Drainage-area scaling and statistical streamflow lag characterized spatial and temporal correspondence along 27 direct links, while local models established within-river predictability. Transferability was evaluated using leave-one-river-out validation, with each river system excluded from model training and predicted using relationships learned from the remaining rivers. Streamflow scaled approximately proportionally with drainage area (mean β = 1.017; median = 0.949), although substantial link-level variation occurred. Twenty-three of 27 links exhibited optimal lags of 0–1 day. Across 20 regional-transfer links, regional XGBoost achieved a mean NSE of 0.692 compared with 0.836 for matched local models, with a median generalization gap of 0.104. None of 13 topographic, land-cover, and soil contrasts were significantly associated with transfer loss, and physical augmentation did not consistently improve prediction. Overall, upstream–downstream relationships retained substantial predictive information when transferred to withheld river systems, but transfer losses were heterogeneous and not consistently explained by static basin contrasts. Whole-river withholding therefore provides a stringent test of regional streamflow transferability in partially gauged cold-region river networks.