Pollution in Rivers

​

Research has extensively documented widespread pollution in U.S. rivers, with 56% of assessed river and stream miles exhibiting poor conditions and persistent contamination (U.S. Environmental Protection Agency, 2019). Industrial discharges, agricultural runoff, wastewater effluents, and atmospheric deposition have been documented as leading contributors to water pollution (Singh et al., 2024). Water contamination sources are categorized into point or nonpoint sources. Point source pollution originates from identifiable locations from which pollutants are discharged, such as from industrial facilities, agricultural activities, and wastewater treatment plants. In this source, pollutants enter straight into the water bodies undiluted and unbuffered, creating hotspots of contamination that often exceed safe thresholds near the discharge point. On the other hand, non-point source pollution results from diffuse sources, including agricultural runoff and atmospheric deposition (National Oceanic and Atmospheric Administration, 2025). Stormwater runoff from nonpoint source pollution is one of the most significant threats to aquatic ecosystems in the United States, as it is linked to the formation of large dead zones with minimal oxygen. 

​

National monitoring records consistently attribute widespread river pollution to three main pollutant classes: nutrients, heavy metals, and organic pollutants. Excessive nutrients, such as nitrogen and phosphorus, lead to eutrophication in bodies of water, depleting oxygen levels and leading to dead zones (Wurtsbaugh, 2019). Furthermore, Ahamad et al. highlight how heavy metals, including mercury, lead, and arsenic, have become increasingly important due to their non-degradable nature, toxicity even at low concentrations, and tendency to bioaccumulate in aquatic organisms (2019). Moreover, previous research has considered how multiple organic compounds, such as pesticides and PCBs, when present in water, are toxic, carcinogenic, and mutagenic, causing several diseases in humans (Agarwal, 2013). 

​

Polychlorinated Biphenyls (PCBs)

​

Among organic contaminants, Polychlorinated Biphenyls (PCBs) stand out for their durability in sediments (U.S. Environmental Protection Agency, 2019). These synthetic organochlorines were widely manufactured in the mid-20th century for electrical equipment, paints, and sealants, until the EPA enforced disposal regulations in 1979 in the United States due to toxicity and persistence in the environment (U.S. Environmental Protection Agency, 2025). In rivers, PCBs are strongly hydrophobic and poorly soluble, so they settle into fine-grained sediments, where they can appear to stay inactive, but high flow events periodically remobilize the contaminated sediment, again posing risks to ecosystems (Cucchiara, 2023). Since they resist degradation, PCBs remain available for consumption, bioaccumulating and biomagnifying through aquatic food webs. Hence, their high hydrophobicity, low water solubility, and resistance to degradation result in multiple decades of environmental resistance and exposure potential (U.S. Environmental Protection Agency, 2025).

​

Extensive studies by the EPA have revealed that PCBs can affect an animal’s immune, reproductive, nervous, and endocrine systems, resulting in a multitude of cancer and serious non-cancer health effects (U.S. Environmental Protection Agency, 2025). Additionally, studies in humans demonstrate that PCBs are carcinogenic and are linked to other health effects, such as low birth weight, thyroid disease, and learning, memory, and immune system disorders (U.S. Environmental Protection Agency, 2025). Therefore, their properties and effects make PCBs a priority contaminant in rivers and a central focus for remediation evaluation.

​

The Hudson River and Dredging

​

Regulatory recognition of PCB contamination in the Hudson River preceded formal remedial decisions by nearly a decade, as the New York State Department of Environmental Conservation’s 1978 Summary of Hudson River PCB Study Results states that the PCB contamination in the Hudson River had been identified as early as 1969 (Cucchiara, 2023). In 1975, General Electric’s facilities were identified as the largest contributor of PCBs to the river, after the EPA had issued GE a permit to continue discharging the chemical in the river only a few years prior, in 1972. Furthermore, after initial releases, PCBs were transported downstream and accumulated in sediments behind the Fort Edward dam. After the deteriorating dam was removed in 1973, subsequent floods resulted in further downstream distribution of PCBs, considerably complicating cleanup efforts (New York State Department of Environmental Conservation, 2018).

​

With the rise of environmental movement and activism in the 1960s, the Hudson River became a very important focus as the effects of the contamination began to be better understood (Cucchiara, 2023). In 1984, 200 miles of the Hudson were placed on the EPA’s National Priorities List, formalizing the system as a federal Superfund site. Extensive federal and state monitoring over subsequent decades showed that natural recovery alone was insufficient, urging the selection of an active sediment-removal remedy.

​

Consequently, industries started to be held accountable for their actions regarding PCB pollution, and environmental considerations began to influence industrial planning. However, plans to clean up the river in the late 1970s were faced with endless regulatory disputes until the Environmental Protection Agency established that General Electric (GE) needed to be held responsible (Cucchiara, 2023). As a result, after a lengthy battle, in 2002, GE was ordered to clean up a 40-mile stretch of the river through the Hudson River PCB Superfund Dredging Project. The first phase of the dredging project occurred in 2009, while the second phase of dredging began in 2011 and ended in the fall of 2015. Over the six seasons of dredging, approximately 2.75 million cubic yards of contaminated sediment containing about 140,614 kg of PCBs were removed from the river bottom (New York State Department of Environmental Conservation, 2018). 

​

Prior to the implementation of dredging, policymakers debated whether the expected effectiveness of sediment removal justified the short-term risks. The EPA endorsed dredging as the method of choice for the project because it involves the direct removal of contaminated sediments. However, opponents argued that it would not be worth the cost, as 65 to 70 percent of the PCBs would remain (Cucchiara, 2023). It was also argued that dredging would stir up the sediments, which would worsen the contamination. GE maintained this perspective, arguing that “PCBs in the Hudson are down 90%” and that “dredging is careless” (Cucchiara, 2023).

​

Existing Methods for Pollutant Load Estimation

​

Since progress from remediation is evaluated by reductions in downstream export and long-term trends, agencies and researchers have employed load estimation methods to quantify PCB transport. Load is defined as the mass of a substance that passes through a particular point of a river, such as a monitoring station, in a specified amount of time. One widely used approach for estimating pollutant loads is regression analysis, particularly when data is incomplete. Regression models, such as linear and multiple regression, require that there be a strong linear association between flow and concentration, allowing researchers to predict river contaminations (Meals et al., 2013). A study by Christensen et al. applied regression models to estimate alkalinity, chloride, sulfate, and atrazine concentrations by using real-time physical measurements as independent variables. Similarly, Richards uses sample data from the Cuyahoga and Choptank Rivers to build multiple regression models linking concentration and flow under varying conditions, helping sediment and nitrogen load estimates (Richards, n.d.). Moreover, ratio estimators offer a lower-complexity alternative when continuous flow and intermittent concentration samples are available, often with bias corrections and seasonal stratification (Meals et al., 2013). Another significant approach involves integration techniques, which rely on definite integrals to calculate the total load, offering accurate assessments for environmental monitoring (Meals et al., 2013). For post-remediation evaluation where conditions vary year to year, windowed trapezoidal integration over multi-year blocks provides stable comparisons of export relevant for decision making (Meals et al., 2013).