Regional patterns and drivers of total nitrogen trends in the Chesapeake Bay watershed: Insights from machine learning approaches and management implications

Spatiotemporal distribution patterns and coupling effects of aquatic environmental factors in the dry-wet season over a decade from the Beibu Gulf, South China Sea

Since the 21st century, the Beibu Gulf area has been affected by increasing anthropogenic activities, which makes the coastal aquatic ecosystem extremely concerning. However, the comprehensive exploration and analysis of the long-term scale behavior change characteristics of various water quality environment factors is still limited. Through comprehensively detecting coastal surface water environmental behavior information from 33 locations in the Beibu Gulf from 2005 to 2015, we revealed and quantified mutual response characteristics and patterns of various environmental indicators. The main environmental pollution indicators (e.g., COD, NH4+, NO3-, and DIP) showed a gradual decrease in concentration from the coast to the offshore sea area, and significantly increases during the wet season. The semi-enclosed Maowei Sea exhibited the most prominent performance with significant differences compared to other regions in Beibu Gulf. The average Chlorophyll-a (Chla) content in the coastal area of the Beibu Gulf during the wet season was more than twice that of the dry season, yet the interaction pattern between Chla and environmental factors in the two seasons was opposite to its concentration behavior, accompanied by a closely significant relationship with thermohaline structure and the input of nitrogen and phosphorous nutrients. The multivariate statistical analysis results of total nutrient dynamics suggested that the Beibu Gulf was clearly divided into different regions in both dry and wet season clusters. The present study can provide a comprehensive perspective for the spatial and temporal migration patterns and transformation laws of coastal water environmental factor, which should contribute to improve the prevention countermeasure of nutrient pollution in coastal environment.

Development of a basin-scale total nitrogen prediction model by integrating clustering and regression methods

Nutrient runoff into rivers caused by human activity has led to global eutrophication issues. The Nakdong River in South Korea is currently facing significant challenges related to eutrophication and harmful algal blooms, underscoring the critical importance of managing total nitrogen (T-N) levels. However, traditional methods of indoor analysis, which depend on sampling, are labor-intensive and face limitations in collecting high-frequency data. Despite advancements in sensor allowing for the measurement of various parameters, sensors still cannot directly measure T-N, necessitating surrogate regression methods. Therefore, we conducted T-N predictions using a water quality dataset collected from 2018 to 2022 at 157 observatories within the Nakdong River basin. To account for the water quality characteristics of each location, we employed a clustering technique to divide the basin and compared a Gaussian mixture model with K-means clustering. Moreover, optimal regressor for each cluster was selected by comparing multiple linear regression (MLR), random forest, and XGBoost. The results showed that forming four clusters via K-means clustering was the most suitable approach and MLR was reasonably accurate for all clusters. Subsequently, recursive feature elimination cross-validation was used to identify suitable parameters for T-N prediction, thus leading to the construction of high-accuracy T-N prediction models. Clustering was useful not only for improving the regressors but also for spatially analyzing the water quality characteristics of the Nakdong River. The MLR model can reveal causal relationships and thus is useful for decision-making. The results of this study revealed that the combination of a simple linear regression model and clustering method can be applied to a wide watershed. The clustering-based regression model showed potential for accurately predicting T-N at the basin level and is expected to contribute to nationwide water quality management through future applications in various fields.

FlowSOM clustering - A novel pattern recognition approach for water research: Application to a hyper-arid coastal aquifer system

Monitoring and understanding of water resources have become essential in designing effective and sustainable management strategies to overcome the growing water quality challenges. In this context, the utilization of unsupervised learning techniques for evaluating environmental tracers has facilitated the exploration of sources and dynamics of groundwater systems through pattern recognition. However, conventional techniques may overlook spatial and temporal non-linearities present in water research data. This paper introduces the adaptation of FlowSOM, a pioneering approach that combines self-organizing maps (SOM) and minimal spanning trees (MST), with the fast-greedy network clustering algorithm to unravel intricate relationships within multivariate water quality datasets. By capturing connections within the data, this ensemble tool enhances clustering and pattern recognition. Applied to the complex water quality context of the hyper-arid transboundary Caplina/Concordia coastal aquifer system (Peru/Chile), the FlowSOM network and clustering yielded compelling results in pattern recognition of the aquifer salinization. Analyzing 143 groundwater samples across eight variables, including major ions, the approach supports the identification of distinct clusters and connections between them. Three primary sources of salinization were identified: river percolation, slow lateral aquitard recharge, and seawater intrusion. The analysis demonstrated the superiority of FlowSOM clustering over traditional techniques in the case study, producing clusters that align more closely with the actual hydrogeochemical pattern. The outcomes broaden the utilization of multivariate analysis in water research, presenting a comprehensive approach to support the understanding of groundwater systems.

Optimizing the development of contaminated land in China: Exploring machine-learning to identify risk markers

Often available for use, previously developed land, which includes residential and commercial/industrial areas, presents a significant challenge due to the risk to human health. China's 2018 release of health risk assessment standards for land reuse aimed to bridge this gap in soil quality standards. Despite this, the absence of representative indicators strains risk managers economically and operationally. We improved China's land redevelopment approach by leveraging a dataset of 297,275 soil samples from 352 contaminated sites, employing machine learning. Our method incorporating soil quality standards from seven countries to discern patterns for establishing a cost-effective evaluative framework. Our research findings demonstrated that detection costs could be curtailed by 60% while maintaining consistency with international soil standards (prediction accuracy = 90-98%). Our findings deepen insights into soil pollution, proposing a more efficient risk assessment system for land redevelopment, addressing the current dearth of expertise in evaluating land development in China.

Our national nutrient reduction needs: Applying a conservation prioritization framework to US agricultural lands

Targeted conservation approaches seek to focus resources on areas where they can deliver the greatest benefits and are recognized as key to reducing nonpoint source nutrients from agricultural landscapes into sensitive receiving waters. Moreover, there is growing recognition of the importance and complementarity of in-field and edge-of-field conservation for reaching nutrient reduction goals. Here we provide a generic prioritization that can help with spatial targeting and applied it across the conterminous US (CONUS). The prioritization begins with identifying areas with high agricultural nutrient surplus, i.e., where the most nitrogen (N) and/or phosphorus (P) inputs are left on the landscape after crop harvest. Subwatersheds with high surplus included 52% and 50% of CONUS subwatersheds for N and P, respectively, and were located predominantly in the Midwest for N, in the South for P, and in California for both N and P. Then we identified the most suitable conservation strategies using a hierarchy of metrics including nutrient use efficiency (proportion of new nutrient inputs removed by crop harvest), tile drainage, existing buffers for agricultural run-off, and wetland restoration potential. In-field nutrient input reduction emerged as a priority because nutrient use efficiency fell below a high but achievable goal of 0.7 (30% of nutrients applied are not utilized) in 45% and 44% of CONUS subwatersheds for N and P, respectively. In many parts of the southern and western US, in-field conservation (i.e., reducing inputs + preventing nutrients from leaving fields) alone was likely the optimal strategy as agriculture was already well-buffered. However, stacking in-field conservation with additional edge-of-field buffering would be important to conservation strategies in 35% and 29% of CONUS subwatersheds for N and P, respectively. Nutrient use efficiencies were often high enough in the Midwest that proposed strategies focused more on preventing nutrients from leaving fields, managing tile effluent, and buffering agricultural fields. Almost all major river basins would benefit from a variety of nutrient reduction conservation strategies, underscoring the potential of targeted approaches to help limit excess nutrients in surface and ground waters.

The changing nitrogen landscape of United States streams: Declining deposition and increasing organic nitrogen

Air quality regulations have led to decreased nitrogen (N) and sulfur deposition across the conterminous United States (CONUS) during the last several decades, particularly in the eastern parts. But it is unclear if declining deposition has altered stream N at large scales. We compared watershed N inputs with N chemistry from over 2,000 CONUS streams where deposition was the largest N input to the watershed. Weighted change analysis showed that deposition declined across most watersheds, especially in the Eastern CONUS. Nationally, declining N deposition was not associated with significant large-scale declines in stream nitrate concentration. Instead, significant increases in stream dissolved organic carbon (DOC) and total organic N (TON) were widespread across regions. Possible mechanisms behind these increases include declines in acidity and/or ionic strength drivers, changes in carbon availability, and/or climate variables. Our results also reveal a declining trend of DOC/TON ratio over the entire study period, primarily influenced by the trend in the Eastern region, suggesting the rate of increase in stream TON exceeded the rate of increase in DOC concentration during this period. Our results illustrate the complexity of nutrient cycling that links long-term atmospheric deposition to water quality. More research is needed to understand how increased dissolved organic N could affect aquatic ecosystems and downstream riverine nutrient export.

Nutrient transport following water transfer through the world's largest water diversion channel

Nutrient levels in the artificial channel constructed for the Middle Route Project are significant indicators of water quality safety and aquatic ecological integrity for this large, inter-basin scheme. However, the distribution and transport of nutrients along the channel were poorly understood. Based on a time-series dataset as well as mass balance and material flow analysis methods, the water and nutrient transport fluxes in the Middle Route of the South-to-North Water Diversion Project were identified in this study. The results indicate that the nutrient concentrations varied considerably with time, but there was no significant difference among the 30 stations of the main channel. Seasonal temperature difference was the major factor in the large fluctuations of water quality indicators over time. The nutrient loadings varied with the water volume outputs from the main channel to the water-receiving cities. Atmospheric deposition was an important source of nutrients in the main channel, accounting for 9.13%, 20.6%, and 0.635% of the nitrogen, phosphorus, and sulfur input from the Danjiangkou Reservoir, respectively. In 2021, a net accumulation of 988 tons of N, 29 tons of P, and 2,540 tons of S, respectively, were present in the main channel. The increase of these external and internal nutrient loadings would cause water quality fluctuation and deterioration in some local sections of the main channel. Our study quantified the spatial and temporal patterns of nutrient transport in the Middle Route and revealed the ecological effects on the aquatic environment, assisting authorities on the project to develop effective water conservation strategies.

Application of classification machine learning algorithms for characterizing nutrient transport in a clay plain agricultural watershed

Excess nutrients in surface water and groundwater can lead to water quality deterioration in available water resources. Thus, the classification of nutrient concentrations in water resources has gained significant attention during recent decades. Machine learning (ML) algorithms are considered an efficient tool to describe nutrient loss from agricultural land to surface water and groundwater. Previous studies have applied regression and classification ML algorithms to predict nutrient concentrations in surface water and/or groundwater, or to categorize an output variable using a limited number of input variables. However, there have been no studies that examined the application of different ML classification algorithms in agricultural settings to classify various output variables using a wide range of input variables. In this study, twenty-four ML classification algorithms were implemented on a dataset from three locations within the Upper Parkhill watershed, an agricultural watershed in southern Ontario, Canada. Nutrient concentrations in surface water were classified using geochemical and physical water parameters of surface water and groundwater (e.g., pH), climate and field conditions as the input variables. The performance of these algorithms was evaluated using four evaluation metrics (e.g., classification accuracy) to identify the optimal algorithm for classifying the output variables. Ensemble bagged trees was found to be the optimal ML algorithm for classifying nitrate concentration in surface water (accuracy of 90.9%), while the weighted KNN was the most appropriate algorithm for categorizing the total phosphorus concentration (accuracy of 87%). The ensemble subspace discriminant algorithm gave the highest overall classification accuracy for the concentration of soluble reactive phosphorus and total dissolved phosphorus in surface water with an accuracy of 79.2% and 77.9%, respectively. This study exemplifies that ML algorithms can be used to signify exceedance of recommended concentrations of nutrients in surface waters in agricultural watersheds. Results are useful for decision makers to develop nutrient management strategies.

Shifts in the composition of nitrogen deposition in the conterminous United States are discernable in stream chemistry

Across the conterminous United States (U.S.), the composition of atmospheric nitrogen (N) deposition is changing spatially and temporally. Previously, deposition was dominated by oxidized N, but now reduced N (ammonia [NH3] + ammonium [NH4+]) is equivalent to oxidized N when deposition is averaged across the entire nation and, in some areas, reduced N dominates deposition. To evaluate if there are effects of this change on stream chemistry at the national scale, estimates of N deposition form (oxidized or reduced) from the National Atmospheric Deposition Program Total Deposition data were coupled with stream measurements from the U.S. Environmental Protection Agency (EPA) National Rivers and Streams Assessments (three stream surveys between 2000 and 2014). A recent fine-scaled N input inventory was used to identify watersheds (<1000 km2) where atmospheric deposition is the largest N source (n = 1906). Within these more atmospherically-influenced watersheds there was a clear temporal shift from a greater proportion of sites dominated by oxidized N deposition to a greater proportion of sites dominated by reduced forms of N deposition. We found a significant positive correlation between oxidized N deposition and stream NO3- concentrations, whereas the correlation between reduced N deposition and stream NO3- concentrations were significant but weaker. Sites dominated by atmospheric inputs of reduced N forms had higher stream total organic N and total N despite lower total N deposition on average. This higher stream concentration of total N is mainly driven by the higher concentration of total organic N, suggesting an interaction between elevated reduced N in deposition and living components of the ecosystem or soil organic matter dynamics. Regardless of the proportion of reduced to oxidized N forms in deposition, stream NH4+ concentrations were generally low, suggesting that N deposited in a reduced form is rapidly immobilized, nitrified and/or assimilated by watershed processes.

Nutrient limitation of phytoplankton in three tributaries of Chesapeake Bay: Detecting responses following nutrient reductions

Many coastal ecosystems suffer from eutrophication, algal blooms, and dead zones due to excessive anthropogenic inputs of nitrogen (N) and phosphorus (P). This has led to regional restoration efforts that focus on managing watershed loads of N and P. In Chesapeake Bay, the largest estuary in the United States, dual nutrient reductions of N and P have been pursued since the 1980s. However, it remains unclear whether nutrient limitation - an indicator of restriction of algal growth by supplies of N and P - has changed in the tributaries of Chesapeake Bay following decades of reduction efforts. Toward that end, we analyzed historical data from nutrient-addition bioassay experiments and data from the Chesapeake Bay long-term water-quality monitoring program for six stations in three tidal tributaries (i.e., Patuxent, Potomac, and Choptank Rivers). Classification and regression tree (CART) models were developed using concurrent collections of water-quality parameters for each bioassay monitoring location during 1990-2003, which satisfactorily predicted the bioassay-based measures of nutrient limitation (classification accuracy = 96%). Predictions from the CART models using water-quality monitoring data showed enhanced nutrient limitation over the period of 1985-2020 at four of the six stations, including the downstream station in each of these three tributaries. These results indicate detectable, long-term water-quality improvements in the tidal tributaries. Overall, this research provides a new analytical tool for detecting signs of ecosystem recovery following nutrient reductions. More broadly, the approach can be adapted to other waterbodies with long-term bioassays and water-quality data sets to detect ecosystem recovery.