Global patterns and key drivers of stream nitrogen concentration: A machine learning approach

Unveiling PFAS hazard in European surface waters using an interpretable machine-learning model

Per- and polyfluoroalkyl substances (PFAS), commonly known as "forever chemicals", are ubiquitous in surface waters and potentially threaten human health and ecosystems. Despite extensive monitoring efforts, PFAS risk in European surface waters remain poorly understood, as performing PFAS analyses in all surface waters is remarkably challenging. This study developed two machine-learning models to generate the first maps depicting the concentration levels and ecological risks of PFAS in continuous surface waters across 44 European countries, at a 2-km spatial resolution. We estimated that nearly eight thousand individuals were affected by surface waters with PFAS concentrations exceeding the European Drinking Water guideline of 100 ng/L. The prediction maps identified surface waters with high ecological risk and PFAS concentration (>100 ng/L), primarily in Germany, the Netherlands, Portugal, Spain, and Finland. Furthermore, we quantified the distance to the nearest PFAS point sources as the most critical factor (14%-19%) influencing the concentrations and ecological risks of PFAS. Importantly, we determined a threshold distance (4.1-4.9 km) from PFAS point sources, below which PFAS hazards in surface waters could be elevated. Our findings advance the understanding of spatial PFAS pollution in European surface waters and provide a guideline threshold to inform targeted regulatory measures aimed at mitigating PFAS hazards.

Microbial Metabolic Limitations and Their Relationships with Sediment Organic Carbon Across Lake Salinity Gradient in Tibetan Plateau

Inland lakes, contributing substantially to the global storage of sediment organic carbon (SOC), are subject to marked changes in salinity due to climate warming. The imbalance in the supply of resources, such as carbon, nitrogen, and phosphorus, in sediments leads to microbial metabolic limitations (MMLs). This, in turn, triggers the secretion of extracellular enzymes by microorganisms to mine for deficient resources by decomposing complex organic carbon. This process is a rate-limiting step in the degradation of organic carbon and, as a result, has the potential to regulate organic carbon stocks. However, the general understanding of MML patterns and their relationships with SOC content along lake salinity gradients remains elusive. This study examined 25 lakes on the Tibetan Plateau with salinity ranging from 0.13‱ to 31.06‱, analyzing MMLs through enzymatic stoichiometry. The results showed that sediment microbial metabolism was mainly limited by carbon and nitrogen, with stronger limitations at higher salinity. Water salinity and sediment pH were the main factors influencing microbial limitations, either directly or indirectly, through their effects on nutrients and microbial diversity. Additionally, the SOC content was negatively correlated with microbial carbon limitation, a relationship weakened when salinity and pH were controlled. These findings suggest that the decrease in SOC with increased salinity or pH could be driven by stronger microbial carbon limitations, offering insights into the impact of salinity changes on SOC stocks in inland lakes due to climate change.

Enhancing prediction and inference of daily in-stream nutrient and sediment concentrations using an extreme gradient boosting based water quality estimation tool - XGBest

Estimating constituent loads in streams and rivers is a crucial but challenging task due to low-frequency sampling in most watersheds. While predictive modeling can augment sparsely sampled water quality data, it can be challenging due to the complex and multifaceted interactions between several sub-watershed eco-hydrological processes. Traditional water quality prediction models, typically calibrated for individual sites, struggle to fully capture these interactions. This study introduces XGBest, a machine learning-based tool, that integrates hydrological data, land cover, and physical watershed attributes at a regional scale to predict daily concentrations of Total Nitrogen (TN), Total Phosphorus (TP), and Total Suspended Solids (TSS). XGBest leverages 29 environmental variables, including daily and antecedent discharge, temporal features, and landscape characteristics, to comprehensively evaluate water quality dynamics across a large hydrologic region. To explore the robustness of the developed tool, XGBest was validated using observed water quality data in three different hydrologic regions in the eastern United States, encompassing 499 water quality monitoring sites characterized by diverse hydro-climatic conditions and watershed attributes. This study also employed the legacy United States Geological Survey (USGS) tools - LOADEST and WRTDS as benchmarks to evaluate the performance of XGBest in these regions. The results demonstrated that XGBest outperformed LOADEST and WRTDS in predictive accuracy and revealed critical insights into the spatial and temporal variability of nutrient and sediment loads. In addition, SHapley Additive exPlanations (SHAP) values highlighted the importance of integrating static and dynamic watershed attributes, such as land cover, antecedent discharge, and seasonality, in capturing the complex concentration-discharge (C-Q) relationships. This study positions XGBest as a robust and scalable water quality prediction tool that bridges the gap between hydrology and broader environmental management. By combining multiple environmental factors into a unified predictive framework, XGBest enhances our understanding of water quality and supports more effective environmental monitoring and management strategies.

Multi-evidences investigation into spatiotemporal variety, sources tracing, and health risk assessment of surface water nitrogen contamination in China

A comprehensive understanding of nitrogen pollution status, especially the identification of sources and fate of nitrate is essential for effective water quality management at the local scale. However, the nitrogen contamination of surface water across China was poorly understood at the national scale. A dataset related to nitrogen was established based on 111 pieces of literature from 2000 to 2020 in this study. The spatiotemporal variability, source tracing, health risk assessment, and drivers of China's surface water nitrogen pollution were analyzed by integrating multiple methods. These results revealed a significant spatiotemporal heterogeneity in the nitrogen concentration of surface water across China. Spatially, the Haihe River Basin and Yellow River Basin were the basins where surface water was seriously contaminated by nitrogen in China, while the surface water of Southwest Basin was less affected. Temporally, significant differences were observed in the nitrogen content of surface water in the Songhua and Liaohe River Basin, Pearl River Basin, Southeast Basin, and Yellow River Basin. There were 1%, 1%, 12%, and 46% probability exceeding the unacceptable risk level (HI＞1) for children in the Songhua and Liaohe River Basin, Pearl River Basin, Haihe River Basin, and Yellow River Basin, respectively. The primary sources of surface water nitrate in China were found to be domestic sewage and manure (37.7%), soil nitrogen (31.7%), and chemical fertilizer (26.9%), with a limited contribution from atmospheric precipitation (3.7%). Human activities determined the current spatiotemporal distribution of nitrogen contamination in China as well as the future development trend. This research could provide scientifically reasonable recommendations for the containment of surface water nitrogen contamination in China and even globally.

The state of nitrogen in rivers and streams across sub-Saharan Africa

The nutrient status of rivers and streams is less researched in sub-Saharan Africa than in many other inhabited regions of the world. Given the expected population growth, intensification of agriculture, increased pressure on natural ecosystems and projected climate change in sub-Saharan Africa, it is crucial to quantify and understand drivers behind spatiotemporal patterns of nitrogen concentrations and loads in rivers and streams. Such knowledge can support sustainable management of water resources with the goal to provide clean water, create and maintain healthy ecosystems and prevent excessive pollution of water resources with nitrogen compounds, as is found in large parts of North America, Europe and Asia. This review provides a synthesis of the current available data from peer-reviewed literature (n = 243) on particulate and dissolved nitrogen in rivers and streams in sub-Saharan Africa, looking into seasonal and land cover-related differences. The review shows that data on nitrogen concentrations in rivers and streams is available for 32 out of the 48 countries (67 %) in sub-Saharan Africa, highlighting large data gaps given the size of the region. Differences in nitrogen concentrations between land cover types are reported, with highest median total nitrogen (3.9 mg N L-1) and nitrate (1.2 mg N L-1) concentrations observed at sites characterised by settlement and industry. In contrast, natural land cover types, like forest, have higher median (N:P) ratios (> 14.6) than cropland and urban areas (< 12.0). The analysis of paired samples from dry and wet seasons reveals varying effects of seasonality on the concentration of different nitrogen compounds between land cover types. However, the processes driving these spatiotemporal differences are still poorly understood. These findings highlight the need for a targeted research agenda for Africa to advance our understanding of the role of rivers and streams in nitrogen cycling in different ecosystems and their interaction with anthropogenic and natural drivers of change.

Unravelling nitrate transformation mechanisms in karst catchments through the coupling of high-frequency sensor data and machine learning

Nitrate dynamics within a catchment are critical to the earth's system process, yet the intricate details of its transport and transformation at high resolutions remain elusive. Hydrological effects on nitrate dynamics in particular have not been thoroughly assessed previously and this knowledge gap hampers our understanding and effective management of nitrogen cycling in watersheds. Here, machine learning (ML) models were employed to reconstruct the annual variation trend in nitrate dynamics and isotopes within a typical karst catchment. Random forest model demonstrates promising potential in predicting nitrate concentration and its isotopes, surpassing other ML models (including Long Short-term Memory, Convolutional Neural Network, and Support Vector Machine) in performance. The ML-modeled NO3--N concentrations, δ15N-NO3-, and δ18O-NO3- values were in close agreement with field data (NSE values of 0.95, 0.80, and 0.53, respectively), which are notably challenging to achieve for process models. During the transition from dry to wet period, approximately 23.0 % of the annual precipitation (∼269.1 mm) was identified as the threshold for triggering a rapid response in the wet period. The modeled nitrate isotope values were significantly supported by the field data, suggesting seasonal variations of nitrogen sources, with precipitation as the primary driving force for fertilizer sources. Mixing of multiple sources appeared to be the main control of the transport and transformation of nitrate during the rising limb in the wet period, whereas process control (denitrification) took precedence during the falling limb, and the fate of nitrate was controlled by biogeochemical processes during the dry period.

Dynamic patterns and potential drivers of river water quality in a coastal city: Insights from a machine-learning-based framework and water management

River water quality continues to deteriorate under the coupled effects of climate change and human activities. Machine learning (ML) is a promising approach for analyzing water quality. Nevertheless, the spatiotemporal dynamics of river water quality and their potential mechanisms in changing environments remain incomprehensively understood through available ML-based researches. Here, we developed a ML-based framework integrating a self-organizing map (SOM) model with a random forest (RF) model. This framework was applied to simultaneously investigate the spatiotemporal patterns and potential drivers of river permanganate (CODMn), ammonia nitrogen (NH3-N), and total phosphorus (TP) dynamics across 34 sites from 2010 to 2020 in a coastal city threatened by deteriorating water environment in southeastern China. The sites were divided into two clusters in the spatial context with different water quality conditions. The year of 2015 for NH3-N and 2018 for CODMn and TP were identified as the key turning points of water quality variations. Features including sewage discharge, population dynamics, percentage of cultivated land, and fertilizer application contributed greatly to water quality deterioration. The increase in forest vegetation reflected by percentage of forest and leaf area index may improve water quality. The ML-based modeling framework demonstrated a promising way to address the spatiotemporal dynamics of river water quality, and provided insights for water management in a coastal city with intensifying human-nature interactions.

Pollution loads in the middle-lower Yangtze river by coupling water quality models with machine learning

Pollution control and environmental protection of the Yangtze River have received major attention in China. However, modeling the river's pollution load remains challenging due to limited monitoring and unclear spatiotemporal distribution of pollution sources. Specifically, anthropogenic activities' contribution to the pollution have been underestimated in previous research. Here, we coupled a hydrodynamic-based water quality (HWQ) model with a machine learning (ML) model, namely attention-based Gated Recurrent Unit, to decipher the daily pollution loads (i.e., chemical oxygen demand, COD; total phosphorus, TP) and their sources in the Middle-Lower Yangtze River from 2014 to 2018. The coupled HWQ-ML model outperformed the standalone ML model with KGE values ranging 0.77-0.91 for COD and 0.47-0.64 for TP, while also reducing parameter uncertainty. When examining the relative contributions at the Middle Yangtze River Hankou cross-section, we observed that the main stream and tributaries, lateral anthropogenic discharges, and parameter uncertainty contributed 15, 66, and 19% to COD, and 58, 35, and 7% to TP, respectively. For the Lower Yangtze River Datong cross-section, the contributions were 6, 69, and 25% for COD and 41, 42, and 17% for TP. According to the attention weights of the coupled model, the primary drivers of lateral anthropogenic pollution sources, in descending order of importance, were temperature, date, and precipitation, reflecting seasonal pollution discharge, industrial effluent, and first flush effect and combined sewer overflows, respectively. This study emphasizes the synergy between physical modeling and machine learning, offering new insights into pollution load dynamics in the Yangtze River.

Assessing Chlorophyll-a Variations in Caspian Sea during the COVID-19 Pandemic

The COVID-19 pandemic's disruptions to human activities prompted serious environmental changes. Here, we assessed the variations in coastal water quality along the Caspian Sea, with a focus on the Iranian coastline, during the lockdown. Utilizing Chlorophyll-a data from MODIS-AQUA satellite from 2015 to 2023 and Singular Spectrum Analysis for temporal trends, we found a 22% Chlorophyll-a concentration decrease along the coast, from 3.2 to 2.5 mg/m³. Additionally, using a deep learning algorithm known as Long Short-Term Memory Networks, we found that, in the absence of lockdown, the Chlorophyll-a concentration would have been 20% higher during the 2020-2023 period. Furthermore, our spatial analysis revealed that 98% of areas experienced about 18% Chlorophyll-a decline. The identified improvement in coastal water quality presents significant opportunities for policymakers to enact regulations and make local administrative decisions aimed at curbing coastal water pollution, particularly in areas experiencing considerable anthropogenic stress.

Toxicity threshold and ecological risk of nitrate in rivers based on endocrine-disrupting effects: A case study in the Luan River basin, China

Nitrate, as a crucial nutrient, is consistently targeted for controlling water eutrophication globally. However, there is considerable evidence suggesting that nitrate has endocrine-disrupting potential on aquatic organisms. In this study, the sensitivity of various adverse effects to nitrate nitrogen (nitrate-N) was compared, and a toxicity threshold based on endocrine-disrupting effects was derived. The spatiotemporal variations of nitrate-N concentrations in the Luan River basin were investigated, and the associated aquatic ecological risks were evaluated using a comprehensive approach. The results showed that reproduction and development were the most sensitive endpoints to nitrate, and their distribution exhibited significant differences compared to behavior. The derived threshold based on endocrine-disrupting effects was 0.65 mgL-1, providing adequate protection for the aquatic ecosystem. In the Luan River basin, the mean nitrate-N concentrations during winter (4.4 mgL-1) were significantly higher than those observed in spring (0.7 mgL-1) and summer (1.2 mgL-1). Tributary inputs had an important influence on the spatial characteristics of nitrate-N in the mainstream, primarily due to agricultural and population-related contamination. The risk quotients (RQ) during winter, summer, and spring were evaluated as 6.7, 1.8, and 1.1, respectively, and the frequency of exposure concentrations exceeding the threshold was 100 %, 64.3 %, and 42.5 %, respectively. At the ecosystem level, nitrate posed intermediate risks to aquatic organisms during winter and summer in the Luan River basin and at the national scale in China. We suggest that nitrate pollution control should not solely focus on water eutrophication but also consider the endocrine disruptive effect on aquatic animals.

Response of microbial communities to exogenous nitrate nitrogen input in black and odorous sediment

Since nitrate nitrogen (NO3--N) input has proved an effective approach for the treatment of black and odorous river waterbody, it was controversial whether the total nitrogen concentration standard should be raised when the effluent from the sewage treatment plant is discharged into the polluted river. To reveal the effect of exogenous nitrate (NO3--N) on black odorous waterbody, sediments with different features from contaminated rivers were collected, and the changes of physical and chemical characteristics and microbial community structure in sediments before and after the addition of exogenous NO3--N were investigated. The results showed that after the input of NO3--N, reducing substances such as acid volatile sulfide (AVS) in the sediment decreased by 80 % on average, ferrous (Fe2+) decreased by 50 %, yet the changing trend of ammonia nitrogen (NH4+-N) in some sediment samples increased while others decreased. High-throughput sequencing results showed that the abundance of Thiobacillus at most sites increased significantly, becoming the dominant genus in the sediment, and the abundance of functional genes in the metabolome increased, such as soxA, soxX, soxY, soxZ. Network analysis showed that sediment microorganisms evolved from a single sulfur oxidation ecological function to diverse ecological functions, such as nitrogen cycle nirB, nirD, nirK, nosZ, and aerobic decomposition. In summary, inputting an appropriate amount of exogenous NO3--N is beneficial for restoring and maintaining the oxidation states of river sediment ecosystems.

Unveiling Microbial Nitrogen Metabolism in Rivers using a Machine Learning Approach

Microbial nitrogen metabolism is a complicated and key process in mediating environmental pollution and greenhouse gas emissions in rivers. However, the interactive drivers of microbial nitrogen metabolism in rivers have not been identified. Here, we analyze the microbial nitrogen metabolism patterns in 105 rivers in China driven by 26 environmental and socioeconomic factors using an interpretable causal machine learning (ICML) framework. ICML better recognizes the complex relationships between factors and microbial nitrogen metabolism than traditional linear regression models. Furthermore, tipping points and concentration windows were proposed to precisely regulate microbial nitrogen metabolism. For example, concentrations of dissolved organic carbon (DOC) below tipping points of 6.2 and 4.2 mg/L easily reduce bacterial denitrification and nitrification, respectively. The concentration windows for NO3--N (15.9-18.0 mg/L) and DOC (9.1-10.8 mg/L) enabled the highest abundance of denitrifying bacteria on a national scale. The integration of ICML models and field data clarifies the important drivers of microbial nitrogen metabolism, supporting the precise regulation of nitrogen pollution and river ecological management.