Spatial and seasonal characteristics of river water chemistry in the Taizi River in Northeast China

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.

Spatiotemporal analysis and prediction of water quality in Pearl River, China, using multivariate statistical techniques and data-driven model

Identifying spatiotemporal variation patterns and predicting future water quality are critical for rational and effective surface water management. In this study, an exploratory analysis and forecast workflow for water quality in Pearl River, Guangzhou, China, was established based on the 4-h interval dataset selected from 10 stations for water quality monitoring from 2019 to 2021. The multiple statistical techniques, such as cluster analysis (CA), principal component analysis (PCA), correlation analysis (CoA), and redundancy analysis (RDA), as well as data-driven model (i.e., gated recurrent unit (GRU)), were applied for assessing and predicting the water quality in the basin. The investigated sampling stations were classified into 3 categories based on differences in water quality, i.e., low, moderate, and high pollution regions. The average water quality indexes (WQI) values ranged from 38.43 to 92.63. Nitrogen was the most dominant pollutant, with high TN concentrations of 0.81-7.67 mg/L. Surface runoff, atmospheric deposition, and anthropogenic activities were the major contributors affecting the spatiotemporal variations in water quality. The decline in river water quality during the wet season was mainly attributed to increased surface runoff and extensive human activities. Furthermore, the short-term prediction of river water quality was achieved using the GRU model. The result indicated that for both DLCK and DTJ stations, the WQI for the 5-day lead time were predicted with accuracies of 0.82; for the LXH station, the WQI for the 3-day lead time was forecasted with an accuracy of 0.83. The finding of this study will shed a light on an effective reference and systematic support for spatio-seasonal variation and prediction patterns of water quality.

A data-driven framework for spatiotemporal characteristics, complexity dynamics, and environmental risk evaluation of river water quality

To evaluate the evolution of river water quality in a changing environment, measuring the objective water quality is critical for understanding the rules of river water pollution. Based on the sample entropy theory and a nonlinear statistical method, this study aims to identify the spatiotemporal dynamics of water quality and its complexity in the Yangtze River basin using time series data, to separate the contributions of human activity and climate change to water quality, and to establish a data-driven risk assessment framework for the spatial (potential risk) and temporal (direct risk) aspects of water pollution. The results demonstrate that the spatiotemporal dynamics of water quality and sample entropy in each monitoring section are closely related to the characteristics of the corresponding location. The water quality of the main stream is superior, and its complexity is less than that of the tributaries. Cascade reservoir operation and vegetation status, agricultural production, and rainfall patterns exert great influences in the upper, middle, and lower reaches, respectively. Dam construction, urban agglomeration development, and interactions between river and lake are also influencing factors. An attributional analysis found that climate change and human activities negatively contributed to the evolution of NH₃-N concentration in most of the monitored sections, and the average relative contribution rates of human activities to changes in water quality in the main and tributary streams were -55.46% and -48.49%, respectively. In addition, the construction of data-driven risk assessment framework can efficiently and accurately assess the potential and direct water pollution risks of rivers.

Pattern recognition of water quality variance in Yamuna River (India) using hierarchical agglomerative cluster and principal component analyses

The monitoring and assessment of a river system is a complex process and not restricted to urban areas only. The discharge of wastewater drains in the river increases the river system complexity further. The abstraction of freshwater at regular intervals and the discharge of the wastewater from various sources cause significant spatial and temporal variation in water quality. The multivariate statistical analysis is performed to identify water quality parameters' variability on the 5-year dataset from four monitoring sites. Hierarchical agglomerative cluster analysis (HACA) and principal component analysis (PCA) are applied to characterize the water quality parameters and identify the significant pollution sources. The clusters are formed considering the similarities between parameters, and eigenvalues are determined from the covariance of parameters. The box plots are designed to identify the spatial and temporal variations. The highest variability of the first principal component is 60.78% of the total variance at the second sampling location, the ITO bridge. The significant varifactors obtained from the PCA indicate the parameters responsible for the maximum variation in water quality. The study reveals the importance of multivariate statistical techniques in identifying a pattern of variability of parameters and developing management strategies to improve river water quality by identifying dominant parameters causing the maximum degradation in water quality.

Water quality assessment and pollution source apportionment in a highly regulated river of Northeast China

Dams and sluices break down the river continuum, alter the river hydrological regime, and intercept the migration processes of nutrients and pollutants. The regulation of dams and sluices will have great impacts on water quality characteristics in the river basin. In this study, variable fuzzy pattern recognition model (VFPR), principal component analysis/factor analysis (PCA/FA), and the absolute principal component score-multiple linear regression (APCS-MLR) were used to assess the water quality and identify the potential pollution sources in a highly regulated river of Northeast China. A set of water quality variables at three stations were measured from January 2015 to August 2017. The water quality assessment results showed that there were spatial and temporal variations of water quality and the total nitrogen (TN) and fecal coliforms (F. coli) were the major pollution factors of the study river section. Four pollution sources, including industrial effluent source, domestic sewage source, meteorological factor and atmospheric deposition source, and agricultural non-point source, were identified in dry and wet seasons using the PCA/FA method. The APCS-MLR results showed that the industrial effluent source was the main pollution source in dry seasons and had a decrease in wet seasons. While the mean contribution of the domestic sewage source had an increase in wet seasons, influenced by the sewage overflow and the flushing of pollutants during the extreme precipitation, the construction of dams decreased the flow obviously in wet seasons and increased in dry seasons. The increase in pollutants caused by storm runoff and the reduction of dilution water in the river channel could be the main reason for the water quality degradation in wet seasons.

Spatial characteristics and influencing factors of river pollution in China

Based on recent water quality data collected from 763 monitoring sections nationwide, this study examined the concentration of major pollutants in China's major rivers. A spatial autocorrelation analysis confirmed that river pollution was spatially uneven and clustered. While pollution of surface water was a nationwide concern, most serious water pollution happened in the Huai, Hai, Yellow, and Liao river Basins in Northern China. The results of the spatial regression analysis showed that GDP per capita, surface water stock, population, and economic structure were all significantly correlated with surface water pollution, with population having strongest impact, followed by level of economic development. By investigating the common characteristics shared by the "hotspot" cities where serious water pollution occurred, this study recommended a regional or basin approach to assessing water quality and controlling river pollution that cuts across jurisdiction boundaries. While China has made considerable progress in improving water productivity, there is still enormous potential in water conservation. It is also imperative to restructure local economy and develop water-efficient, less polluting industries and services. PRACTITIONER POINTS: River pollution in China was spatially uneven and clustered. Most serious water pollution happened in the Huai, Hai, Yellow, and Liao river basins in Northern China. GDP per capita, surface water stock, population, and economic structure correlated with surface water pollution, with population having strongest impact. A regional or basin approach was recommended to assess water quality and controlling river pollution across jurisdiction boundaries. It is also imperative to restructure local economy and develop water-efficient, less polluting industries and services.

Determining the macroinvertebrate community indicators and relevant environmental predictors of the Hun-Tai River Basin (Northeast China): A study based on community patterning

It is essential to understand the patterning of biota and environmental influencing factors for proper rehabilitation and management at the river basin scale. The Hun-Tai River Basin was extensively sampled four times for macroinvertebrate community and environmental variables during one year. Self-Organizing Maps (SOMs) were used to reveal the aggregation patterns of the 355 samples. Three community types (i.e., clusters) were found (at the family level) based on the community composition, which showed a clearly gradient by combining them with the representative environmental variables: minimally impacted source area, intermediately anthropogenic impacted sites, and highly anthropogenic impacted downstream area, respectively. This gradient was corroborated by the decreasing trends in density and diversity of macroinvertebrates. Distance from source, total phosphorus and water temperature were identified as the most important variables that distinguished the delineated communities. In addition, the sampling season, substrate type, pH and the percentage of grassland were also identified as relevant variables. These results demonstrated that macroinvertebrates communities are structured in a hierarchical manner where geographic and water quality prevail over temporal (season) and habitat (substrate type) features at the basin scale. In addition, it implied that the local-scale environment variables affected macroinvertebrates under the longitudinal gradient of the geographical and anthropogenic pressure. More than one family was identified as the indicator for each type of community. Abundance contributed significantly for distinguishing the indicators, while Baetidae with higher density indicated minimally and intermediately impacted area and lower density indicated highly impacted area. Therefore, we suggested the use of abundance data in community patterning and classification, especially in the identification of the indicator taxa.

Identifying nitrate sources and transformations in Taizi River Basin, Northeast China

The nitrate (NO₃^-) pollution of aquatic systems in Northeast China is a severe problem. To identify NO₃^- sources and transformations in different zones with different land uses in the Taizi River Basin, ion-exchange methods were utilized to determine the concentrations and isotopic compositions (δ¹⁵N and δ¹⁸O) of NO₃^- and chloride (Cl^-). Results showed that Cl^- concentrations ranged from 2.7 to 73.4 mg/L. Cl^- concentrations were the highest in zone 8 and the lowest in zone 1. NO₃^- concentrations ranged from 0.3 to 27.4 mg/L and were the highest in zone 1 and the lowest in zone 8. During the sampling period, δ¹⁵N-NO₃^- values varied from 3.8 to 37.2‰, and δ¹⁸O-NO₃^- values ranged from -0.5 to 10.4‰. δ¹⁵N-NO₃^- values were the highest in zone 9 and the lowest in zone 1. The differences in physicochemical parameters and NO₃^- isotopes may be affected by land use and biogeochemical nitrogen processes in different zones. The combined analysis of dual isotopes (δ¹⁵N-NO₃^- and δ¹⁸O-NO₃^-) and NO₃^-/Cl^- versus Cl^- showed that different sources contributed NO₃^- to different zones during the sampling period. Soil N, manure, and sewage were the main NO₃^- sources in the Taizi River Basin. In zones 1 to 6, the δ¹⁵N-NO₃^- values of almost all samples were more than 10‰, NO₃^-/Cl^- values were high, and Cl^- molar concentration was low during the sampling period. These findings suggested that the volatilization and nitrification of soil NH₄⁺ might be related to NO₃^- sources in zones 1 to 6. A 1:1 to 2:1 linear relationship between δ¹⁵N-NO₃^- and δ¹⁸O-NO₃^- combined with the significantly negative relationship between ln (NO₃^-) and δ¹⁸O-NO₃^- indicated that denitrification affected NO₃^- distribution in zones 8 to 9 during the sampling period. These results can provide useful information to control NO₃^- concentrations in different zones in Taizi River Basin.

Evaluation of potential factors affecting deriving conductivity benchmark by utilizing weighting methods in Hun-Tai River Basin, Northeastern China

Specific conductivity is an increasingly important stressor for freshwater ecosystems. Interacting with other environmental factors, it may lead to habitat degradation and biodiversity loss. However, it is still poorly understood how the effect of specific conductivity on freshwater organisms is confounded by other environmental factors. In this study, a weight-of-evidence method was applied to evaluate the potential environmental factors that may confound the effect of specific conductivity on macroinvertebrate structure communities and identify the confounders affecting deriving conductivity benchmark in Hun-Tai River Basin, China. A total of seven potential environmental factors were assessed by six types of evidence (i.e., correlation of cause and confounder, correlation of effect and confounder, the contingency of high level cause and confounder, the removal of confounder, levels of confounder known to cause effects, and multivariate statistics for confounding). Results showed that effects of dissolved oxygen (DO), fecal coliform, habitat score, total phosphorus (TP), pH, and temperature on the relationship between sensitive genera loss and specific conductivity were minimal and manageable. NH₃-N was identified as a confounder affecting deriving conductivity benchmark for macroinvertebrate. The potential confounding by high NH₃-N was minimized by removing sites with NH₃-N > 2.0 mg/L from the data set. Our study tailored the weighting method previously developed by USEPA to use field data to develop causal relationships for basin-scale applications and may provide useful information for pollution remediation and natural resource management.

Characteristics and source identification of dissolved trace elements in the Jinshui River of the South Qinling Mts., China

Dissolved trace elements and physiochemical parameters were analyzed to investigate their physicochemical characteristics and identify their sources at 12 sampling sites of the Jinshui River in the South Qinling Mts., China from October 2006 to November 2008. The two-factor ANOVA indicated significant temporal variations of the dissolved Cu, Fe, Sr, Si, and V (p < 0.001 or p < 0.05). With the exception of Sr (p < 0.001), no significant spatial variations were found. Distributions and concentrations of the dissolved trace elements displayed that dissolved Cu, Fe, Sr, Si, V, and Cr were originated from chemical weathering and leaching from the soil and bedrock. Dissolved Cu, Fe, Sr, As, and Si were also from anthropogenic inputs (farming and domestic effluents). Correlation and regression analysis showed that the chemical and physical processes of dissolved Cu was influenced by water temperature and dissolved oxygen (DO) to some degree. Dissolved Fe and Sr were affected by colloid destabilization or sedimentary inputs. Concentrations of dissolved Si were slightly controlled by biological uptake. Principal component analysis confirmed that Fe, Sr, and V resulted from domestic effluents, agricultural runoff, and confluence, whereas As, Cu, and Si were from agricultural activities, and Cr and Zn through natural processes. The research results provide a reference for ecological restoration and protection of the river environment in the Qinling Mts., China.

Use of multivariate statistical techniques for the evaluation of temporal and spatial variations in water quality of the Kaduna River, Nigeria

Multivariate statistical techniques, such as cluster analysis (CA) and principal component analysis/factor analysis (PCA/FA), were used to investigate the temporal and spatial variations and to interpret large and complex water quality data sets collected from the Kaduna River. Kaduna River is the main tributary of Niger River in Nigeria and represents the common situation of most natural rivers including spatial patterns of pollutants. The water samples were collected monthly for 5 years (2008-2012) from eight sampling stations located along the river. In all samples, 17 parameters of water quality were determined: total dissolved solids (TDS), pH, Thard, dissolved oxygen (DO), 5-day biochemical oxygen demand (BOD5), chemical oxygen demand (COD), NH4-N, Cl, SO4, Ca, Mg, total coliform (TColi), turbidity, electrical conductivity (EC), HCO3 (-), NO3 (-), and temperature (T). Hierarchical CA grouped 12 months into two seasons (dry and wet seasons) and classified eight sampling stations into two groups (low- and high-pollution regions) based on seasonal differences and different levels of pollution, respectively. PCA/FA for each group formed by CA helped to identify spatiotemporal dynamics of water quality in Kaduna River. CA illustrated that water quality progressively deteriorated from headwater to downstream areas. The results of PCA/FA determined that 78.7 % of the total variance in low pollution region was explained by five factor, that is, natural and organic, mineral, microbial, organic, and nutrient, and 87.6 % of total variance in high pollution region was explained by six factors, that is, microbial, organic, mineral, natural, nutrient, and organic. Varifactors obtained from FA indicated that the parameters responsible for water quality variations are resulted from agricultural runoff, natural pollution, domestic, municipal, and industrial wastewater. Mann-Whitney U test results revealed that TDS, pH, DO, T, EC, TColi, turbidity, total hardness (THard), Mg, Ca, NO3 (-), COD, and BOD were identified as significant variables affecting temporal variation in river water, and TDS, EC, and TColi were identified as significant variables affecting spatial variation. In addition, box-whisker plots facilitated and supported multivariate analysis results. This study illustrates the usefulness of multivariate statistical techniques for classification and processing of large and complex data sets of water quality parameters, identification of latent pollution factors/sources and their spatial-temporal variations, and determination of the corresponding significant parameters in river water quality.