Water Quality Data Sets Research Articles

Development of entropy and deviation-based water quality index: Case of river Ganga, India

• Development of an objective and rational framework ( M E D -WQI) for water quality assessment that uses entropy and deviation-based approach. • Evaluation of M ED -WQI using synthetic dataset by comparing with existing Composite WQI. • Assessment of river Ganga's water quality status using M ED -WQI. In this study, a comprehensive multiplicative water quality index, M ED -WQI, for surface water quality assessment has been developed that uses sub-index functions based on deviation from maximum contaminant level of respective parameters. The parametric weights in M ED -WQI were computed using entropy based approach to eliminate subjectivity. Using a synthetic dataset the performance of M ED -WQI has been compared with Composite WQI (CWQI) that is based on Saaty’s analytical hierarchical process. The results indicated that in comparison to CWQI, M ED -WQI approach provides an objective and rational framework that eliminates clustering effect in providing the water quality status. Further, M ED -WQI has been applied to an exhaustive water quality dataset of river Ganga, one of the major perennial rivers of India. Out of 224 sampling locations, 167 sites have been associated with excellent or good water quality class whereas 57 sites are identified with either fair, poor, or heavily-polluted water quality class. Furthermore, the water quality classes correlated well with the type of anthropogenic activities carried out at the site.

Artificial ecosystem optimization with Deep Learning Enabled Water Quality Prediction and Classification model

The majority of what is needed to maintain life is found in the approximately 70 percent of the earth's surface that is composed of water. Water quality has been deteriorating at an alarming rate as a direct result of rapid industrialization and urbanisation, which has led to a rise in the prevalence of serious diseases. In the past, determining the quality of water was typically accomplished by employing labor-intensive, time-consuming, and statistically pricey laboratory investigations, which renders the prevalent concept of real-time monitoring meaningless. The worrisome effect of poor water quality mandates the necessity of an alternative model that is both rapid and economical to implement. There has been a lot of talk about using artificial intelligence to forecast and model water quality as a means of preventing and reducing water pollution. An artificial ecosystem optimization with Deep Learning Enabled Water Quality Prediction and Classification (AEODL-WQPC) model is presented in this paper. The primary objectives of the AEODL-WQPC model that is being given are the prediction and categorization of different levels of water quality. As a first processing step, the data normalization technique is used to the provided AEODL-WQPC model so that this goal can be achieved. In addition to this, an optimal stacked bidirectional gated recurrent unit (OSBiGRU) model is used to forecast the water quality index (WQI), and the Adam optimizer is utilised in order to fine-tune the model's parameters. AEO with enhanced Elman Neural Network (AEO-IENN) model is utilised for the categorization of water quality. This model is characterized by the fact that the AEO algorithm effectively tunes the parameters associated to the ENN model. For the purposes of the experimental validation of the AEODL-WQPC model, a benchmark water quality dataset obtained from the Kaggle repository is utilised. The research that compared several models found that the AEODL-WQPC model had superior results to more recent state of the art methods.

An improved graph convolutional network with feature and temporal attention for multivariate water quality prediction.

The analysis and prediction of water quality are of great significance to water quality management and pollution control. In general, current water quality prediction methods are often aimed at single indicator, while the prediction effect is not ideal for multivariate water quality data. At the same time, there may be some correlations between multiple indicators which the conventional prediction models cannot capture. To resolve these problems, this paper proposes a deep learning model: Graph Convolutional Network with Feature and Temporal Attention (FTGCN), realizing the prediction for multivariable water quality data. Firstly, a feature attention mechanism based on multi-head self-attention is designed to capture the potential correlations between water indicators. Then, a temporal prediction module including temporal convolution and bidirectional GRU with a temporal attention mechanism is designed to deal with temporal dependencies of time series. Moreover, an adaptive graph learning mechanism is introduced to extract hidden associations between water quality indicators. An auto-regression module is also added to solve the disadvantage of non-linear nature of neural networks. Finally, an evolutionary algorithm is adopted to optimize the parameters of the proposed model. Our model is applied on four real-world water quality datasets, compared with other models for multivariate time series forecasting. Experimental results demonstrate that the proposed model has a better performance in water quality prediction than others by two indices.

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.

Spatial and temporal variations in riverine mercury in the Mackenzie River Basin, Canada, from community-based water quality monitoring data

Arctic rivers deliver ~40 t yr−1 of mercury (Hg) to the Arctic Ocean, ~6 % of which is from the Mackenzie River Basin (MRB), a region warming at ~3 times the mean hemispheric rate. How this will affect Hg transfer to ecosystems of the Beaufort Sea is a worrying issue. To help address this question, we analyzed >500 measurements of Hg and other water properties from 22 rivers collected in 2012–2018 by communities of the MRB. This new dataset provides a more comprehensive view of riverine Hg variations across the basin than was previously available. We find that rivers issued from mountains in the western MRB contribute the largest share of Hg in the Mackenzie River, 60–95 % of it being carried as fine suspended solids and probably sourced from riverbank erosion and thaw slumps. In contrast, lowland rivers of the central and eastern MRB contribute larger shares of dissolved Hg (up to 78 %), likely from recent atmospheric deposition through precipitation. Using load modelling constrained by the new water quality dataset, we estimate that the three largest western tributaries (Liard, Peel and Arctic Red rivers) of the Mackenzie contribute 60 % of the annual MRB THg export and DHg export to the Beaufort Sea during freshet, as well as 51 % of DHg export, while supplying 60 % of freshet discharge. Load modelling also reveals a sustained decline in DHg loads of ~13 kg yr−1 between 2001 and 2016 in the lower Mackenzie River, which likely reflect a decreasing trend in atmospheric Hg deposition over most of northwestern Canada during this period. This study highlights the value of community-based water quality monitoring in helping to support assessments of riverine Hg in the MRB in support of the Minamata Convention on Mercury.

Sampling re-design increases power to detect change in the Great Barrier Reef's inshore water quality.

Monitoring programs are fundamental to understanding the state and trend of aquatic ecosystems. Sampling designs are a crucial component of monitoring programs and ensure that measurements evaluate progress toward clearly stated management objectives, which provides a mechanism for adaptive management. Here, we use a well-established marine monitoring program for inshore water quality in the Great Barrier Reef (GBR), Australia to investigate whether a sampling re-design has increased the program’s capacity to meet its primary objectives. Specifically, we use bootstrap resampling to assess the change in statistical power to detect temporal water quality trends in a 15-year inshore marine water quality data set that includes data from both before and after the sampling re-design. We perform a comprehensive power analysis for six water quality analytes at four separate study areas in the GBR Marine Park and find that the sampling re-design (i) increased power to detect trends in 23 of the 24 analyte-study area combinations, and (ii) resulted in an average increase in power of 34% to detect increasing or decreasing trends in water quality analytes. This increase in power is attributed more to the addition of sampling locations than increasing the sampling rate. Therefore, the sampling re-design has substantially increased the capacity of the program to detect temporal trends in inshore marine water quality. Further improvements in sampling design need to focus on the program’s capability to reliably detect trends within realistic timeframes where inshore improvements to water quality can be expected to occur.

Coupling mountain and lowland watershed models to characterize nutrient loading: An eight-year investigation in Lake Chaohu Basin

Quantifying nitrogen (N) and phosphorus (P) sources and their contributions at a watershed scale is a primary step to design best management practices. This task is challenging for a mixed watershed with both mountain and lowland areas, because most existing watershed models were developed for mountain areas with clear water flow directions. To solve this challenge, we coupled the Soil and Water Assessment Tool (SWAT) for mountain watersheds, Phosphorus Dynamic model for lowland Polder systems (PDP) and Nitrogen Dynamic model for lowland Polder systems (NDP) for lowland artificial watersheds (i.e., polders) to characterize N and P loading for a typical mountain-lowland mixed watershed, Lake Chaohu Basin in China. The coupled model was calibrated based on eight-year (2011–2018) meteorological, hydrological and water quality datasets, and had a satisfactory performance for both hydrological and water quality variables (R2 > 0.6, NS > 0.6). Our eight-year investigation in the five watersheds of Lake Chaohu Basin revealed that 1) the estimated N and P loading had an annual averaged amount of 10,832 and 781 t/yr, and an intensity of 15.72 and 1.33 kg/ha/yr, respectively. Three watersheds (Nanfei, Pai and Baishitian River Watersheds) had relatively high loading intensity for N (35.16, 18.34 and 13.34 kg/ha/yr) and P (2.49, 1.25 and 2.72 kg/ha/yr), and contributed 67.6% (N) and 68.5% (P) to the loading of the five investigated watersheds. 2) Polders in Lake Chaohu Basin can be N and P sources or sinks for their surrounding rivers, depending on the polder-river interaction. Within Nanfei, Pai and Baishitian River Watersheds, N and P retention ability was higher in polders than that in mountain watersheds. 3) Population density, precipitation intensity and soil erosion were critical drivers of N and P loss and showed positive correlation with N and P loss intensity. This study demonstrated the high value model coupling to characterize N and P loading in a mountain-lowland mixed watershed system, and thus to improve our understanding on its N and P cycling.

Research on the Application of Multimodal-Based Machine Learning Algorithms to Water Quality Classification

With the development of society and the accelerated industrialization, the problem of water pollution has become increasingly prominent. In order to stop the gathering and diffusion of harmful substances in water bodies, leading to further deterioration of water quality and more serious environmental problems, environmental management departments have developed a series of pollutant discharge standards to prevent water pollution in real time. Common testing methods are the colorimetric method and TDS (total dissolved solids) value testing method, which are mostly through water bodies that contain acid, alkali, salt, and other indicators of the concentration test, to produce an assessment of water quality. However, the traditional methods of water quality testing, whether in the measurement time or in the accuracy of the test, are certain defects. In order to be able to quickly detect the concentration of water quality indicators in water bodies, timely response and treatment of highly polluted water bodies are urgently needed. In this paper, we propose a water quality detection classification model based on multimodal machine learning algorithm. Firstly, we preprocessed and analyzed the collected water quality dataset and determined the reasonable and perfect water quality classification influencing factors. Then, we successively built 15 kinds of classification models based on machine learning algorithms for water quality detection. At the same time, we evaluated the performance of each model. From the four evaluation indexes of precision, recall rate, F1 value, and accuracy, respectively, the real value is compared with the predicted value of each model. The experimental results show that sulfate, pH, solids, and hardness are the important influencing factors to perform water quality testing. And the three models XGBoost (Extreme Gradient Boosting), CatBoost (Categorical Boosting), and LGBM (Light Gradient Boosting Machine) have better performances in conducting water quality testing. Finally, we further optimized the classification models based on XGBoost, CatBoost, and LGBM by using two major tools: cross-validation and hyperparameter tuning.

The plight of Najafgarh drain in NCT of Delhi, India: assessment of the sources, statistical water quality evaluation, and fate of water pollutants.

The Najafgarh drain is the first major drain that joins the Yamuna River at Wazirabad in Delhi, India, and is known to contribute to the maximum pollution load to this river. The drain is originally an extension of the Sahibi River and was intentionally constructed as a canal to carry stormwater, but presently, it is carrying more of sewage, agricultural, and industrial effluents received through various small and large secondary drains. The present study has analyzed the water quality status of this interconnected system, i.e., the Najafgarh drain, its associated secondary drains, and the Yamuna River for physicochemical parameters (n = 16), microbiological parameter (n = 1), and heavy metal concentrations (n = 8). The analysis of the surface water samples collected during pre- and post-monsoon seasons showed that secondary drain discharges significantly impacted the water quality of the Najafgarh drain, which in turn affected the Yamuna River. Out of the eight selected secondary drains for this study, the Goyla dairy outlet came out as the most polluted site in terms of organic pollutants while the Basaidarapur drain was loaded with heavy metal contaminants. Statistical tools comprising hierarchical cluster analysis (HCA), Pearson's correlation, and principal component analysis (PCA) were further implemented on the water quality dataset for a better understanding of the possible sources of contamination for organic and inorganic pollutants in the selected sampling sites. The present study, thus, might help in providing key highlights to the policymakers for effective regulation and management of the point source discharges in Najafgarh drain, which will ultimately restrict its pollution loadings in Yamuna River, Delhi, and also help in the restoration of this important water body.

Anthropogenic induced physicochemical gradients and associated macroinvertebrate community changes in derived savannah stream in Nigeria: Implication for biotic assessment

Understanding the impact of anthropogenic activities on various ecosystem matrices is fundamental to the application of biotic assessment. Here we explored the spatial changes of macroinvertebrate community structure under varying physicochemical gradients in Adoka Stream at Eziani Town, Southeast Nigeria. The water and macroinvertebrate samples were obtained from three stations positioned along the stretch of the stream with station 1 taken as a reference to others. We adopted Principal Component Analysis (PCA) and Canonical Correspondence Analysis (CCA) to analyse the monitored dataset comprising of 16 parameters with the aim to identify the important physicochemical parameters, and further evaluated the relationship between physicochemical parameters and macroinvertebrates assemblages. The PCA in first two rotated components (RCs) showed that turbidity, EC, NO₃, PO₄, SO₄, Cl, BOD₅, COD, DO, pH, and HCO₃ explained 57% of variance in the water quality dataset and associated macroinvertebrates assemblage. Based on the CCA, temperature, pH, turbidity, EC, NO₃, SO₄, DO, HCO₃, Ca, and Mg were important to predicting changes of macroinvertebrates assemblage structure in the stream. The observed variations of the water quality were reflected on macroinvertebrate assemblage across the stations. High abundance, taxa richness and pollution sensitive taxa of Trichoptera were obtained for station 1 with good water quality while stations 2 and 3 with significant influence from anthropogenic activities which resulted to degrading water quality had low abundance and taxa richness as well as dominance of pollution-tolerant taxa of Diptera groups. Dolophilodes urceolus and Leptonema natalense were the dominant taxa of the trichopterans while individuals of taxa Chironomus transvaalensis, Paraphaenocladius sp., and Pentaneura nilotica dominated the dipterans. In conclusion, our results provided useful information that supports the protocol for biotic assessment and also baseline data on physicochemical conditions and macroinvertebrate composition of the stream.

Long range multi-step water quality forecasting using iterative ensembling

Real-life water quality monitoring applications such as aquaculture domains and water resource management need long range multi-step prediction for disaster control. However, prediction accuracy usually degrades gradually as the prediction target timepoint is further away from the current timepoint. To address this, recent water quality forecasting methods mostly rely on complex deep learning models. In this paper, we propose a simple time-variant iterative ensembling method that strives to significantly improve the performance of a given arbitrary long range multi-step time series predictor for water quality data with minimal increase in computational cost. With the given predictor, our proposed method iteratively uses ensembles of predicted values for preceding steps to improve the prediction accuracy for the succeeding steps. The iterative ensembling operation is performed on the trained model and only at the inference stage, and so does not need any further computing-intensive training for the performance improvement. We experimentally show that the proposed method is effective with 7 predictors and 9 water quality datasets of various types, and it outperforms the state-of-the-art results in those datasets by around 2%–29% in mean absolute error (MAE), root mean squared error (RMSE) and mean absolute percentage error (MAPE) metrics. Similar improvement has also been found in two other metrics such as normalized Nash–Sutcliffe model efficiency coefficient (NNSE) metric and Taylor diagram plot. Overall, the proposed iterative ensembling is a promising approach for multi-step long range water quality prediction for high-frequency water quality monitoring systems.

Perfomance analysis of Naive Bayes method with data weighting

Classification using naive bayes algorithm for air quality dataset has an accuracy rate of 39.97%. This result is considered not good and by using all existing data attributes. By doing pre-processing, namely feature selection using the gain ratio algorithm, the accuracy of the Naive Bayes algorithm increases to 61.76%. This proves that the gain ratio algorithm can improve the performance of the naive bayes algorithm for air quality dataset classification. Classification using naive bayes algorithm for air quality dataset. While the Water Quality dataset has an accuracy rate of 93.18%. These results are considered good and by using all the existing data attributes. By doing pre-processing, namely feature selection using the gain ratio algorithm, the accuracy of the Naive Bayes algorithm increases to 95.73%. This proves that the gain ratio algorithm can improve the performance of the naive bayes algorithm for air quality dataset classification. Classification using Naive Bayes algorithm for Water Quality dataset. Based on the tests that have been carried out on all data, it can be seen that the Weight nave Bayes classification model can provide better accuracy values ​​because there is a change in the weighting of the attribute values ​​in the dataset used. The value of the weighted Gain ratio is used to calculate the probability in Nave Bayes, which is a parameter to see the relationship between each attribute in the data, and is used as the basis for the weighting of each attribute of the dataset. The higher the Gain ratio of an attribute, the greater the relationship to the data class. So that the accuracy value increases than the accuracy value generated by the Naïve Bayes classification model. The increase in accuracy in the Naïve Bayes classification model is due to the number of weights from the attribute selection in the Gain ratio.

Fuzzy risk-based allocation of pollution load using extended trading-ratio system

Abstract Today, one of the most important aspects of urban planning and management is the issue of environmental protection. It is necessary to consider the effects of urban development on the environment in urban planning to achieve sustainable economic and industrial development. In this paper, an optimal planning structure has been developed to reduce the pollution load of Khorramabad River, Lorestan Province, Iran. The developed fuzzy trading-ratio system was programmed based on risk-based fuzzy analysis for nine dischargers of biochemical oxygen demand (BOD5) as a water quality index and optimized using a genetic algorithm. The calibrated and verified model was utilized to simulate the BOD5 concentration at checkpoints of the river using four data sets of water quality collected from 2018 to 2021 in August (2018, 2019 and 2020 for calibration and 2021 for verification). The results showed that BOD5 exchange in the downstream stations is in critical condition. Optimization to reduce the cost of wastewater treatment showed that the proposed model could be economically improved by about 11%. The feasible domain of risk changes was assessed at three levels of 30, 60 and 90%, with the maximum value of the objective function calculated for the alcohol factory and the minimum value obtained for the flour factory.

Spatial distribution of fibropapillomatosis in green turtles along the Queensland coast and an investigation into the influence of water quality on prevalence

AbstractFibropapillomatosis (FP) is a tumor‐forming disease which affects all species of marine turtle, but predominantly the green turtle (Chelonia mydas). Expression of this disease is thought to be precipitated by poor environmental conditions and often linked to anthropogenically induced environmental changes. Although FP is a globally distributed disease, targeted studies on the spatial distribution of the disease in Australia are limited. Here, we present the first comprehensive report of FP prevalence in Queensland, Australia. A retrospective analysis of 25,645 capture records for 15 sites along the Queensland coast were used to determine FP prevalence and trends in foraging green turtles. Within this data set, 791 turtles (3.1%) with FP tumors were recorded. Our analysis showed that prevalence varies between sites and years, with juvenile turtles being the most frequently affected by the disease. We found that survey method has a significant influence on the apparent FP prevalence detected at each site. That is, surveys which were explicitly FP‐targeted detected higher numbers of individual turtles with FP, and therefore generated higher prevalence rates than comprehensive population surveys. We also report the first attempt at developing water quality indices (WQIs) to compare with FP prevalence data in foraging green turtles. The WQIs were built from metrics published in a range of peer‐reviewed papers, reports, and based on expert opinion. Despite utilizing an extensive data set, a relationship between FP prevalence and WQI rankings at each site could not be quantified. The analysis was confounded by a range of limitations, including data gaps, varying temporal scales and data capture methods in the FP prevalence, and water quality data sets. This study has significant implications for management as it highlights the benefits of designing and collecting centralized data that can be integrated and used across multiple projects or programs.

Autochthonous sources and drought conditions drive anomalous oxygen-consuming pollution increase in a sluice-controlled reservoir in eastern China

Freshwater reservoirs are an important type of inland waterbody. However, they can suffer from oxygen-consuming pollution, which can seriously threaten drinking water safety and negatively impact the health of aquatic ecosystems. Oxygen-consuming pollutants originate from both allochthonous and autochthonous sources, and have temporally and spatially heterogeneous drivers. Datanggang Reservoir, China, is located in a small agricultural watershed; it is controlled by multiple sluice gates. Anomalously high oxygen consumption indicators were observed in this reservoir in March 2021. Here, it was hypothesized that autochthonous sources were the primary drivers of oxygen-consuming pollution in the reservoir under drought conditions. Datasets of water quality, precipitation, primary productivity, and sediment were used to analyze water quality trends in the reservoir and inflow rivers, demonstrating the effects of allochthonous inputs and autochthonous pollution. No correlation was found between reservoir oxygen consumption indicators and allochthonous inputs; reservoir oxygen consumption indicators and chlorophyll-a concentration were significantly positively correlated (p < 0.05). Substantially lower precipitation and higher water temperature and pH (compared to historical levels) were also observed before the pollution event. Therefore, during this period the hydrological conditions, water temperature, pH, and other variables caused by short-term drought conditions may have facilitated phytoplankton growth in the reservoir. This contributed to a large increase in autochthonous oxygen-consuming pollutants, as reflected by the abnormally high indicators. Sediments contaminated with organic matter may also have been an important contributor. As the effects of environmental management and pollution control continue to emerge, exogenous pollutants imported from the land to reservoirs are currently effectively controlled. However, endogenous pollutants driven by a variety of factors, such as meteorology and hydrology, will likely become the main drivers of short-term changes in oxygen-consuming pollution in freshwater reservoirs in the foreseeable future.

Evaluation of sampling frequency impact on the accuracy of water quality status as determined considering different water quality monitoring objectives

Water quality sampling is a key element in tracking water quality monitoring objectives. However, frequencies adapted by different agencies might not be sufficient to provide an accurate indication of water quality status. In this study, data from low- and high-resolution water quality datasets were analyzed to determine the extent to which monitoring objectives could be achieved with different sampling frequencies, with a view to providing recommendations and best practices for water quality monitoring frequency in places with limited resources with which to implement a high-frequency monitoring plan. Water quality data from two watersheds (Maumee River and Raisin River) located in the Western Lake Erie Basin (WLEB) were used since these watersheds have consistent records over substantial periods of time, and the water quality data available have a high resolution (at least daily). The water quality constituents analyzed included suspended solids (SS), total phosphorus (TP), soluble reactive phosphorus (SRP), and nitrate + nitrite (NO2+3). Sources of pollutants for watersheds located in the WLEB include contributions from point sources like discharges from sewage treatment plants and non-point sources such as agricultural and urban storm runoff. Weekly, bi-weekly, monthly, and seasonal datasets were created from the original datasets, following different sampling rules based on the day of the week, week of the month, and month of the year. The resulting datasets were then compared to the original dataset to determine how the sampling frequency would affect the results obtained in a water quality assessment when different monitoring objectives are considered. Results indicated that constituents easily transported by water (such as sediments and nutrients) require more than 50 samples/year to provide a small error (< 10%) with a confidence interval of 95%. Monthly and seasonal sampling were found appropriate to report a stream's prevailing water quality status and statistical properties. However, these resolutions might not be sufficient to capture long-term trends, in which case bi-weekly samples would be preferable. Limitations of low-resolution sampling frequency could be overcome by including rainfall events and random sampling during specific time windows as part of the monitoring plan.

Assessment of the Potential of Coordinating Two Interacting Monitoring Networks within the Lerma-Santiago Hydrologic System in Mexico

Water quality monitoring networks in the global south often display inefficiencies because monitoring strategies are frequently designed based on subjective professional judgments to define the temporal and spatial attributes of the networks, leading to poor cost–benefit relationships. The Lerma-Santiago Hydrological System (LSHS) in Mexico currently experiences severe environmental degradation caused by uncontrolled pollutant emissions from urban centers, agricultural, livestock, and industrial activities settled in the basin. While both the national and state authorities monitor this hydrological system, there has never been an effort to assess the monitoring efficiency of these two networks. The aim of the present study was to assess through multivariate statistical analyses the potential for coordination between these two interacting networks. For this purpose, two independent large water quality datasets with temporal and spatial attributes measured by two different authorities (the federal and the state) were used to identify those sites where coordination should be rationalized and those parameters that should continue to be monitored. The case study herein presented highlights the duplication in efforts to monitor surface water quality in the Lerma-Santiago hydrologic system, which implies a lack of coordination between the authorities and shows that water quality monitoring networks have not been reassessed since they were first implemented. Furthermore, using the case study of the Lerma-Santiago in Mexico, we expanded on various deficiencies, such as the use of different sampling frequencies and analytical methods by the authorities and inefficient communication among federal and state authorities. This study has revealed a large potential for coordinating two water quality monitoring networks (WQMN) in the Lerma-Santiago Hydrological System and a methodological approach that may be used to assess this potential. Coordination strategies for WQMNs can lead to significant cost reductions, extended network reach, and higher overall data quality in developing countries with limited financial resources and technical capabilities.

Water quantity and quality datasets for the simulation of nonlinearities for watershed adaptability to hydroclimate extremes in agricultural landscapes

Water quantity and quality datasets for the simulation of nonlinearities for watershed adaptability to hydroclimate extremes in agricultural landscapes

Multi-task learning framework for predicting water quality using non-linear machine learning technique

Predicting the quality of water is a very important issue in an ecosystem and it can be used to control the increase of water contamination. Also, water quality prediction is a prominent complex non-linear multi-target learning problem and extracting a relevant subset of features from a large number of features with multiple targets is a challenging task. Existing water quality prediction model not focused on multi-target learning process simultaneously and not identifying the non-linear relationship between the features and target variables. Therefore, this study proposes a multi-task learning method dealing with multi-target regression using non-linear machine learning technique. Finally, experiments are conducted to build a prediction model based on the proposed methods to evaluate accuracy on water quality dataset. The experimental results indicate that our method increases the overall accuracy of the experimental dataset compared with the existing methods with the reduced number of significant features.

Groundwater monitoring near oil sands development: Insights from regional water quality datasets in the Alberta Oil Sands Region (AOSR)

Study regionThis study is carried out across a 142,000 km2 area within the Alberta Oil Sands Region (AOSR), Alberta, Canada. Study focusGroundwater quality data for the AOSR are compiled and interpreted to provide information on regional water quality to inform groundwater monitoring and land use planning. A database of 546 water quality parameters measured between 1958 and 2015 from 5118 water wells is compiled, cleaned, and analyzed by hydrostratigraphic unit (HSU). New hydrologic insights for the regionBaseline water quality conditions were found to vary in the 12 main HSU’s, with wide ranges in total dissolved solids and geochemical facies, reflecting variable lithology and geochemical processes. Median concentrations for multiple parameters exceeded “interim trigger values” under consideration by government regulators. Statistically significant temporal changes in water quality were detected in the 2000’s in isolated areas of Surficial Sands aquifer, the Cretaceous and Devonian formations in the North Athabasca Oil Sands, and in Quaternary aquifers in the South Athabasca Oil Sands and Cold Lake Beaver River. Temporal anomalies occur in areas with enhanced vertical connectivity due to the presence of buried channels, incised rivers, or where the Colorado Group is thin or completely absent. The compiled dataset highlights the role of geochemical data in identifying aquifer connectivity and monitoring priority. Lack of publicly available data for key aquifers near some mining areas are noted.