Water Quality Data Sets Research Articles

Water Quality Prediction Method Based on Multi-Source Transfer Learning for Water Environmental IoT System

Water environmental Internet of Things (IoT) system, which is composed of multiple monitoring points equipped with various water quality IoT devices, provides the possibility for accurate water quality prediction. In the same water area, water flows and exchanges between multiple monitoring points, resulting in an adjacency effect in the water quality information. However, traditional water quality prediction methods only use the water quality information of one monitoring point, ignoring the information of nearby monitoring points. In this paper, we propose a water quality prediction method based on multi-source transfer learning for a water environmental IoT system, in order to effectively use the water quality information of nearby monitoring points to improve the prediction accuracy. First, a water quality prediction framework based on multi-source transfer learning is constructed. Specifically, the common features in water quality samples of multiple nearby monitoring points and target monitoring points are extracted and then aligned. According to the aligned features of water quality samples, the water quality prediction models based on an echo state network at multiple nearby monitoring points are established with distributed computing, and then the prediction results of distributed water quality prediction models are integrated. Second, the prediction parameters of multi-source transfer learning are optimized. Specifically, the back propagates population deviation based on multiple iterations, reducing the feature alignment bias and the model alignment bias to improve the prediction accuracy. Finally, the proposed method is applied in the actual water quality dataset of Hong Kong. The experimental results demonstrate that the proposed method can make full use of the water quality information of multiple nearby monitoring points to train several water quality prediction models and reduce the prediction bias.

A Long-Term Water Quality Prediction Method Based on the Temporal Convolutional Network in Smart Mariculture

In smart mariculture, traditional methods are not only difficult to adapt to the complex, dynamic and changeable environment in open waters, but also have many problems, such as poor accuracy, high time complexity and poor long-term prediction. To solve these deficiencies, a new water quality prediction method based on TCN (temporal convolutional network) is proposed to predict dissolved oxygen, water temperature, and pH. The TCN prediction network can extract time series features and in-depth data features by introducing dilated causal convolution, and has a good effect of long-term prediction. At the same time, it is predicted that the network can process time series data in parallel, which greatly improves the time throughput of the model. Firstly, we arrange the 23,000 sets of water quality data collected in the cages according to time. Secondly, we use the Pearson correlation coefficient method to analyze the correlation information between water quality parameters. Finally, a long-term prediction model of water quality parameters based on a time domain convolutional network is constructed by using prior information and pre-processed water quality data. Experimental results show that long-term prediction method based on TCN has higher accuracy and less time complexity, compared with RNN (recurrent neural network), SRU (simple recurrent unit), BI-SRU (bi-directional simple recurrent unit), GRU (gated recurrent unit) and LSTM (long short-term memory). The prediction accuracy can reach up to 91.91%. The time costs of training model and prediction are reduced by an average of 64.92% and 7.24%, respectively.

No widespread signature of the COVID-19 quarantine period on water quality across a spectrum of coastal systems in the United States of America

During the recent COVID-19 related quarantine period, anecdotal evidence emerged pointing to a rapid, sharp improvement in water quality in some localities. Here we present results from an analysis of the impacts of the COVID-19 quarantine period using two long-term coastal water quality datasets. These datasets rely on sampling that operates at appropriate timescales to quantify the influence of reduced human activity on coastal water quality and span coastal ecosystems ranging from low human influence to highly urbanized systems. We tested two hypotheses: 1) reduced tourism during the COVID-19 quarantine period would lead to improved coastal water quality, and 2) water quality improvements would scale to the level of human influence, meaning that highly urbanized or tourist-centric watersheds would see greater improvement than more rural watersheds. A localized reduction in fecal indicator bacteria was observed in four highly impacted regions of the Texas (USA) coast, but this pattern was not widespread. In less impacted regions, the signature of natural, decadal environmental variability (e.g., dissolved oxygen and turbidity) overwhelmed any potential signature of reduced human activity. Results from this study add to the growing body of literature on the environmental impacts of the COVID-19 quarantine period, and when considered with existing literature, emphasize that coastal water quality improvements appear to be ephemeral and reserved for the most severely affected (by human activity) systems. Furthermore, results show the importance of assessing COVID-19 signatures against long-term, decadal datasets that adequately reveal a system's natural variation.

Using the Mavic 2 Pro drone for basic water quality assessment

This paper assessed the capability of the Dà-Jiāng Innovations (DJI) Mavic 2 Pro Drone (unmanned aerial vehicle – UAV) for the collection and delivery of river water samples for basic water quality assessments. The primary objective of this paper was to evaluate how this UAV model could help in generating large water quality data sets in the developing world to assist in the design and implementation of water quality monitoring and assessment programs, which are often a challenge due to data paucity and resources. We hypothesized that the traditional approach (portable hand meters) to measuring in-situ water parameters, including pH, dissolved oxygen, electrical conductivity, and turbidity could not yield significant water quality data variations from those collected by the Mavic 2 Pro. The UAV was equipped with a plastic bottle attached to a three-meter rigid thin line for sample collection. Samples were collected at stations 50 m apart over a 300 m river length. The drone captured samples in wind conditions of about 10.1 km/h with ease. About 350 mL of samples were collected per mission. A paired t-test was performed to determine the parameter differences between the two approaches. We conclude that, given similar environmental, physical conditions and pilot experience, Mavic 2 Pro can generate large and much more reliable datasets at faster rates than the traditional approach. The drone also avoided obstacles with ease, a perfect technology for use in rural rivers. However, pilot efficiency and precision, including agitation during flight require further investigations considering their potential parameter influences. Similar future tests should investigate the performance of this drone model and data reliability over a long river course to ascertain its capability and suitability in various conditions in ecological applications.

Two novelty learning models developed based on deep cascade forest to address the environmental imbalanced issues: A case study of drinking water quality prediction

Environmental quality data sets are typically imbalanced, because environmental pollution events are rarely observed in daily life. Prediction of imbalanced data sets is a major challenge in machine learning. Our recent work has shown deep cascade forest (DCF), as a base learning model, is promising to be recommended for environmental quality prediction. Although some traditional models were improved by introducing the cost matrix, little is known about whether cost matrix could enhance the prediction performance of DCF. Additionally, feature extraction is also an important way to potentially improve the model's ability to predict the imbalanced data. Here, we developed two novelty learning models based on DCF: cost-sensitive DCF (CS-DCF) and DCF that combines unsupervised learning models and greedy methods (USM-DCF-G). Subsequently, CS-DCF and USM-DCF-G were successfully verified by an imbalanced drinking water quality data set. Our data presented both CS-DCF and USM-DCF-G show better prediction performance than that of DCF alone did. In particular, USM-DCF-G shows the best performance with the highest F1-score (95.12 ± 2.56%), after feature extraction and selection by using unsupervised learning models and greedy methods. Thus, the two learning models, especially USM-DCF-G, were promising learning models to address environmental imbalanced issues and accurately predict environmental quality.

Is water quality better in wet years or dry years in river-connected lakes? A case study from Dongting Lake, China

Water quality is essential for lake systems, which are not only influenced by climate change and human activities but are also controlled by high-frequency hydrological rhythms. However, the differences in water quality during different hydrological periods have not been addressed in detail. Here, a 15-year water quality dataset (2004–2018) was selected to explore the variation trends and their relationship with water level in different hydrological periods in Dongting Lake, a typical river-connected lake in China. The hydrological periods were classified into hydrological years and hydrological phases based on the characteristics of water level fluctuations. The results showed that annual variation in the water level in Dongting Lake fluctuated between 23.63 and 25.81 m from 2004 to 2018, and also displayed considerable water level differences ranging from 7.66 m (dry years) to 9.97 m (wet years) within a year. The water level of the lake phase showed significant differences among the different hydrological years. The concentration of TP, CODMn, and NH3-N showed significant decreasing trends, whereas that of TN showed a significant increasing trend from 2004 to 2018. The TN concentration in wet years was significantly higher than that in dry years, which could be attributed to sewage discharge and hydrological conditions. The contributions of the lake phase to the total sewage discharge successively decreased from 64.54% in wet years to 59.47% in dry years, while the river phase showed the opposite trends, ranging from 35.46% to 40.53%, reflecting the strong relationship between water regimes and pollutant fluxes. A regression analysis indicated the different responses of water quality to water level fluctuations over hydrological years, and water quality in different hydrological phases clearly separated the lake and river phases, highlighting the influence of water level fluctuations on water quality within a year. To maintain the water quality of Dongting Lake, the control of external load should not be relaxed, and hydrological regulation should be actively carried out within each year.

Effect of management on water quality and perception of ecosystem services provided by an urban lake

Costadone L, Sytsma MD, Pan Y, Rosenkranz M. 2021. Effect of management on water quality and perception of ecosystem services provided by an urban lake. Lake Reserv Manage. 37:418–430. Integrating the perception of water quality in management efforts can be an effective way of developing and evaluating management plans targeted at preserving lake ecosystem benefits. We present a case study of Oswego Lake (OR), an urban lake that has been intensively managed for the past 20 yr to preserve valuable esthetic and recreational benefits. We combined the analysis of a long-term water quality dataset with survey data to assess whether management efforts over time successfully met the expectations of people using the lake. The synergistic impact of both in-lake and watershed management activities significantly reduced whole-lake nutrient concentrations; however, high external phosphorus loading from native soil sources and the surrounding urbanized watershed likely continued to contribute to episodes of cyanobacteria blooms during the summer and partially limited in-lake management efforts. Although there was no statistically significant change in mean summer water transparency over the 20 yr management period, more than 60% of the people that have access to the lake thought the water quality had improved since they started using the lake. The lake was also perceived as “suitable” for esthetic and recreational enjoyment by users. As with other lakes in highly developed urban watersheds, Oswego Lake requires ongoing interventions to maintain adequate water quality. Management challenges include coordination of multiple stakeholders and jurisdictions with sometimes conflicting goals and constituencies, maintaining a long-term monitoring program that can produce consistent high-quality data, and keeping in touch with user groups to ensure that lake water quality meets expectations.

Using an unmanned aerial vehicle water sampler to gather data in a pit-lake mining environment to assess closure and monitoring.

Residual pit lakes from mining are often dangerous to sample for water quality. Thus, pit lakes may be rarely (or never) sampled. This study developed new technology in which water-sampling devices, mounted on an unmanned aerial vehicle (UAV), were used to sample three pit lakes in Nevada, USA, during 1week in 2017. Water-quality datasets from two of the three pit lakes on public lands, Dexter and Clipper, are presented here. The current conditions of the Dexter pit lake were assessed by examining cation and anion concentration changes that have occurred over a 17-year period since the pit lake was last sampled in 2000. Data gathered during this sampling campaign assessed 2017 conditions of the Dexter and Clipper pit lakes by comparing constituent concentrations to the Nevada Division of Environmental Protection (NDEP) pit lake water-quality requirements, indicating that selenium concentrations exceeded regulatory standards. We compared our sampling data for Dexter lake to prior water-quality data from the Dexter pit lake collected in 1999 and 2000. This comparison for the Dexter pit lake indicates that evapoconcentration may have caused increasing cation and anion concentrations. This UAV sampling approach can potentially incorporate the use of additional multiparameter probes: pH, oxygen concentration, turbidity, or chlorophyll. Some limitations of this UAV water-sampling methodology are battery duration, weather conditions, and payload capacity.

A comprehensive evaluation of surface water quality and potential health risk assessments of Sisa river, Kumasi

Surface water quality deterioration is one of the primary issues of the twenty-first century due to the high demand for drinking and irrigation purposes in most communities. The purpose of this study was to investigate the suitability of the Sisa river in Kumasi Metropolis, as well as the critical factors affecting its water quality and human health risk among children and adults. The findings indicated that the surface water was alkaline and, in some samples, turbidity, alkalinity, color, microbial, Fe, and Cr concentrations exceeded the standard permissible limits of WHO. Water from individual sampling sites contained Escherichia coli, total, and fecal coliform counts. There was no strong significant correlation among the levels of the heavy metals, suggesting their lack of control in their distribution, source, and transport pathway. Principal component analysis (PCA) revealed three factors that account for about 98.86 % of the total variance in the surface water quality data set. Moreover, PCA findings revealed that geogenic and anthropogenic sources were responsible for variations in physicochemical and microbiological parameters of the study area. Fe and Cr were found to have high toxic levels among the heavy metals analyzed. The heavy metal assessment index indicated moderate to high hazards. Cu had a hazard index greater than one at Duase and Atonsu, indicating a possible non-carcinogenic risk. The overall carcinogenic risk values were greater than 10−6, indicating the possibility of cancer. Water quality index values ranging from 97.95 to 137.22 percent showed that about 75 % of the samples were of poor quality. Based on these results, we recommended proper water management strategies and policies should be implemented as soon as possible to manage and protect the water resources.

Impact of lockdown on the environmental quality along the Indian coast and a tropical estuary

The COVID-19 lockdown significantly improved the global environmental quality by restricted anthropogenic activities. We aimed to understand the response of Indian coastal environmental quality and an estuarine (Zuari estuary, central west coast of India) water quality to the lockdown using satellite-derived data sets and in-situ measurements, respectively. The in-situ study was carried out for two weeks (during August and September 2020) to measure chlorophyll-a and macro-nutrients. Recent 5-years (2015–2019; March–August) data sets of aerosol optical depth (AOD), rainfall, air temperature, sea surface temperature (SST), chlorophyll-a, particulate inorganic carbon (PIC) and particulate organic carbon (POC) were compared with lockdown period (2020). The AOD was decreased during the lockdown period (23 ± 11%) as compared with recent 5-years shows the significant impact of lockdown on industrial, transport and urbanization activities. The water quality parameters and chlorophyll-a biomass have remained similar to the past data sets, indicating the insignificant effect of lockdown on these parameters. The results of the case study were compared with the previous reports, which showed a decrease in the concentration of nitrate, phosphate and silicate and comparable concentration of chlorophyll-a. Recent studies show significant impacts of lockdown on the atmospheric parameters and water quality parameters, however the present study did not observe any significant variations in the in-situ and satellite derived data sets of water quality and chlorophyll-a. We have discussed the possible reasons underlying in the less/undetectable response of the chlorophyll-a and water quality of the Indian coastal waters and estuaries during the lockdown.

Water Quality and Phytoplankton Distribution of the Lower Kinabatangan River Catchment, Sabah

A study on water quality and phytoplankton distribution was carried out at the Lower Kinabatangan River Catchment, Sabah in November 2013, January 2014 and March 2014. The objectives were to study the surface water quality of the Lower Kinabatangan River Catchment; to identify the composition of phytoplankton in three different types of land use in Sukau, Kinabatangan; and to determine spatial and temporal variations of water quality in Sukau, Kinabatangan. Three sampling stations were selected to represent different types of land use, consisting of oil palm plantation (OP), secondary forests (SF) and oxbow lake (OB). Based on Interim National Water Quality Standards (INWQS) for Malaysia, the parameters were categorized within Class I to Class IV. Statistical analyses ANOVA one-way, paired sample t-test and discriminant analysis have been carried out to both water quality and total monthly precipitation data sets. The distribution of phytoplankton in Kinabatangan River consisted of 5 divisions: the Bacillariophyta, Chlorophyta, Cyanophyta, Cryptophyta and Euglenophyta. Chlorophyta recorded the highest diversity, with 10 species recorded out of 17 species found of the Lower Kinabatangan River Catchment. Discriminant analysis suggested that UV-visible absorption coefficients at 254 and 340 nm were dominant in samples from OP and SF. Temporal variations showed that parameters suspended sediment, UV-visible absorption coefficients at 254 and 340 nm were dominant in samples from collected in January 2014.

Water Pollution Characteristics and Source Apportionment in Rapid Urbanization Region of the Lower Yangtze River: Considering the Qinhuai River Catchment

Clarifying the current situation of regional water pollutants and the relationship between pollutants and pollution sources is considered essential for managing the water environment. Water quality identification index (WQI), cluster analysis (CA), positive matrix factorization (PMF), and stable isotope analysis in R (SIAR) were employed to interpret a large and complex water quality data set of the Qinhuai River catchment generated during 2015 to 2019 to monitor of 11 parameters at 29 different sampling sites. WQI analysis indicated that water quality in Qinhuai River catchment is considered to have "moderate pollution," and an improving trend of water quality was observed at the interannual scale. TN was the most deteriorated of all pollution parameters. CA and PMF results on the spatial scale revealed that sampling sites located at downtown of Nanjing and Lishui District or Jangling University town were highly polluted due to the sewage from domestic sewage and business service sewage (28.88%) as well as industrial wastewater (27.43%), while sampling sites located at Hushu Street Administrative District, Ergan River, and Sangan River were slightly polluted by rural domestic wastewater and garbage (28.79%), and agricultural non-point source pollution (24.3%). The middle-lower reaches (Jiangning Development Zone and Moling Street) and middle reaches (Lukou Street Administrative District) were moderately polluted by industrial wastewater (27.25%), sewage from domestic wastewater and business service wastewater (31.62%) as well as inner sources (24.76%). The SIAR results showed that NO3--N was the main nitrogen form, and the NO3--N mainly originated from sewage (61%) and soil (34%) in the Yuntaishan River sub-catchment. These results will aid in the development of measures required to control water pollution in river catchments.

A study on water quality prediction by a hybrid CNN-LSTM model with attention mechanism.

The water environment plays an essential role in the mangrove wetland ecosystem. Predicting water quality will help us better protect water resources from pollution, allowing the mangrove ecosystem to perform its normal ecological role. New approaches to solve such nonlinear problems need further research since the complexity of water quality data and they are easily affected by the noise. In this paper, we propose a water quality prediction model named CNN-LSTM with Attention (CLA) to predict the water quality variables. We conduct a case study on the water quality dataset of Beilun Estuary to predict pH and NH3-N. Linear interpolation and wavelet techniques are used for missing data filling and data denoising, respectively. The hybrid model CNN-LSTM is highly capable of resolving nonlinear time series prediction problems, and the attention mechanism captures longer time dependence. The experimental results show that our model outperforms other ones, and can predict with different time lags in a stable manner.

Modelling Freshwater Eutrophication with Limited Limnological Data Using Artificial Neural Networks

Artificial Neural Networks (ANNs) have wide applications in aquatic ecology and specifically in modelling water quality and biotic responses to environmental predictors. However, data scarcity is a common problem that raises the need to optimize modelling approaches to overcome data limitations. With this paper, we investigate the optimal k-fold cross validation in building an ANN using a small water-quality data set. The ANN was created to model the chlorophyll-a levels of a shallow eutrophic lake (Mikri Prespa) located in N. Greece. The typical water quality parameters serving as the ANN’s inputs are pH, dissolved oxygen, water temperature, phosphorus, nitrogen, electric conductivity, and Secchi disk depth. The available data set was small, containing only 89 data samples. For that reason, k-fold cross validation was used for training the ANN. To find the optimal k value for the k-fold cross validation, several values of k were tested (ranging from 3 to 30). Additionally, the leave-one-out (LOO) cross validation, which is an extreme case of the k-fold cross validation, was also applied. The ANN’s performance indices showed a clear trend to be improved as the k number was increased, while the best results were calculated for the LOO cross validation as expected. The computational times were calculated for each k value, where it was found the computational time is relatively low when applying the more expensive LOO cross validation; therefore, the LOO is recommended. Finally, a sensitivity analysis was examined using the ANN to investigate the interactions of the input parameters with the Chlorophyll-a, and hence examining the potential use of the ANN as a water management tool for nutrient control.

Water-Quality Data Imputation with a High Percentage of Missing Values: A Machine Learning Approach

The monitoring of surface-water quality followed by water-quality modeling and analysis are essential for generating effective strategies in surface-water-resource management. However, worldwide, particularly in developing countries, water-quality studies are limited due to the lack of a complete and reliable dataset of surface-water-quality variables. In this context, several statistical and machine-learning models were assessed for imputing water-quality data at six monitoring stations located in the Santa Lucía Chico river (Uruguay), a mixed lotic and lentic river system. The challenge of this study is represented by the high percentage of missing data (between 50% and 70%) and the high temporal and spatial variability that characterizes the water-quality variables. The competing algorithms implement univariate and multivariate imputation methods (inverse distance weighting (IDW), Random Forest Regressor (RFR), Ridge (R), Bayesian Ridge (BR), AdaBoost (AB), Hubber Regressor (HR), Support Vector Regressor (SVR) and K-nearest neighbors Regressor (KNNR)). According to the results, more than 76% of the imputation outcomes are considered “satisfactory” (NSE > 0.45). The imputation performance shows better results at the monitoring stations located inside the reservoir than those positioned along the mainstream. IDW was the model with the best imputation results, followed by RFR, HR and SVR. The approach proposed in this study is expected to aid water-resource researchers and managers in augmenting water-quality datasets and overcoming the missing data issue to increase the number of future studies related to the water-quality matter.

Analysis of C4.5 Algorithm of Water Quality Dataset

The C4.5 algorithm still has weaknesses in predicting or classifying data if a large number of classes are used which can lead to increased decision-making time. So an approach is needed to improve the performance of the C4.5 algorithm with the selected split attributes that use the application of the average gain value to help predictions. The C4.5 algorithm is one of the Decision Tree methods in the classification process using the information entropy concept. The C4.5 algorithm uses the split criteria from ID3, the Gain Ratio is a modification of the method. The ID3 algorithm uses Information Gain (IG) for the split attribute criteria, while the C4.5 algorithm with Gain Ratio (GR), where the root value comes from high gain. The conclusion of the tests that have been carried out using the Water Quality dataset in the C4.5 method has an accuracy rate of 91.30%, with a classification error rate of 8.70%. Successful implementation using the C4.5 method in predicting the Water Quality dataset.

Characterizing the river water quality in China: Recent progress and on-going challenges

Food production systems, urbanization, and other anthropogenic activities dramatically alter natural hydrological and nutrient cycles, and are primarily responsible for water quality impairments in China's rivers. This study compiled a 16-year (2003–2018) dataset of river water quality (161,337 records from 2424 sites), watershed/landscape features, and meteorological conditions to investigate the spatial water quality patterns and underlying drivers of river impairment (defined as water quality worse than Class V according to China's Environmental Quality Standards for Surface Waters, GB3838-2002) at a national scale. Our analysis provided evidence of a distinct water quality improvement with a gradual decrease in the frequency of prevalence of anoxic conditions, an alleviation of the severity of heavy metal pollution, whereas the cultural eutrophication has only been moderately mitigated between 2003 and 2018. We also identified significant spatial variation with relatively poorer water quality in eastern China, where 17.2% of the sampling sites registered poor water quality conditions, compared with only 4.6% in western China. Total phosphorus (TP) and ammonia-nitrogen (NH3-N) are collectively responsible for >85% of the identified incidences of impaired conditions. Bayesian modelling was used to delineate the most significant covariates of TP/NH3-N riverine levels in six large river basins (Liao, Hai, Yellow, Yangtze, Huai, and Pearl). Water quality impairments are predominantly shaped by anthropogenic drivers (82.5% for TP, 79.5% for NH3-N), whereas natural factors appear to play a secondary role (20.5% for TP, 17.5% for NH3-N). Two indicator variables of urbanization (urban areal extent and nighttime light intensity) and farmland areal extent were the strongest predictors of riverine TP/NH3-N levels and collectively accounted for most of the ambient nutrient variability. We concluded that there is still a long way to go in order to eradicate eutrophication and realize acceptable ecological conditions. The design of the remedial measures must be tailored to the site-specific landscape characteristics, meteorological conditions, and should also consider the increasing importance of non-point source pollution and internal nutrient loading.

Discussion on the existing methodology of entropy-weights in water quality indexing and proposal for a modification of the expected conflicts.

The present research focuses on addressing various ambiguities in the existing method of integrating information entropy and water quality, thereby presenting a novel approach for an entropy-weighted water quality index. A three-dimensional water quality dataset is considered in the proposed method, the third dimension being the sampling frequency factor. The probability of observed values adhering to desirable limits prescribed by a standard code is estimated, leading to the computation of information entropy and, eventually, entropy weights. These weights are then used for the computation of the Modified Entropy-weight Water Quality Index (MEWQI) values. To verify the proposed method's applicability, the water quality dataset of Deepor Beel, India, was considered. IS 10500: 2012 was used for estimating MEWQI values. Results showed an excellent correlation with the observed dataset and their uncertainties of occurrence. The reliability and correctness of the proposed methodology were finally confirmed through both cluster analysis and sensitivity analysis. The cluster analysis showed remarkable associations with the computed MEWQI values, while the sensitivity analysis proved that no particular parameter was accountable for the contribution of MEWQI values; instead, all parameters exhibited equal contributions. The proposed methodology was thus found to be the most reasonable and reliable as it considered both factors, i.e., measured values concerning standard limits and the uncertainty, necessary for a consistent water quality monitoring program.

Charting a perfect storm of water quality pressures

The agri-food economy can be a significant driver of water quality pressures but the role of hydro-meteorological patterns in a changing climate also requires consideration. For this purpose, an assessment was made of a ten-year synchronous high temporal resolution water quality and hydro-meteorological dataset in Irish agricultural catchments. Changes occurring to rainfall intensity and soil temperature patterns were found to be important drivers of nutrient mobility in soils. There were links between the intensity of the North Atlantic Oscillation over the decade and large shifts in baseline nutrient concentrations in catchments. The data also revealed extreme weather impacts to pollution patterns including short periods of rain induced nutrient flux, that exceeded average annual mass loads in these catchments, and drought influences on point source pollution. These influences need consideration, and may require different mitigation strategies, as links between water quality land use pressure and water quality state in regulatory reviews. In a decade of both increased land use source and hydro-meteorological transport pressures, water quality natural capital in Ireland has faced a perfect storm. Such conditions are difficult to model and only revealed in high temporal resolution datasets.

Water Quality Prediction Method Based on Transfer Learning and Echo State Network

Due to the adjacency effect of space, the water quality information of adjacent monitoring points in the same water area is relevant. However, traditional prediction methods use the historical water quality information of only one monitoring point rather than other nearby monitoring points. It is important to note that the sample quantity has a great influence on the prediction accuracy of the prediction methods which require lots of training samples. In addition, the water quality information is time series. Whether the time sequence of water quality information can be effectively used has direct impact on the prediction accuracy. Therefore, this paper proposes a water quality prediction method based on transfer learning and echo state network (ESN). First, we transfer the features of water quality samples of the nearby monitoring point to the prediction monitoring point based on transfer component analysis, and further select the high-quality samples based on similarity and time sequence. In this way the sample quantity of the prediction monitoring point is increased on the premise of ensuring the sample quality. Second, we construct a water quality prediction model based on ESN which takes the advantage of the time sequence of water quality information, improving the prediction accuracy on the premise of ensuring the training time. Finally, the proposed method is applied in the actual water quality dataset of Hong Kong. The experiment results demonstrate that the proposed method can expand the quantity of training samples and further improve the accuracy of water quality prediction by making full use of the space-time characteristics of water quality information.