An ensemble model for accurate prediction of key water quality parameters in river based on deep learning methods.

Deep learning models provide a more powerful method for accurate and stable prediction of water quality in rivers, which is crucial for the intelligent management and control of the water environment. To increase the accuracy of predicting the water quality parameters and learn more about the impact of complex spatial information based on deep learning models, this study proposes two ensemble models TNX (with temporal attention) and STNX (with spatio-temporal attention) based on seasonal and trend decomposition (STL) method to predict water quality using geo-sensory time series data. Dissolved oxygen, total phosphorus, and ammonia nitrogen were predicted in short-step (1 h, and 2 h) and long-step (12 h, and 24 h) with seven water quality monitoring sites in a river. The ensemble model TNX improved the performance by 2.1%-6.1% and 4.3%-22.0% relative to the best baseline deep learning model for the short-step and long-step water quality prediction, and it can capture the variation pattern of water quality parameters by only predicting the trend component of raw data after STL decomposition. The STNX model, with spatio-temporal attention, obtained 0.5%-2.4% and 2.3%-5.7% higher performance compared to the TNX model for the short-step and long-step water quality prediction, and such improvement was more effective in mitigating the prediction shift patterns of long-step prediction. Moreover, the model interpretation results consistently demonstrated positive relationship patterns across all monitoring sites. However, the significance of seven specific monitoring sites diminished as the distance between the predicted and input monitoring sites increased. This study provides an ensemble modeling approach based on STL decomposition for improving short-step and long-step prediction of river water quality parameter, and understands the impact of complex spatial information on deep learning model.

Pilot investigation on biostability of drinking water distribution systems under water source switching.

Water quality deterioration of drinking water distribution systems (DWDSs) caused by water source switching has been reported previously. However, systematic investigation of the biostability of DWDS under water source switching is limited. Aged pipes, including three commonly used pipe materials dug out from a practical DWDS, were used to systematically investigate the biofilm stability mechanism of DWDS under water source switching to quality-improved water. An increase in adenosine triphosphate (ATP) concentration in the bulk water during the initial stage of the switching period was observed, indicating the risk of biofilm release through aged pipe surfaces after water source switching. Sloughing of biofilms might contribute to temporary instability. From day 35, the ATP concentration in the polyethylene (PE) and plastic stainless steel composite (PS) pipes were maintained at approximately 2.40 and 2.56 ng/L, respectively. In contrast, the ATP concentration in the ductile iron (DI) pipes was higher, at approximately 3.43 ng/L from day 42. The water quality variation could cause areas of the biofilm to slough and reduce the biomass of biofilm, causing partial alteration of the microbial community. 16S rRNA gene amplicon sequencing-based functional prediction revealed that the biofilm could increase the abundance of chlorine-resistant bacteria attributed to the increase in Pseudomonas and Methylobacterium after switching to quality-improved water. Moreover, the profiles of specific pathways linked to human diseases, antibiotic resistance, and antibiotic biosynthesis revealed that the safety of the biofilm could improve after switching to quality-improved water. KEY POINTS: • The PE and PS biofilm showed improved resistance to water quality perturbation. • Greater number of Methylobacterium was found in the biofilm after water source switching. • 3.16S gene-based metagenomics prediction revealed that the safety of the biofilm under water source switching.

Inhibition of Phenol from Entering into Condensed Freshwater by Activated Persulfate during Solar-Driven Seawater Desalination.

Recently, solar-driven seawater desalination has received extensive attention since it can obtain considerable freshwater by accelerating water evaporation at the air-water interface through solar evaporators. However, the high air-water interface temperature can cause volatile organic compounds (VOCs) to enter condensed freshwater and result in water quality safety risk. In this work, an antioxidative solar evaporator, which was composed of MoS2 as the photothermal material, expandable polyethylene (EPE) foam as the insulation material, polytetrafluoroethylene (PTFE) plate as the corrosion resistant material, and fiberglass membrane (FB) as the seawater delivery material, was fabricated for the first time. The activated persulfate (PS) methods, including peroxymonosulfate (PMS) and peroxodisulfate (PDS), were applied to inhibit phenol from entering condensed freshwater during desalination. The distillation concentration ratio of phenol (RD) was reduced from 76.5% to 0% with the addition of sufficient PMS or PDS, which means that there was no phenol in condensed freshwater. It was found that the Cl- is the main factor in activating PMS, while for PDS, light, and heat are the dominant. Compared with PDS, PMS can make full utilization of the light, heat, Cl- at the evaporator's surface, resulting in more effective inhibition of the phenol from entering condensed freshwater. Finally, though phenol was efficiently removed by the addition of PMS or PDS, the problem of the formation of the halogenated distillation by-products in condensed freshwater should be given more attention in the future.

Removal of antibiotic-resistant genes during drinking water treatment: A review.

Once contaminate the drinking water source, antibiotic resistance genes (ARGs) will propagate in drinking water systems and pose a serious risk to human health. Therefore, the drinking water treatment processes (DWTPs) are critical to manage the risks posed by ARGs. This study summarizes the prevalence of ARGs in raw water sources and treated drinking water worldwide. In addition, the removal efficiency of ARGs and related mechanisms by different DWTPs are reviewed. Abiotic and biotic factors that affect ARGs elimination are also discussed. The data on presence of ARGs in drinking water help come to the conclusion that ARGs pollution is prevalent and deserves a high priority. Generally, DWTPs indeed achieve ARGs removal, but some biological treatment processes such as biological activated carbon filtration may promote antibiotic resistance due to the enrichment of ARGs in the biofilm. The finding that disinfection and membrane filtration are superior to other DWTPs adds weight to the advice that DWTPs should adopt multiple disinfection barriers, as well as keep sufficient chlorine residuals to inhibit re-growth of ARGs during subsequent distribution. Mechanistically, DWTPs obtain direct and inderect ARGs reduction through DNA damage and interception of host bacterias of ARGs. Thus, escaping of intracellular ARGs to extracellular environment, induced by DWTPs, should be advoided. This review provides the theoretical support for developping efficient reduction technologies of ARGs. Future study should focus on ARGs controlling in terms of transmissibility or persistence through DWTPs due to their biological related nature and ubiquitous presence of biofilm in the treatment unit.

Nitrous oxide emission from stormwater biofilters in alternating dry and wet weather.

The nitrous oxide (N2O) flux and its possible production pathways from stormwater biofilters in response to moisture content (MC) due to a shift from dry to wet weather was investigated. In this study, we evaluated the changes in the composition of the bacterial community, the relative abundance of functional genes, and N2O emission rate in laboratory-scale stormwater biofilters in response to changes of MC. The results indicated that N2O flux correlated positively with soil MC (r = 0.722 p < 0.01). We observed a higher rates of N2O flux when shifting from dry to wet conditions. Notably, these values decreased substantially within 8-24 h in response to the rapid decline in MC, and then gradually decreased and stabilized at 4.4-12.0 µg/m2·h. The relative abundance of ammonia-oxidizing and denitrifying bacteria, as well as the relative abundance of functional groups associated with denitrification was higher under conditions of low soil MC than those in the high MC. Furthermore, the abundance of bacterial genes norB (r = 0.716 p < 0.01) and hao (r = 0.917 p < 0.01) was associated with higher N2O emission in high MC soils. Studies with the stable isotope (15N) revealed that 15N enrichment in N2O was primarily via denitrification pathways and labeled ammonium ion (15NH4+). Taken together, our results suggested that nitrifier denitrification is the main pathway generating N2O emission in soils with high MC, which may be caused by the high molar ratio of NH3 to total nitrogen in the influent.

Algae-induced taste and odour problems at low temperatures and the cold stress response hypothesis.

The existence of taste and odour (T&O) in drinking water is one of the principal causes of consumer complaints and is commonly related to algae growth. Numerous studies have confirmed the existence of algal blooms emerging specifically in low-temperature periods, herein referred to as "cold algae"; these include chrysophytes, cryptophytes, dinoflagellates and diatoms. In addition, the adaption mechanisms of these "cold algae" involve high flexibility in their nutrient intake and to the hydrological characteristics of the waters and their high contents of intracellular polyunsaturated fatty acids (PUFAs). Like algae proliferating in higher temperature waters, cold algae can also produce offensive odours. The potential dominant T&O compounds of low-temperature algae probably include saturated/unsaturated aldehydes and even some terpenoids. Among these, the polyunsaturated aldehydes (PUAs), the derivatives of polyunsaturated fatty acids, are the dominant T&O compounds and are probably synthesized during cell rupture. It was found that, for cold algae, low temperature may have a favourable effect on the generation of algae-induced T&O compounds. Furthermore, to better understand the internal mechanisms of algal T&O production, the stress response theory is introduced, which provides ideas for T&O control in raw water and in water treatment. Finally, implications for T&O management are given based on this review. KEY POINTS: • Like algae proliferating in higher temperature waters, cold algae can produce offensive odours. • Low temperatures may have a favourable effect on the generation of algae-induced T&O compounds. • The stress response theory can help to better understand the internal mechanisms of algal T&O production.

The influence mechanism of bioclogging on pollution removal efficiency of vertical flow constructed wetland.

The effect of change of hydraulic characteristic and microbial community on pollution removal efficiency of the infiltration systems in the bioclogging development process remain poorly understood. In this study, therefore, the pollutant removal as a response to hydraulic conductivity reduction and the change of diversity and structure of microbial communities in vertical flow constructed wetlands (VFCWs) was investigated. The results indicated that the richness and diversity of the bacterial communities in the columns at different depths were decreased, and the microbial communities of the genus level were changed in the process of bioclogging. However, the variation of microbial communities has a low impact on the purification performance of VFCWs because the abundance of function groups, respiratory activity, and degradation potentiality of microorganisms remain steady or even get improved in the columns after bioclogging. On the contrary, the hydraulic efficiency of VFCWs decreased greatly by 16.9%, 9.9%, and 57.1% for VFCWs filled with zeolite (Column I), gravel (Column II), and ceramsite (Column III), respectively. The existence of short-circuiting and dead zones in the filter media cause the poor pollution removal efficiency of VFCWs due to the short contact time and decrease of oxygenation renewal, as well as low activity in the dead zone.

A novel method: using an adenosine triphosphate (ATP) luminescence-based assay to rapidly assess the biological stability of drinking water.

The rapid and credible evaluations of the microbial stability of a drinking water distribution system (DWDS) are of great significance for ensuring the safety of drinking water and predicting microbial pollution. Conventional biostability assessment methods mainly focus on bacterial regrowth or evaluation of the level of nutrients that support bacterial regrowth. However, such methods are time-consuming and have many limitations. An adenosine triphosphate (ATP) assay can rapidly measure all active microorganisms and is known to be a useful method to assess the microbial activity of drinking water. The measurement of ATP has been used for more than a decade in the field of drinking water research. This article reviews the application of an ATP luminescence-based method to assess the biostability of drinking water and discusses the feasibility of ATP measurement as a parameter for quickly evaluating this criterion. ATP measurement will help researchers and water managers better monitor the biological stability of drinking water from the source to the consumer's tap. This review covers the: (1) principle and application of the ATP measurement in drinking water quality assessment; (2) comparison of the merits and demerits of several methods for evaluating the biostability of drinking water; (3) discussions on using ATP measurement in evaluating biostability; and (4) improvements in the use of ATP measurement in evaluating biostability. At the end of this review, recommendations were given for better application of the ATP measurement as a parameter for monitoring the microbial quality of drinking water.

Time-Series-Based Leakage Detection Using Multiple Pressure Sensors in Water Distribution Systems.

Leak detection is nowadays an important task for water utilities as leakages in water distribution systems (WDS) increase economic costs significantly and create water resource shortages. Monitoring data such as pressure and flow rate of WDS fluctuate with time. Diagnosis based on time series monitoring data is thought to be more convincing than one-time point data. In this paper, a threshold selection method for the correlation coefficient based on time series data is proposed based on leak scenario falsification, to explore the advantages of data interpretation based on time series for leak detection. The approach utilizes temporal varying correlation between data from multiple pressure sensors, updates the threshold values over time, and scans multiple times for a scanning time window. The effect of scanning time window length on threshold selection is also tested. The performance of the proposed method is tested on a real, full-scale water distribution network using synthetic data, considering the uncertainty of demand and leak flow rates, sensor noise, and so forth. The case study shows that the scanning time window length of 3-6 achieves better performance; the potential of the method for leak detection performance improvement is confirmed, though affected by many factors such as modeling and measurement uncertainties.

A Greedy Sampling Design Algorithm for the Modal Calibration of Nodal Demand in Water Distribution Systems.

This paper presents a greedy optimization algorithm for sampling design to calibrate WDS hydraulic model. The proposed approach starts from the existing sensors and sequentially adds one new sensor at each optimization simulation step. In each step, the algorithm tries to minimize the calibration prediction uncertainty. The new sensor is installed in the location where the uncertainty is greatest but also sensitive to other nodes. The robustness of the proposed approach is tested under different spatial and temporal demand distribution. We found that both the number of sensors and the perturbation ratio affect the calibration accuracy as defined by the average nodal pressure deviation itself and its variability. The plot of the calibration accuracy versus the number of sensors can reasonably guide the trade-off between model calibration accuracy and number of sensors placed or the cost. This proposed approach is superior in calibration accuracy and modeling efficiency when compared to the standard genetic algorithm (SGA) and Monte Carlo Sampling algorithm (MCS).

Refitted inclined plate for improving suspended solids removal in standard storm-water sumps.

Suspended solids (SS) in the storm-water makes up a significant source of total suspended solids in wet weather flow. With appropriate modification and maintenance, the standard sumps in the drainage system can remove SS from storm-water runoff as a best management practice device. To increase the removal efficiency, especially in the condition of high flow rate, inclined plates, based on the shallow pool sedimentation theory, have been designed and refitted to the sump. Its performance under the different surface load and flow rate were evaluated through scale models. The results show that the preliminary design, Model A, had limited removal efficiency, and even played a negative role sometimes due to the concentrated flow in the axis. The optimizations through installing a non-uniform porous baffle (Model B) and adopting inverted V-shaped plates (Model C) were improved, and results show that removal efficiency rate can be increased by around 15-20%, even at high flow rates. Moreover, too many plates cannot improve the removal rate further, because they make the cross-section decline and lead to higher velocity between plates.

Degradation of sulfonamides and formation of trihalomethanes by chlorination after pre-oxidation with Fe(VI).

Sulfonamides are used in human therapy, animal husbandry and agriculture but are not easily biodegradable, and are often detected in surface water. Sulfamethazine (SMZ) and sulfadiazine (SDZ) are two widely used sulfonamide antibiotics that are used heavily in agriculture. In this study, they were degraded in an aqueous system by chlorination after pre-oxidation with ferrate(VI) (FeVIO42-, Fe(VI)), an environmentally friendly oxidation technique that has been shown to be effective in degrading various organics. The kinetics of the degradation were determined as a function of Fe(VI) (0-1.5mg/L), free chlorine (0-1.8mg/L) and temperature (15-35°C). According to the experimental results, SMZ chlorination followed second-order kinetics with increasing Fe(VI) dosage, and the effect of the initial free chlorine concentration on the reaction kinetics with pre-oxidation by Fe(VI) fitted a pseudo-first order model. The rate constants of SDZ and SMZ chlorination at different temperatures were related to the Arrhenius equation. Fe(VI) could reduce the levels of THMs formed and the toxicity of the sulfonamide degradation systems with Fe(VI) doses of 0.5-1.5mg/L, which provides a reference for ensuring water quality in drinking water systems.

An ignored and potential source of taste and odor (T&O) issues-biofilms in drinking water distribution system (DWDS).

It is important for water utilities to provide esthetically acceptable drinking water to the public, because our consumers always initially judge the quality of the tap water by its color, taste, and odor (T&O). Microorganisms in drinking water contribute largely to T&O production and drinking water distribution systems (DWDS) are known to harbor biofilms and microorganisms in bulk water, even in the presence of a disinfectant. These microbes include T&O-causing bacteria, fungi, and algae, which may lead to unwanted effects on the organoleptic quality of distributed water. Importantly, the understanding of types of these microbes and their T&O compound-producing mechanisms is needed to prevent T&O formation during drinking water distribution. Additionally, new disinfection strategies and operation methods of DWDS are also needed for better control of T&O problems in drinking water. This review covers: (1) the microbial species which can produce T&O compounds in DWDS; (2) typical T&O compounds in DWDS and their formation mechanisms by microorganisms; (3) several common factors in DWDS which can influence the growth and T&O generation of microbes; and (4) several strategies to control biofilm and T&O compound formation in DWDS. At the end of this review, recommendations were given based on the conclusion of this review.

Heterogeneous oxidation of diclofenac in the presence of α-MnO2 nanorods: influence of operating factors and mechanism.

Diclofenac (DCF), one of the pharmaceutical and personal care products that has been widely detected in water, was selected as a model pollutant to evaluate the oxidation activity of α-MnO2 nanorods. The results showed that the heterogeneous oxidation process is highly pH dependent, with higher degradation efficiency at lower pH values. The complete removal of DCF was obtained within 80 min at the solution pH value of 2.5. The oxidation kinetics of DCF can be modeled by Langmuir-Hinshelwood equation (R2>0.999). The effects of various operating parameters, including initial solution pH, α-MnO2 dosage, anions, and cations, on the oxidation efficiency were investigated in detail. A possible reaction pathway for DCF was proposed. In addition, it was demonstrated that the α-MnO2 nanorods can be recycled without decreasing their oxidation activity after 10 cycles.

Effect of booster disinfection on the prevalence of microbial antibiotic resistance and bacterial community in a simulated drinking water distribution system.

Booster disinfection was often applied to control the microorganism's growth in long-distance water supply systems. The effect of booster disinfection on the changing patterns of antibiotic resistance and bacterial community was investigated by a simulated water distribution system (SWDS). The results showed that the antibiotic resistance bacteria (ARB) and antibiotic resistance genes (ARGs) were initially removed after dosing disinfectants (chlorine and chloramine), but then increased with the increasing water age. However, the relative abundance of ARGs significantly increased after booster disinfection both in buck water and biofilm, then decreased along the pipeline. The pipe materials and disinfectant type also affected the antibiotic resistance. Chlorine was more efficient in controlling microbes and ARGs than chloramine. Compared with UPVC and PE pipes, SS pipes had the lowest total bacteria, ARB concentration, and ARB percentage, mainly due to higher disinfectant residuals and a smoother surface. The significant correlation (rs = 0.77, p < 0.001) of the 16S rRNA genes was observed between buck water and biofilm, while the correlations of targeted ARGs were found to be weak. Bray-Curtis similarity index indicated that booster disinfection significantly changed the biofilm bacterial community and the disinfectant type also had a marked impact on the bacterial community. At the genus level, the relative abundance of Pseudomonas, Sphingomonas, and Methylobacterium significantly increased after booster disinfection. Mycobacterium increased after chloramination while decreased after chlorination, indicating Mycobacterium might resist chloramine. Pseudomonas, Methylobacterium, and Phreatobacter were found to correlate well with the relative abundance of ARGs. These results highlighted antibiotic resistance shift and bacterial community alteration after booster disinfection, which may be helpful in controlling potential microbial risk in drinking water.

Degradation of organic pollutants by the Cl-/PMS process.

The viewpoints on whether high concentrations of chloride ion (Cl-) promote or inhibit the oxidation activity of activated persulfates are still inconclusive. Furthermore, the degradation of organic pollutants by the persulfates in the presence of high Cl- concentrations without any activation medium has not yet been studied. In this work, the efficiency and mechanism of degradation of organic pollutants such as carbamazepine (CBZ), sulfadiazine (SDZ), and phenol (PN) by Cl--activated PMS (denoted as Cl-/PMS) were investigated. Results showed that Cl- could effectively activate PMS for the complete removal of CBZ, SDZ, and PN with reaction kinetic constants of 0.4516 min-1, 0.01753 min-1, and 0.06805 min-1, respectively. Parameters such as PMS dose, Cl- concentration, solution pH, and initial concentrations of organic pollutants that affect the degradation efficiencies of the Cl-/PMS process were optimized. Unlike conventional activated persulfates, it was confirmed that the free chlorine was the main active species in the Cl-/PMS process. Finally, the degradation by-products of CBZ and SDZ as well as their toxicity were detected, and a possible degradation pathway for CBZ and SDZ was proposed. Though higher toxic chlorinated by-products were generated, the Cl-/PMS process was still an efficient oxidation method for the removal of organic pollutants in aqueous solutions which contain high concentrations of Cl-.

An approximate solution for two-dimensional groundwater infiltration in sewer systems.

Estimating groundwater infiltration into sewer systems is important for wastewater treatment operators and municipalities. This paper presents an approximate solution for steady-state groundwater infiltration into sewer systems through line defects. The groundwater table was assumed to be horizontal and the aquifer homogeneous and isotropic. Mobius transformation technique and equivalent circumference method were introduced to solve the governing equation. The infiltration rate is found to be controlled by the hydraulic conductivity of the surrounding soil, the total hydraulic head above the sewer pipe, the size of the sewer pipe, the position of the defect, and the size of the defect.

Comparison of Uv/PDS and UV/H2O2 processes for the degradation of atenolol in water.

UV/H2O2 and UV/peroxodisulfate (PDS) processes were adopted to degrade a typical beta-blocker atenolol (ATL). The degradation efficiencies under various operational parameters (oxidant dosage, pH, HCO3-, humic acid (HA), NO3-, and Cl-) were compared. Principal factor analysis was also performed with a statistical method for the two processes. It was found that increasing the specific dosage of the two peroxides ([peroxide]0/[ATL]0) ranging from 1:1 to 8:1 led to a faster degradation rate but also higher peroxide residual. Within the pH range 3-11, the optimum pH was 7 for the UV/PDS process and elevating pH benefitted the UV/H2O2 process. The presence of HCO3-, HA, and Cl- adversely affected ATL oxidation in both processes. The NO3- concentration 1-3 mmol/L accelerated the destruction of ATL by the UV/PDS process, but further increase of NO3- concentration retarded the degradation process, contrary to the case in the UV/H2O2 process. The rank orders of effects caused by the six operational parameters were pH approximately specific dosage > [HA]0 > [NO3-]0 > [HCO3-]0 > [Cl-]0 for the UV/H2O2 process and specific dosage > pH > [HA]0 > [NO3-]0 > [HCO3-]0 > [Cl-]0 for the UV/PDS process. The UV/PDS process was more sensitive to changes in operational parameters than the UV/H2O2 process but more efficient in ATL removal under the same conditions.

Characterization of typical taste and odor compounds formed by Microcystis aeruginosa.

Production and characteristics of typical taste and odor (T&O) compounds by Microcystis aeruginosa were investigated. A few terpenoid chemicals, including 2-MIB, beta-cyclocitral, and beta-ionone, and a few sulfur compounds, such as dimethyl sulfide and dimethyl trisulfide, were detected. beta-Cyclocitral and beta-carotene concentrations were observed to be relevant to the growth phases of Microcystis. During the stable growth phase, 41-865 fg/cell of beta-cyclocitral were found in the laboratory culture. beta-Cyclocitral concentrations correlated closely with beta-carotene concentrations, with the correlation coefficient R2 = 0.96, as it is formed from the cleavage reaction of beta-carotene. For dead cell cases, a high concentration of dimethyl trisulfide was detected at 3.48-6.37 fg/cell. Four T&O compounds, including beta-cyclocitral, beta-ionone, heptanal and dimethyl trisulfide, were tested and found to be able to inhibit and damage Microcystis cells to varying degrees. Among these chemicals, beta-cyclocitral has the strongest ability to quickly rupture cells.