Daily Dissolved Oxygen Research Articles

An efficient data fusion model based on Bayesian model averaging for robust water quality prediction using deep learning strategies

Accurate monitoring of dissolved oxygen (DO) levels is critical for stakeholders to effectively safeguard water resources and aquatic ecosystem health. This research presents an innovative data fusion framework based on Bayesian model averaging (BMA) by the combination of several neuroscience models (deep learning methodologies) including multilayer perceptron neural network (MLPNN), recurrent neural network (RNN), convolutional neural network (CNN), gated recurrent unit (GRU), long short-term memory (LSTM), and seasonal autoregressive integrated moving average with exogenous variables (SARIMAX). BMA has the capacity to greatly enhance the results attained by standalone approaches. In this study, two feature selection methods such as mutual information (MI) and recursive feature elimination (RFE) applied to select effective predictors and investigate the importance of each input parameters. The techniques were evaluated based on four different metrics and, finally, to demonstrate the usefulness and effectiveness of the newly implemented strategy, it was applied proficiently at the USGS stations, 01427510 and 02336152 within the USA. The findings from analyzing data based on two stations confirmed that the suggested approach (BMA) outperformed other methods such as MLPNN, RNN, CNN, LSTM, and SARIMAX when it came to predicting the levels of DO on a daily basis. In terms of RMSE, MAE and R2, BMA yielded 0.272 mg/L, 0.216 mg/L, and 0.975 using MI technique and 0.320 mg/L, 0.261 mg/L, and 0.965 using RFE method at 01427510 USGS station, respectively. Similarly, based on RMSE, MAE and R2, BMA produced DO prediction by RMSE = 0.352 mg/L, 0.264 mg/L, and 0.968 by MI approach and RMSE = 0.378 mg/L, 0.282 mg/L, and 0.963 via RFE process at 02336152 USGS station, respectively. After analyzing different combinations of input that obtained from feature selection paradigms, it was observed that the variables with the greatest impact on daily dissolved oxygen levels are the dissolved oxygen levels in the previous time period and the water temperature. On the other hand, the pH and turbidity have minimal influence on the daily DO. Finally, the results of this study confirmed that the BMA tool can be efficiently applied to predict DO concentration based on deep learning model outputs.

Hybrid machine learning models for prediction of daily dissolved oxygen

Measuring water quality parameters is a significant step in many hydrological assessments. Dissolved oxygen (DO) is one of these parameters that is an indicator of water quality. Hence, this study offers two novel intelligent models, i.e., the integration of biogeography-based optimization (BBO) and atom search optimization (ASO) with artificial neural network (ANN), to predict the daily DO. These methods are comparatively assessed and validated against several benchmark techniques. Five-year (2014–2019) water quality data of a USGS station called Rock Creek (Station number 01648010) is used for implementing the proposed model. In this sense, the models first learn the DO behavior using 80 % of the data and they then predict the DO for the fifth year. As per the performed sensitivity analysis, the water temperature was selected as the most effective parameter in the DO prediction. Trying different population sizes determined an optimal configuration of the employed models and assessing the accuracy of the results revealed that the proposed models can nicely perceive the DO pattern with around 4 % mean absolute percentage error (MAPE) and 97.5 % correlation. In the testing phase, the BBO-ANN and ASO-ANN models predicted the DO of the fifth year with MAPEs 2.3848 and 2.5170 %, and correlations of 0.99186 and 0.99135, respectively. Moreover, the suggested BBO-ANN and ASO-ANN outperformed some similar hybrids from the existing literature. Lastly, an explicit formula is derived from the BBO-ANN for convenient prediction of the DO.

Predicting river dissolved oxygen time series based on stand-alone models and hybrid wavelet-based models

Accurate prediction of dissolved oxygen time series is important for improving the water environment and aiding water resource management. In this study, four stand-alone models including multiple linear regression (MLR), support vector machine (SVM), artificial neural network (ANN) and random forest (RF), and four hybrid models based on wavelet transform (WT) including WT-MLR, WT-SVM, WT-ANN and WT-RF were used to predict the daily dissolved oxygen (DO) at 1–5-day lead times in the Dongjiang River Basin, China. To make the prediction robust, the maximal information coefficient (MIC) was used to capture comprehensive information between DO and explanatory variables. The 5-fold cross validation grid search approach was used to optimize parameters of machine learning tools. Two types of frameworks of WT: direct framework (i.e., only the explanatory variables were decomposed) and multicomponent framework (i.e., both explanatory variables and target variables were decomposed) were used to construct hybrid models. The results show that MIC extracts four optimal explanatory variables: previous DO, water temperature, air temperature and air pressure. Four evaluation parameters including correlation coefficient (R), Nash-Sutcliffe efficiency (NSE), mean absolute error (MAE) and root mean square error (RMSE) indicate that the prediction accuracy decreases as the lead time changes from 1 to 5 days. In terms of the stand-alone models, MLR model outperforms the other three models with higher NSE values of 0.616–0.921, and lower RMSE values of 0.503–1.111. With regard to the hybrid models, WT-ANN and WT-MLR models exhibit higher performance, and multicomponent framework performs better than direct framework in all hybrid models. In general, the multicomponent framework of WT can improve the prediction accuracy of stand-alone models at a certain degree, while the direct framework shows no obvious advantage.

Prediction of dissolved oxygen in urban rivers at the Three Gorges Reservoir, China: extreme learning machines (ELM) versus artificial neural network (ANN)

AbstractIn the present study, two non-linear mathematical modelling approaches, namely, extreme learning machine (ELM) and multilayer perceptron neural network (MLPNN) were developed to predict daily dissolved oxygen (DO) concentrations. Water quality data from four urban rivers in the backwater zone of the Three Gorges Reservoir, China were used. The water quality data selected consisted of daily observed water temperature, pH, permanganate index, ammonia nitrogen, electrical conductivity, chemical oxygen demand, total nitrogen, total phosphorus and DO. The accuracy of the ELM model was compared with the standard MLPNN using several error statistics such as root mean squared error, mean absolute error, the coefficient of correlation and the Willmott index of agreement. Results showed that the ELM and MLPNN models perform well for the Wubu River, acceptably for the Yipin River and moderately for the Huaxi River, while poor model performance was obtained at the Tributary of Huaxi River. Model performance is negatively correlated with pollution level in each river. The MLPNN model slightly outperforms the ELM model in DO prediction. Overall, it can be concluded that MLPNN and ELM models can be applied for DO prediction in low-impacted rivers, while they may not be appropriate for DO modelling for highly polluted rivers.This article has been made Open Access thanks to the kind support of CAWQ/ACQE (https://www.cawq.ca).

Development and evaluation of a real-time forecasting framework for daily water quality forecasts for Lake Chaohu to Lead time of six days

The socioeconomic benefits associated with informative water quality forecasts for large lakes are becoming increasingly evident. However, it remains an enormous challenge to produce forecasts of water quality variables that are accurate enough to meet public demand. In this study, we developed and evaluated a new forecast framework for real-time forecasting of daily dissolved oxygen (DO), ammonium nitrogen (NH), total phosphorus (TP) and total nitrogen (TN) concentrations at lead times from one to six days for Lake Chaohu, the fifth largest freshwater lake in China. The forecast framework is based on a 3-D hydrodynamic ecological model referred to as EcoLake. We used hydrological, meteorological and water quality data from multiple sources to generate initial conditions and forcing functions. Solar radiation and inflows from tributaries which are not readily available were calculated using forecasted cloud cover and rainfall. Forecast skill was evaluated based on 122 forecasts produced on different days in 2017 and for each of the 12 sampling sites. Results indicate that the skill of the forecast framework varies considerably across water quality variables, sampling sites, and lead times. Generally, the forecast framework is more skillful than the persistence forecasts, which use the most recent observations as forecasts. The TN forecasts tend to be the most skillful with a mean RMSE skill score of 28.5% averaged across the six lead times. The DO forecasts tend to have the lowest skill with an average value of 10.9%. Model sensitivity experiments further revealed that errors in the raw air temperature and wind speed forecasts have a noticeable impact on the overall skill of DO and NH forecasts. The forecast framework proposed here could be a useful operational forecasting tool to enhance the effectiveness of the drinking water supply and public health protection based on the water quality management of Lake Chaohu.

Stream response to an extreme drought-induced defoliation event

We assessed stream ecosystem-level response to a drought-induced defoliation event by gypsy moth caterpillars (Lymantria dispar) with high-frequency water quality sensors. The defoliation event was compared to the prior year of data. Based on long-term records of precipitation and drought indices, the drought of 2015–2016 in Rhode Island, USA was an extreme climatic event that preceded and likely precipitated the defoliation from insect infestation. Canopy cover in the riparian area was reduced by over 50% increasing light availability which warmed the stream and stimulated autotrophic activity. Frass and leaf detritus contributed particulate carbon and organic nutrients to the stream. Based on locally calibrated s::can spectro::lyser data, nitrate concentration and flux did not significantly increase during defoliation while orthophosphate concentration and flux did significantly increase during part of the defoliation period. Lower mean daily dissolved oxygen (DO) levels and wider diel cycles of DO indicated higher biological activity during the defoliation event. Stream metabolism metrics were also significantly higher during defoliation and pointed to heterotrophic activity dominating in the stream. The increases in stream metabolism were low compared to other studies; in streams with higher nutrient levels (e.g., in agricultural or urban watersheds) the increase in light and temperature could have a stronger influence on stream metabolism. The in-stream metabolic processes and nutrient fluxes observed in response to the drought-driven defoliation event resulted from the long-term deployment of high-frequency water sensors. The proliferation of these water sensors now enable studies that assess ecosystem responses to stochastic, unusual disturbances.

Modelling daily dissolved oxygen concentration using least square support vector machine, multivariate adaptive regression splines and M5 model tree

In the present study, three types of artificial intelligence techniques, least square support vector machine (LSSVM), multivariate adaptive regression splines (MARS) and M5 model tree (M5T) are applied for modeling daily dissolved oxygen (DO) concentration using several water quality variables as inputs. The DO concentration and water quality variables data from three stations operated by the United States Geological Survey (USGS) were used for developing the three models. The water quality data selected consisted of daily measured of water temperature (TE, °C), pH (std. unit), specific conductance (SC, μS/cm) and discharge (DI cfs), are used as inputs to the LSSVM, MARS and M5T models. The three models were applied for each station separately and compared to each other. According to the results obtained, it was found that: (i) the DO concentration could be successfully estimated using the three models and (ii) the best model among all others differs from one station to another.

Modeling daily dissolved oxygen concentration using modified response surface method and artificial neural network: a comparative study

In the present study, two nonlinear mathematical modeling approaches, namely modified response surface method (MRSM) and multilayer perceptron neural network (MLPNN) were developed and compared for modeling daily dissolved oxygen (DO) concentration. The DO concentration and water quality variables data for several years, available from four stations operated by the United States Geological Survey, were used for developing the two models. The water quality data selected consisted of daily measured river discharge, water pH, specific conductance, water turbidity, and DO. The response surface methodology is modified based on the two steps for calibrating process. In the first regression step, the normalized input data were calibrated based on a linear function and then transferred by an inverse power function. In the second regression step, the input data from first step were used to calibrate a highly nonlinear third-order polynomial function. The accuracy of the proposed nonlinear MRSM is compared with the standard MLPNN using several error statistics such as root-mean-square error, mean absolute error, mean bias error, the coefficient of correlation, the Nash–Sutcliffe efficiency, and the Willmott index of agreement. The results obtained indicate that MRSM model performed best in comparison with the MLPNN for the all four stations.

Comparing A Bayesian and Fuzzy Number Approach to Uncertainty Quantification in Short-Term Dissolved Oxygen Prediction

A new autoregressive-type, updating fuzzy linear regression method is proposed to predict daily dissolved oxygen (DO) concentration in a highly urbanized riverine environment. Results of this model are compared to results from an updating Bayesian regression model. Both methods use lagged daily DO (at four different lags) as the independent variable. Uncertainty in the models is represented by a fuzzy number based approach in the first case, and by a Bayesian framework in the second. Real-time data from the Bow River in Calgary, Canada is used to calibrate the models sequentially to mimic a real-time updating model. Four different performance metrics were used to measure the performance of each model. Lastly, the input data resolution is reduced to measure the impact on model performance. Results show that the physical system can be adequately characterized using only one year of data. Both approaches can capture the general trend of daily DO, but the fuzzy number based method can better capture the changes in observed variability. The metrics for both models are comparable, with the one-day lag case categorized as “very good”; however, the performance reduces at higher lags. The fuzzy number method captures more low DO events than the Bayesian approach, with a much lower mean squared error. A possibility to probability transformation is used to highlight the risk of low DO days for the fuzzy case. Lastly, reducing the input data resolution from 96 to 6 points per day has a minimal impact on model performance, suggesting the limited efficacy or utility in increasing sampling rates.

Predicting river aquatic productivity and dissolved oxygen before and after dam removal

The potential changes to dissolved oxygen concentrations caused by combined sewer overflows (CSOs) are an important consideration before and after dam removal in many urban rivers. An old 2-m-high dam was recently (autumn 2012) removed in Columbus, Ohio, USA, from the fourth-order urban Lower Olentangy River, which has several major CSOs. Two simulation models were developed using STELLA® 9.1.4 to investigate the interactions of hydrology, water quality, and CSOs in the Olentangy River reservoir before removal of the Fifth Avenue Dam. The first simulation model, focusing on daily dissolved oxygen (DO), was created to predict DO changes in the pollution found in CSOs before and then 120h after dam removal. Field DO data in 2009, 2011, and 2012 were used to calibrate and validate the model. The second simulation model, a weekly water quality model, was created to simulate one-year streamflow, gross primary productivity (GPP), and DO changes before dam removal, and to predict long-term water quality changes 2 years after dam removal. Field streamflow and water quality data from 2004 and 2009 were used to calibrate that model. Results from both field data analysis and model simulations suggest that there are great differences in water quality between base flow and flooding periods and that DO concentrations will continue to be significantly impacted by CSO discharges even after dam removal.

A Method and Rationale for Deriving Nutrient Criteria for Small Rivers and Streams in Ohio

A mechanistic understanding of the effects of nutrient enrichment in lotic systems has been advanced over the last two decades such that identification of management thresholds for the prevention of eutrophication is now possible. This study describes relationships among primary nutrients (phosphorus and nitrogen), benthic chlorophyll a concentrations, daily dissolved oxygen (DO) concentrations, and the condition of macroinvertebrate and fish communities in small rivers and streams in Ohio, USA. Clear associations between nutrients, secondary response indicators (i.e., benthic chlorophyll and DO), and biological condition were found, and change points between the various indicators were identified for use in water quality criteria for nutrients in small rivers and streams (<1300 km(2)). A change point in benthic chlorophyll a density was detected at an inorganic nitrogen concentration of 0.435 mg/l (+/-0.599 SD), and a total phosphorus (TP) concentration of 0.038 mg/l (+/-0.085 SD). Daily variation in DO concentration was significantly related to benthic chlorophyll concentration and canopy cover, and a change point in 24-h DO concentration range was detected at a benthic chlorophyll level of 182 mg/m(2). The condition of macroinvertebrate communities was related to benthic chlorophyll concentration and both minimum and 24-h range of DO concentration. The condition of fish communities was best explained by habitat quality. The thresholds found in relationships between the stressor and the response variables, when interpreted in light of the uncertainty surrounding individual change points, may now serve as a framework for nutrient criteria in water quality standards.

Production Responses of Channel Catfish to Minimum Daily Dissolved Oxygen Concentrations in Earthen Ponds

AbstractThis study determined the effects of the minimum daily dissolved oxygen (DO) concentration on the production parameters of channel catfish Ictalurus punctatus in earthen ponds. Fifteen 1‐acre ponds (five ponds per treatment) were managed as high‐oxygen (minimum DO concentrations averaging 4.37 ppm or 54% air saturation from June through September), medium‐oxygen (minimum DO concentrations averaging 2.68 ppm or 33.2% air saturation), or low‐oxygen treatments (minimum DO concentrations averaging 2.32 ppm or 28.7% air saturation) using one 5‐hp electric paddlewheel aerator per pond. Fish in the high‐, medium‐, and low‐oxygen treatment ponds were fed a mean total of 14,008, 13,212, and 12,607 lb/acre of 28%‐protein floating feed, respectively. Net production paralleled the total amount of feed fed, averaging 5,772, 5,278, and 5,113 lb/acre in the high‐, medium‐, and low‐oxygen treatments, respectively. Individual fish weight at harvest also showed a similar trend, averaging 1.37, 1.33, and 1.30 lb in the high‐, medium‐, and low‐oxygen treatments, respectively. No visible stress responses were observed in any ponds during this study. Total aeration averaged 5,245, 2,518, and 1,337 hp‐h/acre in the high‐, medium‐, and low‐oxygen treatments, respectively. Treatments with higher minimum DO concentrations had significantly higher nitrite‐nitrogen, suspended solids, chlorophyll a, and pH and lower Secchi disk visibility, alkalinity, and hardness; however, no water quality parameters exceeded the normal acceptable range for channel catfish. While the cost of electricity must be considered, maintaining a minimum daily DO concentration of 2.3–2.5 ppm is suggested as a compromise between maximizing both the amount of feed fed and fish production while minimizing aeration costs.

Dissolved oxygen fluctuations in organically and inorganically fertilized walleye ( Stizostedion vitreum) hatchery ponds

We compared autotroph productivity and fish production in walleye ( Stizostedion vitreum) hatchery ponds under two fertilization regimes. Three ponds were fertilized with inorganic N and P (experimental) and three ponds were fertilized with a combination of organic chopped hay, soybean meal, alfalfa meal, and inorganic N and P (control). Daily dissolved oxygen (DO) changes were monitored at three depths. Chlorophyll a concentrations and Secchi depth measurements were taken twice per week. Mean DO concentrations were significantly lower ( P<0.0001, n=684) in the control ponds. Average productivity, calculated by either dawn–dusk–dawn DO changes, or diel DO changes measured at 15-min intervals, were not significantly different between the treatments ( P>0.05) however, chlorophyll a concentrations were significantly higher in control ponds (19.21 vs. 7.09 μg/l in experimental ponds). Secchi depth did not correlate with algal productivity or biomass. Higher alkalinity did not increase algal production or biomass. Even with cooler than normal temperatures, one of the control ponds required aeration to avoid a fish kill. Overall, fish yield and survival were higher in control ponds, suggesting that some other factors associated with organic fertilizers (increased protozoan production, slower release of nutrients) may make them necessary for optimum yields. Adjusting the ratio of organic to inorganic fertilizer could reduce cost and anoxia without reducing fish production.

RESEARCH: Projected Climate Change Effects on Winterkill in Shallow Lakes in the Northern United States.

/ Each winter, hundreds of ice-covered, shallow lakes in the northern United States are aerated to prevent winterkill, the death of fish due to oxygen depletion under the ice. How will the projected climate warming influence winterkill and the need to artificially aerate lakes? To answer this question, a deterministic, one-dimensional year-round water quality model, which simulates daily dissolved oxygen (DO) profiles and associated water temperatures as well as ice/snow covers on lakes, was applied. Past and projected climate scenarios were investigated. The lake parameters required as model input are surface area, maximum depth, and Secchi depth as a measure of radiation attenuation and trophic state. The model is driven by daily weather data. Weather records from 209 stations in the contiguous United States for the period 1961-1979 were used to represent past climate conditions. The projected climate change due to a doubling of atmospheric CO(2) was obtained from the output of the Canadian Climate Center General Circulation Model. To illustrate the effect of projected climate change on lake DO characteristics, we present herein DO information simulated, respectively, with inputs of past climate conditions (1961-1979) and with a projected 2 x CO(2) climate scenario, as well as differences of those values. Specific parameters obtained were minimum under-ice and lake bottom DO concentration in winter, duration of under-ice anoxic conditions (<0.1 mg/liter) and low DO conditions (<3 mg/liter), and percentage of anoxic and low DO lake volumes during the ice cover period. Under current climate conditions winterkill occurs typically in shallow eutrophic lakes of the northern contiguous United States. Climate warming is projected to eliminate winterkill in these lakes. This would be a positive effect of climate warming. Fish species under ice may still experience periods of stress and zero growth due to low DO (<3 mg/liter) conditions under projected climate warming.

Simulated climate change effects on dissolved oxygen characteristics in ice-covered lakes

A deterministic, one-dimensional model is presented which simulates daily dissolved oxygen (DO) profiles and associated water temperatures, ice covers and snow covers for dimictic and polymictic lakes of the temperate zone. The lake parameters required as model input are surface area ( A S), maximum depth ( H MAX), and Secchi depth as a measure of light attenuation and trophic state. The model is driven by daily weather data and operates year-round over multiple years. The model has been validated with extensive data (5976 points). Standard error between simulated and measured dissolved oxygen is 1.9 mg/l. The model is applied to simulate effects of climate change on dissolved oxygen characteristics of 27 lake classes in Minnesota. The projected climate changes due to a doubling of atmospheric CO 2 are obtained from the output of the Canadian Climate Center Global Circulation Model (CCC GCM). Climate change delays the ice formation and shortens the ice cover period. Winter anoxia, even in shallow lakes, therefore disappears under a projected 2 × CO 2 climate condition. This eliminates winterkill in these lakes. Herein, the simulated DO characteristics have been plotted and interpolated graphically in a coordinate system with a lake geometry ratio ( A 0.25 S H MAX ) on one axis and Secchi depth on the other. The lake geometry ratio expresses a lake's susceptibility to stratification. To illustrate the effect of projected climate change on DO characteristics, separate graphs are presented for values simulated with inputs of past climate conditions (1961–1979) and with a projected 2 × CO 2 climate scenario.

Model simulations of dissolved oxygen characteristics of Minnesota lakes: Past and future

A deterministic, one-dimensional, unsteady numerical model has been developed, tested, and applied to simulate mean daily dissolved oxygen (DO) characteristics in 27 lake classes in the state of Minnesota. Reaeration and photosynthesis are the oxygen sources, while respiration, sedimentary, and biochemical water column oxygen demand are the sinks of oxygen in the model. The lake classes are differentiated by surface area (A s), maximum depth (H max), and trophic status expressed as Secchi depth (Z s). Because lake stratification is most important to lake oxygen dynamics, simulated DO characteristics are plotted in terms of a stratification parameterA s/H max 0.25 and Secchi depthZ s. Simulations provide DO profiles on a daily time scale. Specific DO characteristics of ecological and environmental interest are epilimnetic DO, hypolimnetic DO, DO gradient from surface to bottom, and DO minima and maxima. Specific results are as follows: Simulated mean daily and weekly DO values in the epilimnion of all lakes for both past and future climate scenarios are near saturation over the summer season. Hypolimnetic DO values depend strongly on lake morphometry, trophic status, and time throughout the summer season. Future climate conditions are specified as the historical records from 1955 to 1979, adjusted (monthly) by the 2 × CO2 GISS model output to account for doubling of atmospheric CO2. With this climate change, weekly averaged epilimnetic DO is projected to drop by less than 2 mg/liter, and will remain above 7 mg/liter throughout the open water season. The hypolimnetic DO reductions after climate change are on the order of 2–8 mg/liter. Periods of anoxia are longer by as much as 80 days. Those changes would alter water quality dynamics in lakes and have a profound effect on lake ecosystems including indigenous fishes. The results presented are useful for evaluating environmental management options.